PROCEEDINGS OF FULL PAPERS. 2 nd Conference on Sustainable Urban Mobility

Transcription

1 PROCEEDINGS OF FULL PAPERS 2 nd Conference on Sustainable Urban Mobility 5 6 May, 2014 Volos Greece Department of Planning and Regional Development, University of Thessaly Under the auspices of the Hellenic Ministry of Infrastructure, Transport and Networks within the framework of the Hellenic Presidency of the EU Council

2 2nd CONFERENCE ON SUSTAINABLE URBAN MOBILITY 5-6 MAY, 2014, VOLOS, GREECE PROCEEDINGS OF FULL PAPERS

3 TABLE OF CONTENTS SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT An Advanced Traveler Information System for Informing Tourists and Promoting Public Transport... 1 Prioritization of Smart Mobility Solutions in the City of Volos SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT A Multi-Periodic Optimization Modeling Approach for the Establishment of a Bike- Sharing Network: a Case-Study of the City of Athens Research on the Development potential of Cycling routes Connecting Green Open Spaces in Athens Shift from Car Cities to Soft Mode Cities? Learning from Best and Worst Practice? 65 SESSION: PUBLIC TRANSPORT PLANNING The Future Search Methodology as a Tool for the Development of a Sustainable Urban Mobility Plan in the Region of Central Macedonia Integrating Urban Vehicle Characteristics into Sustainable Transportation Planning 95 Public Transport Emissions Measurement and Reactive Measures Designing Sustainable Urban Transport Interchanges SESSION: MODELLING PUBLIC TRANSPORT Applications of Traffic Distribution using the Software VISUM and more Environmental Capacity Utilization Foundation and Implementation of a New Composite Accessibility Measure in Individual Level A Descriptive Study on Public Transport User Behaviour from Live Bus Arrivals. 149 SESSION: TRANSIT QUALITY PERFORMANCE Sustainable Urban Mobility Indicators for Medium-Sized Cities. The Case of Serres, Greece Spatio-Temporal Parameters Affecting Sustainable Urban Mobility in Greek Medium Sized Cities: The Case of Volos Environmental Impact Assessment of Urban Public Transport Systems: a Scenario Study for the Northwest Part of City of Volos A Decision Tree Application in Transit Quality of Service in the City of Volos

4 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT

5 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT An Advanced Traveler Information System for Informing Tourists and Promoting Public Transport Maria Morfoulaki a, Kornilia Kotoula a*, Katerina Chrysostomou a, Glykeria Myrovali a, Bartholomew Michael Vassilantonakis a a Hellenic Institute of Transport (HIT), Thessaloniki, Greece *Corresponding Author: nilia@certh.gr, , fax: Abstract Nowadays increasingly more tourists prefer Public Transport for their daily trips in urban, rural and island areas. This fact has changed modal share during touristic periods and positively affected both traffic and environmental conditions. A key parameter for succeeding this shift to more sustainable tourism is the upgrade of offered information and the development of new electronic services easily accessible during the pre- or the en-route phase of the trip. Since 2007, when Bulgaria joined the EU, cross-border traffic between Greece and Bulgaria increased - mainly during holidays and weekends- due to leisure trips to/from major tourist destinations in both countries. The use of private car, being the main mode for leisure trips, together with the lack of adequate information provision to the travelers, either at pre-trip or at en-route level, generated a lot of accessibility and environmental problems to both countries. The present paper presents the development of an Advanced Traveler Information System, EasyTrip, which offers to all travelers of Northern Greece and Southern Bulgaria advanced personalized e-mobility services (including information about all available transport modes, connectivity between urban and interurban public transport systems, route guidance etc), in order to enhance accessibility in the areas and connectivity between the two countries, promoting sustainable mobility, tourism and development. The development of the above mentioned services was based on the travelers needs as they have been identified by the focused target group studies that have been executed and followed the relevant EC standards, state-of-the-art techniques and route optimization algorithms that were developed. Keywords: Information communication technologies, public transport systems, sustainable mobility, sustainable tourism. 1. INTRODUCTION Transport is a key necessity to all human activities and as part of it, transportation systems role is vital for tourism industry. Most visitors arriving in new destinations need and should be informed about all transport modes available to serve their needs and requirements, as only few of them wish or can even afford to hire private transport. Therefore, the application of Information and Communication Technology systems (ICT) in the sector of Transport is of great significance as it provides to 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 1

6 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT all potential users valuable information to relevant mobility issues. Consequently, understanding and adjusting services of public transport systems to tourists needs may confer additional benefits to the area of influence, such as increase in the number of tourists, enhancement of accessibility and equal opportunities for economic development. The present paper aims to present the Advanced Traveler Information System (ATIS) that has been developed in the framework of EasyTrip: GR-BG E-Mobility solutions project that was funded by the European Territorial Cooperation Programme Greece-Bulgaria in order to offer to all travelers of Northern Greece and Southern Bulgaria advanced and personalized electronic mobility services. The paper is structured as follows: following the present introductory part, the second part presents an overview of problems daily being faced by travelers, while the third part presents the methodology that has been followed. The fourth part presents the system architecture, the next one the services that are provided through the platform, the sixth one the data and information requirements for the system development and the seventh one the algorithms and processes that have been used. Finally, the last part contains the conclusions and future expansion opportunities for the system. 2. THE PROBLEM Transport is considered as key factor in the success of sustainable tourism development (Gossling et al., 2009; Page & Connell, 2009). According to Kaul (1985) transportation systems are significant for the development of tourism activities, while Page (2005) highlights the fact that that there is a strong connection between transport and tourism industries. In addition, Albalate & Germa (2010), confirm the close links between these two industries, based on econometric estimations. Hall (1999) argues that transportation systems should offer easy access to tourists travelling within a wider destination area, while Crouch and Ritchie (1999) note that the touristic development depends on the provided infrastructure and facilities regarding the field of transport. Duval (2007) emphasizes the importance of accessibility is such that the ability of a destination to attract tourists is largely contingent on the availability and efficiency of transport needed to travel to that destination. Many research activities have also concluded to the fact that nowadays tourists prefer Public Transport modes for their daily trips in rural, urban and island areas (Currie & Falconer, 2013). However, information is critical for the increase of public transport share and existing sources of information are usually fragmented and it is a difficult task for prospective passengers to access the right information. Timetables are not available from most operators, and where they are available, the quality is poor or difficult to comprehend. The number of daily trips that take place among European countries is increasing, mainly due to open borders between the countries of the European Union, which now allow the free movement of commuters. For example, since 2007, when Bulgaria joined the EU, cross-border traffic between Greece and Bulgaria increased up to 50,000 vehicles / day (ca. 130,000 passengers / day). This increase was observed mainly 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 2

7 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT during the holidays and weekends, due to leisure trips to/from major tourist destinations in both countries. The use of private car as the main mode for these trips has generated a lot of accessibility and environmental problems to both countries. The main question was whether the transportation systems were ready to accept all the above mentioned touristic flows and how the accessibility between the two countries could be improved. The absence of adequate information provision to the travelers between Greece and Bulgaria, either at pre- trip or at en-route level, regarding mobility related issues such as alternative transport means, connectivity between urban and interurban public transport systems both in regional and international level, exact route guidance etc, was the missing point in order to achieve a better accessibility and connectivity between cross-border areas, affecting sustainable tourism, sustainable development, minimization of traffic effects, equal opportunities and fair competition within the entire area. Therefore, the provision of personalized e-mobility services through the Easytrip e- Services platform, using advanced ICT systems and spreading the information to all travelers through channels easily reachable by everyone (mobile phone applications, internet, VMS equipment etc) is being analyzed in the present paper. 3. COVERING THE TOURIST S NEED FOR PUBLIC TRANSPORT (PUT) SYSTEMS INFRORMATION In order to identify the needs of frequent and non-frequent travelers between Greece and Bulgaria, a targeted use needs survey took place. The survey focused on visitors of the areas of interest and was conducted in two different time periods covering both the winter and the summer peak period trips. Both surveys took place at Promachonas border station and the target groups were during the summer, Bulgarian visitors of Greece and during the winter period, Greek travelers to Bansko. The main desired characteristic of the respondents was their familiarity with internet services and/or smartphone applications. In total, 1200 questionnaires were completed. The visitors needs analysis indicated that for their main trip purpose being tourism and leisure, the main selection criterion for mode choice is comfort. Moreover the main selection criterion for route choice is that they are not aware of any other alternative choice (summer period) or that this is the most suitable one (winter period). Most users travel 1 to 10 times per year during winter time and rarely during the summer time to Bulgaria and Greece respectively (Fig. 1). 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 3

8 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Figure 1. Analysis of traveler s profile Regarding the services that travelers between Greece-Bulgaria are interested in, these mainly are; points of interest, parking and road charges information but also alternative routes and modes to their destination (Fig. 2). As a general result the services that the travelers need are these that provide information that is are difficult to find or to be aware of when you are visiting another country. For instance, the Greek road charging system includes road tolls with specific fee, while the Bulgarian system includes the use of vignette, something that most of the visitors are not aware of. Figure 2. Analysis of travelers preferences regarding the e-services that they need to be provided The results of the travelers needs and preferences surveys were used as the basis in order to design the Easytrip E-Mobility solutions system. The main challenge for the design team of the Easytrip system was to cover all the above mentioned requirements, by collecting the appropriate data and information (which was not offered through any other electronic source and was crucial for the provision of the info-services), to develop multimodal routing algorithms in order to inform about optimal routes and 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 4

9 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT modes in urban, national and international level (which required detailed recording of the road and PuT network and its operational characteristics) and finally to design a user friendly interface in order to provide these services through an web-based platform but also through a smartphone application. 4. SYSTEM ARCHITECTURE The primary goal was to develop services that are addressed to all users and are simple, easy and quick to use. The Easytrip E-mobility services are designed in such a way that they easily provide the user with condensed and useful information taking into account its location (location based services) and based on default settings. Moreover, users also have the opportunity to navigate through the 'advanced options ' in order to get information for other areas included in the Easytrip project and according to other criteria. The services are provided in the form of user-friendly, personalized web services and mobile applications for pre-trip and en-route planning and information. The Easytrip platform has been developed mainly based on the n-tier architecture model which provides separated steps of development and processes which can coexist at the final level of service. This architecture provides a flexible and distributable environment on the development/deployment level as well as on the operational level of the whole integrated system. This approach provides three layerlevels functionalities (Fig. 3) which are: Data layer o The data layer hosts all the data storages processes as well as the appropriate procedures and triggers to store and provide the data to the appropriate format and model. Business layer o The business or application layer provides all the necessary tools/functions for the communication and exchange of the necessary data between the data and presentation layer. It also hosts all algorithmic processes needed in order to provide the services. Presentation layer o The presentation layer supports all the functionalities that are needed between the users and the system. It also provides all content and tools/services to the public. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 5

10 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Figure 3. Platform Architecture Each of the above layers performs specific tasks and processes so either the expansion and maintenance of any service hosted by the Easytrip platform or the whole platform itself will be active at any time uninterrupted of any problem or work that has to be done. On all platforms common interfaces are used in order to access the functional and data level. All services are feed with data mainly from specific local authorities (initially from the five Municipalities) and other associated bodies. Such bodies are for example Public Transport operators, Hotels associations, the local Chambers of Commerce etc. The relation between the partners, the third parties and the operation (Input/Output data and means of connection) of the Easytrip platform is presented in the following figure (Fig. 4). Figure 4. Operation of the Easytrip platform 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 6

11 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT The Easytrip system has three major types of users: the Administrators, the Content Providers and the Visitors. Concerning the Administrators, the Central Administrator of the system has the full responsibility of the hardware, data management and proper operation of all services and is the Hellenic Institute of Transport. Additionally, all Project partners are Regional Administrators and responsible for the maintenance and the integrity of the data at their regional level. They should constantly, be in touch with the Main Administrator. Content Providers role is to enter and change static content of specific Easytrip services such as points of interest, public transport information, offers and events. The data providers should take permission by the central administrator and/or in some regions by the regional administrator. Visitor s role is not a functional entity of the system, but represents all people visiting the platform and its services. However, visitors of the platform have the ability to make comments about the services, provide corrections to the data and suggest new data (ex. Points of Interest). 5. EASYTRIP E-MOBILITY SOLUTIONS SERVICES The Easytrip platform aims to encourage public transport trips made by tourists by providing them the relevant information. Based on the above, the Easytrip services were designed in order to provide effective, timely and accurate information to support more efficient traveler decisions and system objectives. The Easytrip E-mobility services are provided in Greek, Bulgarian and English languages, so as to enable smooth and convenient flow of information and service provision to all travelers, despite the language used locally. Figure 5. Screenshots from the Easytrip smartphone application (android, ios) and internet platform The final Easytrip services can be categorized into two different categories, Information services and Routing services, and are the following: 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 7

12 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Information Services Public Transport Information Service Points of Interest (Places) Information Service Offers Information Service Traffic Information service Parking Information Service Environmental Information service Touristic Routes Information Service Road Safety Information Service Weather Information Service Routing Services Car Routing Service Public Transport (multimodal) Routing Service More specifically, regarding the Public Transport Information service, it aims to provide information related to: Urban transport routes, stops and terminals Interurban transport routes, stops and terminals Average travel time Timetables Fares Identify the nearest stop / terminal The search for Public Transport stops/terminals may take place based on the current location or by typing the address, designating the location on the map or entering key words, indicating at the same time the maximum walking distance from the current location to the stop/terminal. The search for routes may take place by choosing the desired route and direction from the respective menu or based on the search for stops/terminals, where the user may choose the routes operated from each station. Based on the search for stops/terminals, a list of all Public Transport stops fulfilling the criteria set by the user, appear on the screen (of the PC or the smartphone) and are highlighted on the map in relation to the designated location. By choosing a stop/terminal either from the list or the map, the user may view the name of the station and the routes operated there from. Based on the search for routes, the user may view on the screen (of the PC or the smartphone) a detailed description of the selected route, including all stops in the form of a list or on the map. Moreover, the user may get information on the fare of the selected route and choose a season in order to get information on the respective timetable for the selected route. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 8

13 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT For every stop/terminal selected on the map or from the list, or obtained as a result of the search for stops/terminals or routes, the user may search for Offers, Events, Points of Interest and other Public Transport stops/terminals based on the selected spot. As regards the Public Transport Routing service, such services are very popular among numerous modern international, European and Greek applications. This service, provided through the Easytrip platform, provides routing alternatives based on different parameters, such as the length and duration of the trip and the traveler can get support on the optimal route either car or by public transport not only for local, but also for national and international trips. Additionally, public transport is highlighted as an environmentally friendly and competitive alternative when travelling from one city to another and provides users with routing recommendations using all transportation means available in the area (rail, intercity bus, local and municipal lines). Users define the points of departure and arrival and they also have the option to see Points of Interest and/or Offers available along the route or at the destination, depending on the routing type. Every time the user searches for a route, the results appear on the map along with detailed directions on how to move, including all available means of public transport, the changes, the distance to be covered and the time needed to reach the final destination. 6. EASYTRIP DATA AND INFORMATION REQUIREMENTS The main problem for the implementation of the transport information services is the need for complete, updated and detailed data. The required data do not only referred to the transport systems characteristics (timetables, routes, frequencies etc) but also to geographical information (detailed GIS networks, road characteristics, etc), detailed points of interest description in all the potential origin or destination pairs and all the other data that is required for covering the users needs. For the Easytrip routing services, detailed network mapping took place and all the appropriate data was collected concerning frequencies, timetables, stops and terminals for both urban and interurban public transportation systems (trains, taxis, urban and interurban buses). Additional data included average travel time for each itinerary, route timetables, cost information and nearest stop/terminal location. The points of interest information services, required data relating to entertainment, health and welfare, shopping, tourist attractions, accommodation, restaurants, culture, education, sports centers etc. All the collected information was formed into categories and subcategories in order to facilitate all users search procedure. Additionally, parking information and policy in the study area (e.g. pedestrian streets,, streets with parking restrictions, controlled short-term or long-term parking, loading and unloading spaces etc), as well as pricing policy and force hours were collected and digitized. Except for the above mentioned extensive data collecting process, a major challenge of the Easytrip platform development was also associated with data management and processing, even at the level before the actual use of data for supporting the provided services to the end users. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 9

14 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT In order to support the Points of interest services (POIs), POIs were acquired fot North Greece. A detailed graphical representation of the network, consisting of approximately links for the region of Greece and Bulgaria is used to support the routing services. Additional data were collected by the project partners in order to facilitate individual regions requirements. All data were collected, processed and transformed to a unified efficient data structure, taking into account the requirement analysis, with respect to quality and integrity of the information. Data are stored in a relation database management system (Microsoft SQL Server). Different sources of information and amount of data led to the need of creating the appropriate policy for update of the data. Web-based management systems were designed and implemented in order to support the update of data by third-party entities. 7. ALGORITHMS AND SERVICES DEVELOPMENT Theoretical results on the hardness of the label constrained shortest path problem can be found in many research works (Dellind et al.; In Lerner et al., 2009). Experimental evaluations of basic algorithms are given in (Mendelzon, 1995), while basic speed-up techniques and bidirectional search have systematically been examined (Barret et al., 2000; Barret et al., 2002; Hart et al., 1968). Route planning in uni-modal scenarios has been undergoing a rapid development in recent years with the fastest techniques yielding query times of several microseconds in road networks (Delling et al.). However, there are limited route planning algorithms that can answer a multi-modal query in a huge combined transportation network within milliseconds (Delling et al.). Cooke and Halsey (1966), provided a dynamic programming algorithm for solving the fastest path problem between any node of a network with time-dependent travel times and a single destination for any possible departure time in the discrete time horizon [0,T]. Each iteration k of their algorithm, compares the temporarily optimal path from each node i and time τ in [0,T] that comprises k-1 or less arcs with the emerging path consisting of k arcs. Based on the approach proposed by Cooke and Halsey (1966), Ziliaskopoulos and Mahmassani (1993), developed a label correcting algorithm for the fastest and minimum cost path problem. An alternative category of algorithms for the time dependent shortest path problem has been based on determining efficient techniques in searching for optimum paths in the timeexpanded network. Along this line, Cai, Kloks, and Wong (1997), proposed an algorithm for the time dependent minimum cost path problem from a single origin to a single destination under the constraint that the total time of the path does not exceed a specified standard value T, i.e., the arrival time at the destination should occur within the time window [0,T]. For the development of the Easytrip Public Transport (multimodal) Routing Service and in order to develop a service that is functional, reliable and flexible, a new algorithm for multimodal routing has been developed. This algorithm takes into account time scheduling in order to suggest the optimal route and is based on a metaheuristic process. In metaheuristic algorithms the optimal solution is identified between a set of promising solutions and for this reason a pre-processing procedure has also been designed in order to accelerate the process of finding the solution (speedup technique). 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 10

15 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT More specifically the main steps of the algorithm development are presented below: Step 1: Definition of the origin and destination stops. Either the closest stop to the geographic points given as input to the algorithm or the second closest stop that is in addition served by at least a predetermined number of different lines is selected. Step 2: Identification of the lines (with direction) that pass through the origin and destination stops Step 3: Control logic process o Check whether walking from the origin to the destination stop is feasible. o Check whether it is feasible to connect the origin with the destination stop with one line. This check is performed with a query to the routes database. o Check whether it is feasible to connect the origin with the destination stop by combining two lines. This process takes place in two loops, one inside the other. o Check whether it is feasible to connect the origin with the destination stop by combining three lines. This process takes place in three loops, one inside the other. For each of the possible solutions, only the solution with the least number of interim stops is selected. This means that the algorithm suggests to the traveler the option will lead him/her faster to the line that will take him to his final destination. Step 4: Calculating the parameters of feasible solutions For each leg of the solution the following are being calculated: o Waiting time for each stop, taking into account departure times o Walking time from a stop to a neighboring one o Travel time (taking into account the path length and the number of intermediate stops The total time of travel is the sum of the above. Step 5: Solutions classification procedure The classification of the feasible solutions is based on the calculation of their parameters and more specifically on the lowest weighted mean of the expected total travel time. 8. CONCLUSIONS It is well known that people in our days have very positive attitudes towards the provision of routing or travel information. The need for such information varies according to the purpose of the trip and its type (urban, interurban or even international). The common interest of all travelers is focused mainly on information provision on routing suggestions, using all available means of transport, on points of interest in their destination as well as on parking and road charges information. The presented platform of EASYTRIP provides the above mentioned information through services of an open and expandable architecture that can easily be expended to more places and data. The main characteristic of the system is the variety and the large number of data that it uses, the open access to the public as well as the upgradable services that are being offered based on up-to-date algorithms. The main objective of the EASYTRIP platform is to promote alternative and more 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 11

16 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT sustainable ways of traveling between and inside the participated locations such as the public transport (bus or rail) while at same time to offer various touristic and travel information. The benefits from its operation are many and in most cases not directly quantifiable, as most of the social and economic benefits from such projects. Succinctly, social and economic benefits of the platform are listed below: Social benefits Improvement of traffic conditions and shift to public transport means versus private car Reduction of travel time by providing real time traffic information Reduction of environmental pollution Improvement of road safety Promotion of the cooperation between local public and private authorities Promotion of the culture and history of the participating in EasyTrip locations Equal opportunities for development through e-promotion for all the commercial enterprises Financial Benefits Increase of tourism demand Strengthening of the local economy through tourism and commerce. Better coordination of investments for development of tourism services under the specific requirements that will be recorded from the users of the platform. In order to ensure the sustainability of such a system, close and long term cooperation between local stakeholders, public transport systems operators, the market representatives and all other interested parties must be achieved. This is the only way to continuously update the databases of the platform and offer up- to-date information to the users. This kind of cooperation can be achieved through the development of specific memorandum of understanding or public and private partnerships between relevant public and private authorities. References Barrett, C., Jacob, R., Marathe, M.V (2000). Formal-Language-Constrained Path Problems. SIAM Journal on Computing 30(3), Barrett, C., Bisset, K., Jacob, R., Konjevod, G., Marathe, M.V. (2002). Classical and Contemporary Shortest Path Problems in Road Networks: Implementation and Experimental Analysis of the TRANSIMS Router. In Möhring, R.H., Raman, R., eds.: ESA LNCS, vol. 2461, pp Springer, Heidelber. Cai, X., Kloks, T., Wong, C.K. (1997). Time-Varying Shortest Path Problems with Constraints., Networks, vol. 29, no. 3, pp Cooke, K.L., Halsey, E. (1966). The Shortest Route Through a Network with Time- Dependent Intermodal Transit Times. Journal of Mathematical Analysis and Applications, vol. 14, pp Crouch, G. I., & Ritchie, J. R. B. (1999). Tourism, competitiveness, and societal Prosperity, Journal of Business Research, 44 (3), nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 12

17 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Currie, C., & Falkoner, P. (2013). Maintaining sustainable island destinations in Scotland: The role of the Transport-tourism relationship, Journal of destination marketing & management, Delling, D., Sanders, P., Schultes, D., Wagner, D. (2009). Engineering Route Planning Algorithms. Source Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 5515 LNCS, pp Delling, D., Pajor, T. & Wagner, D.(2009) Accelerating Multi-Modal Route Planning by Access-Nodes. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Duval, D. (2007). Tourism and Transport. Modes, networks and flows. Clevedon:Channel View Publications Hall, D. R. (1999).Conceptualising tourism transport: inequality and externality Issues Journal of Transport Geography, 7(3), Hart, P.E., Nilsson, N., Raphael, B. (1968). A Formal Basis for the Heuristic Determination of Minimum Cost Path, IEEE Transactions on Systems Science and Cybernetics 4, In Lerner, J., Wagner, D., Zweig, K.A. (2009). Algorithmics of Large and Complex Networks, Volume 5515 of Lecture Notes in Computer Science. (pp ). Springer Kaul,R.(1985) Dynamics of tourism. Atrilogy, 111. NewDelhi: Transportation and Marketing Mendelzon, A.O., Wood, P.T. (1995). Finding Regular Simple Paths in Graph Databases. SIAM Journal on Computing 24(6) Page, S. J. (2005).Tourism and transport. Global perspective (2nd ed.). Harlow: Pearson/Prentice Hall Pearce, D. (1987): Tourism today: A geographical analysis, Harlow: Longman Ziliaskopoulos, A., Mahmassani, H. (1993). Time-dependent, Shortest-Path Algorithm for Real-Time Intelligent Vehicle Highway System Applications., Transportation Research Record, vol. 1403, pp nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 13

18 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Prioritization of Smart Mobility Solutions in the City of Volos S. Charalampidou*, E. Nathanail Department of Planning and Regional Development, University of Thessaly, Volos, Greece E mail: sofharal@prd.uth.gr Department of Civil Engineering, University of Thessaly, Volos, Greece E mail: enath@civ.uth.gr Abstract The field of Intelligent Transport Systems (ITS) is a relatively new one in the area of transport demonstrating rapid development. Intelligent Transportation Systems include the applications of information and communication technologies to the planning and operation of transportation systems in order to relieve congestion, improve safety and enhance productivity and mobility. The main aim of the conducted study was to adopt a smart mobility framework in the city of Volos, in Greece, in order to improve traffic conditions and solve the basic problems facing the city. Before the adoption of the proposed applications it is obligatory to examine the efficiency, effectiveness and social acceptability. For this purpose, a questionnaire survey was carried out where the traffic habits of the citizens of Volos were investigated and their attitudes towards applications of smart mobility were examined. This paper presents the methodological tool of the evaluation of the proposed application and the main results of the survey. The study took place in the city of Volos, in July 2013 and lasted about two months. 341 people participated in the research and answered a questionnaire which consisted of three sections. The participants were of different age, gender and profession, in order to have a more independent sample. In the first section the personal characteristics of the respondents were recorded. In the second their traffic habits were examined (description of the two most frequent movements, purpose for its one and their habit of receiving or not information about the traffic conditions and the travel time). In the third part was examined their willingness of receiving information, the areas of information, as well as their attitudes towards certain applications of smart mobility (Variable Message Signs, smart bus stops, municipal rental bikes and carpooling car sharing). For the data analysis both descriptive statistics and inferential statistics were used. Broadly the results showed that participants respond positively to the proposed applications of smart mobility believing that the adoption of them will facilitate significantly their movements. Keywords: Applications of smart mobility, carpooling, Intelligent Transport Systems, rental bikes, Variable Message Signs. 1. INTRODUCTION In general the field of Intelligent Transport Systems (ITS) research is still a relatively young discipline especially in applications, varying from country to country in levels of acceptance and local applicability. Intelligent Transportation Systems include the applications of information and communication technologies to the planning and operation of transportation systems in order to relieve congestion, improve safety and enhance productivity and mobility [1]. While many think improving a country s transportation system solely means building new roads or repairing aging infrastructures, the future of transportation lies not only 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 14

19 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT in concrete and steel, but also increasingly in using Information Technology. Throughout Europe, investments in new large infrastructures are considered the last option facing the transport problems. Nowadays, the primary aim is the sustainability of transport system. ITS, smart mobility and cooperative systems are all related to new technologies aimed at optimizing the use of infrastructure [2,3]. Information Technology enables elements within the transportation system -vehicles, roads, traffic lights, message signs, etc.- to become intelligent by embedding them with microchips and sensors and empowering them to communicate with each other through wireless technologies. In the leading nations in the world, ITS bring significant improvement in transportation system performance, including reduced congestion and increased safety and traveler convenience [4]. 2. DESCRIPTION OF THE STUDY This paper describes the attempt to adopt a smart mobility framework in the city of Volos, in Greece, in order to improve traffic conditions and solve the basic problems facing the city. Before the adoption of the proposed applications it is obligatory to examine the efficiency, effectiveness and social acceptability. Since the proposed applications concern the city of Volos, the traffic behavior of users, their attitude towards the adoption of innovative mobility applications and their tendency of changing their traffic habits have to be examined. For this purpose, a questionnaire survey was conducted, using the technique of random sampling. The choice of this method is based on the fact that questionnaires are a standard way of analyzing the data, they could be sent to a large number of people, the researcher cannot influence the respondents and also is does not take a lot of time to conduct the survey [5]. The study took place in July 2013 and lasted about two months. The research is focused in the city of Volos, where traffic problems are more pronounced. However, smart mobility systems may be applied to the entire municipality to achieve continuity. Volos is a medium-sized city with population of about people. In order to validate the survey the sample has to be about 383 people (confidence level CL=95%, response distribution P=50% and error a=5%) [3]. Finally, 431 people participated in the survey and answered a questionnaire which consisted of three sections. In the first one the personal characteristics of the respondents were listed. In the second, their traffic habits were examined and in the last section were recorded their attitudes towards certain applications of smart mobility (Variable Message Signs, smart bus stops, municipal rental bikes and carpooling car sharing). The selected choices are internationally representative examples of successful applications of ITS [6,7,8]. Additionally, they are active measures of guidance as the user chooses whether to use them and also satisfy the specific urban and transport needs of the city. 3. RESULTS OF THE SURVEY 3.1. Descriptive statistics Regarding the characteristics of the sample, 53% of the participants were women and 47% men. Among them, 46% were under the age of 25, 53% were between 26 and 59 years old and 1% was over the age of 60. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 15

20 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT In the second part, people were requested to describe the two most frequent movements, the reason for each one and if they are accustomed to receive information about the traffic conditions and the travel time. In order to facilitate the decoding of responses the city of Volos was divided into three zones: the central B, the western A and the eastern C. Figure 1. The study area Figure 2. Origin destination of the first movement According to given answers, most of the movements (both 1 st and 2 nd ) took place in the central districts of city. These results were absolutely expected as the central zone B is the urban center of the city constituting a cluster of recreational activities and services. However, the aim differs between the first and second movement. Regarding the first one, the main purpose (69%) is work education and for the second are other and mostly (57%) the entertainment. Moreover, the majority of respondents (96%) stated that they are not used to receive any information about their movement. Nevertheless, 80% of the participants declared that they want to receive some kind of information before and during the upcoming movement, and only 20% is not interested in receiving any type of information. This probably means that there is no possibility of information about the traffic issues. The information wanted concerns the traffic condition, disruption of transit services, delays, travel time from origin to destination 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 16

21 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT for each mode of transport, availability of parking spaces, combined transports and weather conditions. For the evaluation of the proposed applications, a Likert type scale from 1 to 5 was used, where 1 equals to "not at all" and 5 to "very much". 84% of the participants responded positively in the installation of Variable Message Signs in central city points which will inform users about the traffic conditions. The smart bus stop system which is already applied in certain stops in the city is characterized very effective given that 92% of the participants stated that facilitates their movement (from moderate to very much). Figure 3. Evaluation of Variable Message Signs Figure 4. Evaluation of smart bus stops Most of the respondents, about 85%, are also positive about the use of municipal rental bikes, believing that this system will facilitate significantly their movements. However, they seem to be skeptical about the application of carpooling-car-sharing as there is a dispersion of answers varying from not at all to very much. This reservation is probably due to the short distances of the city of Volos. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 17

22 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Figure 5. Evaluation of municipal rental bikes 3.2. Inferential statistics Figure 6. Evaluation of the carpooling-car sharing system Regarding inferential statistics, the statistical analysis of the responses was carried out using nonparametric tests, which are considered as particularly powerful for analyzing data collected through questionnaire surveys. The parameters which are examined are listed below: Gender (2 classes: male, female) Age (3 classes: <25 years old, 26-59, >60) Origin - destination of the two most frequent movements (3 classes: movements with origin destination the central zone B, movements between the zone B and A or C, movements between the zone A and C) Purpose of its movement (2 classes: work-education, other) Receiving of information (2 classes: yes, no) Except from those variables, the given choices of smart mobility were also examined. As it was mentioned before there were 4 choices: Variable Message Signs Smart bus stops Municipal rental bikes Carpooling-car sharing 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 18

23 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT The correlations between the variables were examined with the use of statistical package SPSS. Specifically was tested the correlation between parameters and choices and the correlation between options. The data consist of qualitative categorical variables. One sample Kolmogorov-Smirnov (Monte Carlo simulation) test was conducted for the normality control. The p values for all the variables were zero which means that the assumption that the variables of this study are normally distributed is rejected at a significance level a = 0.05 or a = 5%. Therefore non parametric tests were executed. Chi-square (X2) test of independence was used to examine the hypothesis that two categorical variables are independent. Kruskal-Wallis testing and Mann- Whitney two-sample U-testing were performed to assess differences among and between the samples in characteristics measured on the Likert scale, respectively. The data processing indicated that there is a strong correlation (p value = 0,000) between gender and proposed applications. Specifically, in all examined variables, women responded more positively than men in the acceptance of innovative technologies. In table 1, the mean values of the responses between the two genders are depicted. Table 1. Means values of the responses between the two genders Variable Male (mean value) Female (mean value) p-value Evaluation of VMS 110,28 309,25 0,000 Evaluation of smart bus stops 113,40 306,50 0,000 Evaluation of municipal rental bikes Evaluation of carpooling car sharing 109,19 310,22 0, ,71 314,17 0,000 The exact spread of responses as regards acceptance (on a Likert scale) of the proposed systems, also indicates differences, as it is illustrated in figure 7, which focuses on the placement of VMS at key points in the city, and is used as a representative graphical analysis of the responses. In this figure it is shown that most of the women (60%) believe that this system will facilitate their movements very much, followed by 40% of women who believe that it will facilitate them much. For the same system, men believe that it will facilitate them much 19%, moderately 46%, followed by a 22% who believe that it will facilitate them little and 13% who assess it as useless. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 19

24 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Figure 7. Correlation between gender and Variable Message Sign Similarly, 100% of women assessed smart bus stops as very useful facilitating their movements, while 10% of men believe that smart bus stops will facilitate their movements very much, 55% much, 18% moderate, 11% little and 6% consider that they would not have any impact on their movements. Regarding the municipal rental bikes 80% of women consider that they will help very much their movements, followed by the rest 20% who believe that municipal rental bikes will help much their movements. Contrary, 0% of men evaluate the municipal rental bikes as very useful in order to make their travels more convenient, 34% believe that they will facilitate much their movements, followed by 32% who consider that they will have moderate impact on their travels, 20% little impact and 14% assessed them as no useful at all. The rates of the evaluation of carpooling-car sharing system seem to be diversified but still women are more positive in their implementation. Specifically, 46% of women believe that the system of carpooling-car sharing will facilitate their movements very much, 43% much and 11% moderately. For the same system, 27% of men believe that it will facilitate moderately their travels, 35% little and 38% believe that it will not have any impact on their movements. From the results derived of the conducted tests (Kruskal-Wallis test) it is obviously observed that there is a statistical significant correlation (p value = 0,000) between Origin Destination and proposed applications. In table 2, the mean values of the responses between the combinations of the three zones of the first movement are depicted. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 20

25 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Table 2. Mean values of the responses between the combinations of the three zones of the first movement Variable Movements with OD the zone B (mean value) Movements between the zone B and A or C (mean value) Movements between the zone A and C p-value Evaluation of VMS 49,88 307,19 150,15 0,000 Evaluation of smart bus stops Evaluation of municipal rental bikes Evaluation of carpooling car sharing 50,91 306,50 150,82 0,000 47,52 307,16 152,01 0,000 45,96 310,94 145,22 0,000 Since the most movements are performed between the central zone B and the zone C or A (54%, figure 2), the proposed applications have higher rates of positive evaluation in these areas. The analysis indicates that the proposed applications of smart mobility will have great impact in the movements between the central zone and the zone A or C, significant impact in the movements between the zone A and C and less impact in the movements within the central zone. A typical example is that of the figure 8 which depicts the correlation between origin - destination of the first movement and municipal rental bikes. The system of municipal rental bikes will facilitate the movements between the central zone B and the zone A or C very much with a proportion of 78% and much 22%. The same system will facilitate the movements between the zone A and C much 57% and moderately 43%. Regarding the movements within the zone B, the municipal rental bikes will facilitate the movements of the users moderately 19%, little 47% and not at all 34%. Similarly, the adoption of smart bus stops will facilitate absolutely 100% the movements between the zone B and A or C and very much 14% followed by much 86% the movements between the zone A and C. On the contrary, the impact of smart bus stops is not significantly in the movements within the zone B as 17% of them will be much facilitated, 42% will be moderately facilitated, followed by 26% and 15% which will be little and not at all facilitated respectively. The adoption of VMS system seems not be differentiated. 59% of the movements between the zone B and A or C will be facilitated very much followed by 41% of them which will be facilitated much. VMS system also positively affects the movements between the zone A and C as 30% of them will be much facilitated and the rest 70% will be facilitated in a moderate level. The least influence seems to have in the movements within the central zone as 18% of them will be moderately facilitated, 53% will be little facilitated and there are some movements 29% which will not be affected at all. Figure 9 depicts clearly that due to the short distances of the city the system of carpooling-car sharing will have the least impact in the movements within the central zone B. As it is clearly observed carpooling application will facilitate very much 45% the movements between the zone B and A or C. The same travels will be much facilitated at about 42% and moderately facilitated 13%. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 21

26 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Figure 8. Correlation between Origin - Destination of the 1st movement and municipal rental bikes Figure 9. Correlation between Origin Destination of the 1 st movement and the carpooling-car sharing application Because of the short distances of the city, carpooling innovative technology will not influence at all 91% the movements within the zone B and will contribute little in facilitating the same movements 9%. The data processing indicates that there is the same correlation between origin destination of the second movement and the four choices of proposed innovative systems. From the tests executed, a statistically significant correlation (p value = 0,000) was observed between the purpose and the proposed applications. All of smart mobility applications had greater impact when the main purpose is other than work-education. While, when the aim is work-education, the respondents seem to be more restrained in 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 22

27 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT assessing applications. Table 3 presents the mean values of the responses between the purposes concerning the evaluation of the four choices. Table 3. Means values of the responses between the purposes of the 1 st movement Variable Work - education (mean value) Another (mean value) p-value Evaluation of VMS 151,32 362,50 0,000 Evaluation of smart bus stops 176,05 306,50 0,000 Evaluation of municipal rental bikes Evaluation of carpooling car sharing 161,48 339,50 0, ,17 358,31 0,000 As it was mentioned above in descriptive statistics analysis, the main purpose of the first movement is work education while for the second is other and specifically the entertainment. As a result the proposed innovative applications seem to facilitate more the second movements whose purpose is other than work education. Figure 10. Correlation between the purpose of the 1st movement and the VMS system 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 23

28 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT Figure 11. Correlation between the purpose of the 2 nd movement and the VMS system Figure 10 and 11 constitute two representative examples of the correlation between the acceptability of VMS application and the purpose of each movement. In the left figure it is clearly stated that all of the first movements (100%) with purpose other than work will be facilitated very much by the installation of VMS systems at key points in the city. Conversely, when the main purpose of the first movement is work-education the proposed system will facilitate then much 44%, moderately 32%, little 15%, while 9% of the movements will not be facilitated at all. For the second movement, when the main purpose is work-education the impact of VMS is not very significant as 26% of the travels will be moderately facilitated, 48% will be little facilitated and 26% will not be influenced at all. However, when the purpose of the second movements differs from work-education, the installation of VMS system will facilitate very much the movements 41%, much 39% and moderately 20%. This observed behavior is due to the fact that the respondents are aware of their daily routes to wok-education while receiving information will be more helpful for no common and familiar movements. The correlations between the purpose of its movement and the other proposed choices (smart bus stops, municipal rental bikes and carpooling-car sharing) are similar of the above example. When testing correlation between age and proposed applications, and receiving information and proposed applications, no statistically significance differences were indicated (p value > 0,01). Regarding the correlation between the four choices (VMS, smart bus stops, municipal rental bikes, and carpooling-car sharing) there is no statistically significant dependence of each other. There is only a slight correlation between the electronic signs, the municipal rental bikes and the carpooling system. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 24

29 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT 4. CONCLUSIONS This paper concentrates on the presentation of the methodological tool aiming to evaluate the effectiveness and efficiency of some specific innovative applications of the transportation system corresponding to the basic traffic problems of the city of Volos. The main results concluding of the conducted survey are that participants evaluate positively all of the proposed applications. The most movements (54%) take place between the central zone B and the zone A or C (west and east part of the city respectively). The occurrence of this phenomenon may be explain by the fact that zone B is the urban center of the city where the most of activities are conducted. The purpose of the first movement is usually (69%) work education and for the second are other and mostly (57%) the entertainment. Moreover, there is a statistically significant correlation (p value = 0,000) between the proposed applications and gender, origin destination and purpose of movements. Based on the results derived from the survey in addition to best practices which has been implemented internationally and the specific transport problems which are recorded in other studies, an integrated system of applications of smart mobility was proposed. The proposed applications follow the categorization of activity areas of the European institution ERTICO, including actions which concerns the safe mobility, eco mobility, cooperative mobility and info mobility [9]. References Kellberger, S., Enhancing the transfer of ITS innovations to the market, ITS state of the art assessment. T-Trans project of European Commission. Available at: pdf [Accessed July 1, 2013]. McQueen, B. and J., Intelligent Transportation Systems Architectures. London: Artech House. Stathopoulos, A. and Karlaftis, M., Transportation systems planning. Athens: Papasotiriou (in Greek). Stephen, E., Explaining international IT application leadership: Intelligent Transportation Systems. Information Technology & Innovation Foundation. Available at: [Accessed June 2, 2013]. Kotzamanis, B., Introduction to research methods. Didactic notes, Volos: University of Thessaly (in Greek). Haque, M. and Debnath, A., Sustainable, safe, smart three key elements of Singapore s evolving transport policies. Elsevier, Transport Policy, 27, pp Available at: [Accessed May 19, 2013]. Garcìa Ortiz, A., Amin, S. and Wootton, J., Intelligent Transportation Systems enabling technologies. Elsevier, Mathematical and Computer Modelling, 22(4-7), pp Available at: 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 25

30 SESSION: ITS TECHNOLOGIES AND SOLUTIONS IN PUBLIC TRANSPORT [Accessed May 19, 2013]. Transit Cooperative Research Program, Car Sharing: where and how it succeeds. Transportation Research Board, TCRP, 108. Available at: [Accessed May 19, 2013]. ERTICO, Intelligent Transport Systems and Services for Europe, Activities. Available at: [Accessed May 19, 2013]. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 26

31 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT

32 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT A Multi-Periodic Optimization Modeling Approach for the Establishment of a Bike-Sharing Network: a Case-Study of the City of Athens Saharidis G.K.D a, Fragkogios A b, Zygouri E b, Mavrotas G c a Department of Mechanical Engineering, School of Engineering, University of Thessaly, Volos, Greece and Kathikas Institute of Research & Technology, Paphos, Cyprus saharidis@gmail.com b Department of Mechanical Engineering, School of Engineering, University of Thessaly, Volos, Greece fragkogiosantonios@gmail.com, el.zigouri.haf@gmail.com c Department of Process Analysis and Plant Design, School of Chemical Engineering, National Technical University of Athens, Athens, Greece mavrotas@chemeng.ntua.gr Abstract This study introduces a novel mathematical formulation that addresses the strategic design of a bicycle sharing network. The developed pure integer linear program takes into consideration the available budget of a city for such a network and optimizes the location of bike stations, the number of their parking slots and the distribution of the bicycle fleet over them in order to meet as much demand as possible and to offer the best services to the users. Given a set of candidate locations of bike stations and the assumed time-dependent demand for bikes at these locations during a single day it is necessary to know where to place the bike stations and how many parking slots and bikes should each one have. The 24 hours of the day are discretized into time intervals, during which different numbers of users come to each station either to pick up or to drop off a bicycle. The mathematical formulation is solved using CPLEX. The proposed approach is implemented on the very center of the city of Athens, Greece. Keywords: bike sharing, integer, mathematical model, multi-periodic. 1. INTRODUCTION Bike-sharing networks have received increasing attention during the last decades and especially in the 21st century as a no-emission option in order to improve the first/last mile connection to other modes of transportation, thus facilitating the mobility in a densely populated city. The bike-sharing network consists of docking stations, bicycles and information technology (IT) interfaces that have been recently introduced to improve the quality offered to the users. There have been three generations of bike-sharing programs over the past half century (1). The 1st generation emerged in 1965 when the White Bikes were spread throughout Amsterdam for public use. The network was not successful as within few days the bikes were stolen or damaged. The next generation was introduced by two small networks in Farso and Grena and in 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 27

33 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Nakskov, Denmark in 1991 and in 1993 respectively. In 1995 the Bycyklen program was launched in Copenhagen, where the bikes were locked in specific stations and unlocked by the users with a coin. However, the bicycles still experienced theft due to the anonymity of the user. Better user tracking was necessary. The first implementation of the 3rd generation bike-sharing networks was Bikeabout in 1996 at Portsmouth University, where students could rent a bike using a magnetic card. Despite its success, it was not until 2005 that this generation flourished with the launch of Velo v with 1500 bikes in Lyon. Two years later, Paris launched Velib and Barcelona launched Bicing, which are two of the most successful networks nowadays. Nowadays, there are more than 550 programs in operation and more than 150 in planning or under construction all over the world (Metrobike, July 2013). This expanding trend of bike-sharing networks necessitates their better planning and design in order that they are successful. The goal of this paper is to propose a novel mathematical formulation to design such networks incorporating the hourly demand estimation, the fixed costs of infrastructure, the proximity and density of stations, as well as their size. Given a set of candidate locations of stations and with a predefined available construction budget the model decides the number and the location of the stations, how large they will be and how many bikes should they have at the beginning of the day in order to meet the assumed demand. The remainder of the paper is organized as follows. Section 2 provides a brief literature review of the main approaches that have been proposed to solve similar problems. Section 3 presents the developed novel mathematical model. In section 4 the case-study for the center of Athens is described followed by the results of the implementation of the model on it. Finally, in section 5 there is a commentary on the proposed model, its broader application and potential areas of future work. 2. LITERATURE REVIEW The allocation of bike stations is essentially a hub location problem which varies depending on the system requirements. Farahani et al. (2013) (2) deal with hub location problems making a review of the mathematical models, the solution methods, the main specifications and the applications of such problems since In the present research the establishment of a bike-sharing network is dealt with as a hub covering problem meaning that an established station-hub covers the demand of the locations that are within a maximum distance from it. However, the network is considered in terms of individual demand points instead of origin-destination pairs, as in (2), because the latter would make the problem too complicated to solve for large bike-sharing networks. Finally, the herein paper introduces the scope of covering only a part of the demand of the locations other than station locations. The rest of the demand is lost with an additional cost. This scope is assumed to better simulate the behavior of the users of a bike-sharing network. Shu et al. (2010) (3) proposed practical models for the design and management of a bicycle-sharing network given the location of the stations. A stochastic network flow model is introduced in order to predict the flow of bicycles within the network and to estimate the number of trips supported by the system, the suitable number of bicycles 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 28

34 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT to be deployed and the number of docks needed in each station examining the viability of periodic re-distribution of bicycles as well. In the herein research the number and the location of the stations is not predefined, as in (3), but part of the design problem. Furthermore, an average value of the demand of each candidate location is assumed according to the recorded usage data of already implemented bike-sharing networks. Finally, the present work addresses the design of such networks and not their management, so no re-distribution aspects are taken into consideration, as in (3). Lin et al. (2011) (4) developed a pure integer non-linear program for the strategic design of a bike-sharing network. Given a set of origins, destinations, candidate bicycle stations and the travel demands from origins to destinations with specific demand processes, it optimizes the location of the stations and bicycle lanes and the required inventory level for sharing bicycles at each station to meet demand. The herein model is a pure integer linear program where no origin-destination flows are assumed, but every location is characterized by time-discretized demand for pick-ups and drop-offs during a single day. This approach is considered to give improved and less complicated simulation of the network s future usage. Additionally, the bike-sharing network is dealt with independently and so the establishment of bicycle lanes is not in the scope of the present research. Sayarshad et al. (2011) (5) introduce a multi-periodic optimization formulation to determine the minimum required bike fleet size that minimizes simultaneously unmet demand, unutilized bikes and the need to transport empty bikes between rental stations. The herein model, also, uses multi-periodic formulation without re-distribution concerns because it addresses only the network design problem and not its usage, as in (5). Martinez et al. (2012) (6) present a heuristic, encompassing a mixed integer linear program, which optimizes the location of bike stations and the fleet dimension, while measuring the required bicycle re-distribution activities. It considers a mixed fleet of regular and electric bikes and several fare collection methods of the system. The present research considers only regular bikes assuming that electric ones are yet to come in such a network. Moreover, it includes no fare policy as this may be decided after the establishment of the bike-sharing network. García-Palomares et al. (2012) (7) use Geographical Information System (GIS) to calculate the spatial distribution of the potential demand for trips, locate stations using location-allocation models, determine station capacity and define the characteristics of the demand for stations. In the herein project the GIS is not used as access to a respective software could not be granted. The demand data are derived by the recorded usage data of already implemented similar bike-sharing networks. Romero et al. (2012) (8) have developed a bi-level mathematical programming model that optimizes the location of public bicycle docking stations with a genetic algorithm. Its lower level is a modal split and assignment model, which reflects the interactions between car and bicycle mode. In the present study no such methodology was used aiming at less complexity and the bike sharing network is dealt with independently from other means of transport, such as the car. Finally, the number of parking slots and bikes required in each station are, also, considered, contrary to (8). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 29

35 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT 3. MODEL FORMULATION 3.1. Problem definition Given a set of candidate locations of bike stations and the time-dependent demand for bikes at these locations during an average day it is necessary to know where to place the bike stations and how many parking slots and bikes should each one have. The available budget of a city for the construction of the whole bike-sharing system is predefined and so are the costs of a single bike, a single parking slot and a single station. So it is a matter of optimization for the model to decide how many stations, bikes and parking slots it will include in its solution. The walking time between the locations is another parameter of the problem used to ensure the proximity of the constructed stations as far as this is possible. As regards demand in each location, it is split into Demand for Pick-Ups, i.e. how many users would like to take a bike from a station, and Demand for Drop-Offs, i.e. how many riders would like to leave a bike into a station. The 24 hours of the day are discretized into time intervals of one hour, during which different numbers of users come to a station either to pick up or drop off a bicycle Mathematical model The problem was formulated as a pure integer linear problem and was solved using CPLEX optimizer through a C++ code. The code was implemented on a laptop computer (Intel 2.67 GHz Core i5 and 4GB of RAM). The model includes the following subscripts and sets, input parameters and decision variables: Subscripts and Sets: i, k N: the candidate locations of bicycle stations t, p T: the time intervals in a single day Input Parameters: : : : : : : : : : the cost of purchasing a single bicycle the cost of establishing a bike station (without any parking slots) the cost of constructing a single parking slot into an established station the walking time from location i to location k (in minutes) the maximum walking time (in minutes) between two candidate locations, of which the one has an established station and the other one does not have a station. the total available budget for the establishment of the whole bike-sharing network the Demand for Pick-Ups from location i during time interval t the Demand for Drop-Offs at location i during time interval t a parameter that equals 1 if the Demand for Drop-Offs is more 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 30

36 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT than the Demand for Pick-Ups until time interval p and 0 otherwise. : the minimum number of parking slots a station could have : the maximum number of parking slots a station could have : the percentage of the demand that is transferred from a location where a station is not established to an established station. It is assumed that if a station is not established at a location, part of its demand is lost (1 ). : the penalty unit cost of the transfer of a unit of demand of a candidate location where a station is not established to an established station per minute of walking time : the penalty cost for a unit of unmet demand : a very large number : a very small number Decision Variables Xk : Zik : DNk : BNkt : BFkt : BEkt : UDBinFkt : UDBinEkt: binary variable, that equals 1 if a station is established at location k and otherwise binary variable, that equals 1 if candidate location i is served by the established station at location k and 0 otherwise general integer variable that equals the number of constructed bicycle parking slots at station k general integer variable, that equals the number of bicycles that are available at station k at the beginning of time interval t general integer variable, that equals the number of bicycles that could leave station k during time interval t, where BNkt bicycles are available general integer variable, that equals the number of bicycles that could arrive at station k during time interval t, where DNk parking slots are established and BNkt bicycles are available binary variable, that equals 1 if a station k cannot serve some Demand for Pick-Ups at time interval t and 0 otherwise (not enough available bicycles). binary variable, that equals 1 if a station k cannot serve some Demand for Drop-Offs at time interval t and 0 otherwise (not enough available parking slots). In Figure 1 the thorough consideration of the problem is explained. N locations i are predefined together with their Demand for Pick-Ups (i.e. ) and Demand for Drop-Offs (i.e. ) at all-time intervals during an average day. The walking time between these N locations is, also, known. It is a matter of optimization how many bike stations will be established and where, so that every location has a nearby station. The locations k, where stations are established, is a subset of the locations i. If budget is not enough to construct stations at all N locations, at some locations i there will be no station. These locations should have a nearby station k no more than a specific walking time away (i.e. ) and only a percentage (i.e. ) of their demand (i.e. and ) is considered to be passed to this station k. The rest of their demand is not served supposing that this part of citizens will not take a bike due to the distance of the station k from their location i. In this way, it is assumed that location i is served by station k, i.e. Zik =1. On the one hand, this transfer of the demand 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 31

37 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT is inevitable as the restricted budget does not allow station to be built at all locations. On the other hand, it is not desirable because it means that the users of the network will have to walk from location i, where they would rather a station to be present, to the established station k and vice versa. This would result into poor service quality offered to the users of the bike-sharing network, as some potential customers will not eventually use the network (i.e. 1 ). Objective Function Figure 1: Network structure of bike-sharing system The objective function of the model is a minimization of three terms: The first term expresses the amount of demand (i.e. *( ) ) that is transferred from a location i to its allocated station k (i.e. Zik ), which are a specific walking time away from one another (i.e. ). Thus, the model will propose a dense distribution of stations, establishing no station at locations with low demand ensuring that they are as close to a station as possible. This term is multiplied by the penalty unit cost to differ its importance from the other two terms. The second and the third term of the objective function are introduced in order to minimize the unmet demand. There is a difference between the parameters and and the variables BFkt and BEkt. The former express the users who would like to pick up and drop off a bike from and to a candidate station location respectively. However, the station k may not have the required bikes or free parking slots to meet these two types of demand respectively. So the number of bikes that eventually leave 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 32

38 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT or arrive at a station k is expressed by the two mentioned variables. Both the parameters and the variables refer to each time interval t. These two terms are multiplied by the same penalty unit cost meaning that no different importance is given to either of them. Constraints 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 33

39 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Constraint (1) warrants that the total cost for the establishment of all stations, the construction of all parking slots in them and the purchase of all bikes does not exceed the available budget. Constraint (2) ensures that the bicycle parking slots (i.e. DNk ) at each constructed station are between the permissible minimum and maximum value (i.e. and ). Constraint (3) ensures that at all time intervals, each station cannot have more bikes than the number of its parking slots. Constraint (4) means that at all time intervals the total number of bicycles at all stations will not exceed the total number of bicycles at the first time interval. This constraint is introduced because the first time interval is assumed to be 4-5am, so at 4am all bikes are considered to be parked into the stations and no user keeps a bike away. During the day a user can keep a bike for more than the duration of the time interval (e.g. one hour) and return it to a station at a later time interval. This happens according to the hourly distribution of the Demand for Pick-Ups and the Demand for Drop-Offs at all stations of the network. So in a given time interval t due to more Demand for Pick-Ups than Demand for Drop-Offs the total number of available bikes at all stations will be less than the initial number. Afterwards, in a later time interval t >t due to more Demand for Drop-Offs than Demand for Pick-Ups the available bikes at all stations will be more than those in time interval t, but not greater than the total number of bikes in t0. This constraint also ensures that the model does 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 34

40 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT not add bikes to the network during the day, i.e. the bike-sharing network is a closed network. Constraint (5) expresses that the number of bicycles at station k at the beginning of time interval t+1 is equal to the ones it had at the beginning of time interval t plus the bikes that arrive minus the ones that leave during time interval t. Constraint (6) guarantees that a location i cannot be served by location k, if a station is not built in location k. Constraint (7) warrants that if a station is constructed at location k this location will be served by its own station. Constraint (8) ensures that each location i may be served by exactly one bike station k. Constraint (9) expresses that a constructed station k can serve only locations which are located within a maximum walking time from it. Constraint (10) guarantees that at every time interval the bicycles that can leave the station can be no more than the available ones. Constraint (11) ensures that at every time interval the bikes that can come to a station can be no more than the free parking slots. Constraint (12) expresses that at every time interval the bikes that can leave a station can be no more than the demand for pick-ups of this station plus a percentage of the demand of all other locations this station serves. Constraint (13) expresses the same as the previous one, but for the demand for drop-offs. Constraint (14) forces the variable UDBinFkt to be 1 if a station k cannot serve some Demand for Pick-Ups during time interval t and 0 otherwise. Constraint (15) forces the variable UDBinEkt to be 1 if a station k cannot serve some Demand for Drop-Offs during time interval t and 0 otherwise. Constraints (16) and (17) guarantee that if there is unsatisfied Demand for Pick-Ups, all available bikes will leave the station and if there is no unsatisfied Demand for Pick- Ups, the whole demand will be met. Constraints (18) and (19) guarantee that if there is unsatisfied Demand for Drop-Offs, all bikes will fill the available slots and if there is no unsatisfied Demand for Drop- Offs, the whole demand will be met. These two constraints are relaxed if the Demand for Drop-Offs is more than the Demand for Pick-Ups until time interval p (i.e. = 1 ), which is a deformation of the assumed demand. This happens when due to the values of the input parameters and, it is assumed that until time interval t the total number of users that want to drop off a bike at all stations are more than the ones that have already picked up a bike. The bike-sharing network is a closed network, so at all time intervals the users who want to arrive at all locations may not be more than the ones that have already left them. With the introduction of this parameter at these two constraints the model is not obliged to meet the whole Demand for Drop-Offs at the time intervals at which this happens. Finally the constraints (20), (21), (22), (23), and (24), (25), (26), (27) are the integrality and the non-negativity constraints, respectively. At this point it is necessary to explain how the model decides the number of a station s parking slots (i.e. DNk ) and its bikes at first time interval (i.e. BNkt0 ). Giving a value at these two variables it determines the values of UDBinFkt and UDBinEkt (constraints 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 35

41 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT (14) and (15)). The latter variables determine the values of the bikes that will leave or come to the station k at the first time interval (i.e. BFkt0 and BEkt0, constraints (16) to (19)). The last ones determine the kt0 kt0 available bikes of the station k at the beginning of the next time interval t1 (i.e. BNkt1, constraint (5)) and so goes on. Heading to minimize unmet demand the model proposes those values of DNk and BNkt0 at each station that will result into having the suitable number of available bikes and free parking slots in the following time intervals given the station s different distribution of demand during the day. 4. ATHENS CASE-STUDY 4.1. Data settings In order to find the optimal design for a bike-sharing network for the city of Athens the estimation of its potential demand is necessary. Generally, it should be noticed that the goal of this research is to develop a generally applicable modeling approach for the design of the bike-sharing network and not the estimation of demand. However, so as to estimate the potential demand of a bike-sharing network for the city of Athens, the three existing in the literature papers were analyzed and one of them was taken into consideration. Jon Froehlich et al. (2009) (9) provide spatiotemporal analysis of the bicycle station usage in Barcelona s shared bicycling network, called Bicing. Neal Lathia et al. (2012) (10) analyze the usage data of the London Barclay Cycle Hire network. Finally, Come Etienne et al. (2012) (11) propose a model to form clusters of the stations of the Velib network of Paris based on their usage data. All three papers contain usage data of constructed bike-sharing networks in three European cities that could be used to predict spatiotemporal demand of such a network that is to be constructed in Athens. The paper for the Velib network of Paris describes stations dynamics in a significantly ampler way than the one used in the two other studies, providing information on arrivals and departures. This fact makes it ideal data for constructing the values of the parameters and. In the present paper there was made an analysis of the usage data of the Velib network of Paris, in order to define potential demand in Athens. It is, also, necessary to mention that the dataset used here corresponds to the average demand in one month in Paris 2 (April 2011). The respective paper (11) results in categorizing the more than 1200 bike stations of Paris into eight different clusters: Spare-Time (1), Spare-Time (2), Parks, Railway Stations, Housing, Employment (1), Employment (2), Mixed. These clusters were derived by the analysis of the usage data of each station during the day. So, for example, if a station has a specific activity during the day with high Demand for Pick-Ups in the morning, when citizens pick up a bike from home to work, and high Demand for Drop-offs in the evening, when citizens return home, this station is set to the Housing cluster. Finally, for each cluster specific average values of each demand are derived. However, in the present project the network is not established yet. So no data is available and each location s demand has to be predicted. Depending on each location s surroundings (high population density or job density or next to metro stations), it is set to a cluster, which has a specific average demand pattern during the day. For the small 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 36

42 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT region of the 1 st Municipality. District of Athens and aiming to make the demand data less complicated, while not losing in validity, the authors chose four of the previously mentioned clusters in order to categorize the candidate locations of the mathematical model. It is considered that the prediction of demand cannot have so many details as usage data recording. For example, no two different clusters are needed to describe the Employment locations. The four clusters used are: Housing, Employment, Subway, Spare Time. The Housing cluster contains candidate locations that are in densely populated areas. A dissymmetry in its profile is noticed with a lot of Demand for Pick-ups but little Demand for Drop-Offs at the morning peak and the reverse at the afternoon peak. The cluster Employment presents a dissymmetry, contrary to the one in Housing. This cluster contains the locations close to a business area. The third cluster, Subway, corresponds to the candidate locations, where a metro station exists. This cluster has the maximum total demand during the whole day, which is indicative of the high activity of metro stations. Finally, the cluster Spare Time includes the locations, which are close to restaurants, coffee bars, shops etc. The demand profiles were derived by the average usage data of the Velib network in whole Paris. The present paper solves the case of establishing a bike-sharing network in the 1 st Municipal District of Athens. The comparison between the population density of Paris (12) and the 1st Municipal District of Athens (13) indicates a ratio of 1, i.e. the two regions have the same population density and thus the same expected usage of the bike-sharing network. However, apart from the population density, the stations distance from the center should, also, be taken into consideration in order to predict the demand in Athens. In the previously mentioned paper for the Velib network, it is noted that the mean activity of a station is more significant if it is located near the center of Paris. The Velib network s stations are from 0.1km to 10km away from the city center ( Les Halles ). The selected candidate locations in this problem are from 0.35km to 3.1km away from the city center ( Syntagma ), which means that they are very close it. So it is assumed that the mean demand of a potential bike-sharing network in the center of Athens is approximately equal to 1.5 times the mean demand in the entire city of Paris. For this reason, the usage data of the Velib network was multiplied by a factor of 1.5. Figure 2 depicts the demand patterns of the selected four clusters of stations during the weekdays in Athens as they were derived by the respective usage data of the Velib network. The candidate stations in the problem of Athens are categorized into these four clusters depending on their location. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 37

43 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 2. a) The hourly Demand for Pick-Ups of each cluster during a weekday in Athens, b) The hourly Demand for Drop-Offs of each cluster during a weekday in Athens However, even in the same cluster the stations will not have the same average values of demand. Some stations may have higher values, while others may have lower ones. So, the total demand is not evenly distributed to each candidate location, but it depends on its distance from the city center and its nearby city characteristics. Thus, each candidate location is given a scaling factor which is multiplied by the average values of demand in the cluster it belongs to and changes them into higher or lower values, while the profile of the cluster remains the same. In the specific problem of the 1 st Municipality District of Athens, the scaling factors of the candidate locations differ from 0.25 (depicts a low-activity location) to 2 (depicts a high-activity location). As mentioned above, the mathematical model needs a set of candidate locations to be already defined. So, the authors of this paper chose 50 candidate locations where bikesharing stations could be constructed, taking into consideration important urban data such as the population density, the job density, the metro stations and the squares. These 50 locations were categorized into the previously described 4 clusters and each one was given a scaling factor of 0.25 to 2 according to the criteria that were mentioned above. The average value of all these factors has to be equal to 1, so that the total amount of demand remains the same. These locations are shown at Figure 3. The shape 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 38

44 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT of each dot is related to the cluster to which it belongs and its size depicts its assumed scaling factor. Figure 3. The chosen candidate station locations categorized in clusters and with their scaling factor. The bounds of the 1 st Municipality District of Athens are also depicted. (Reference: The walking time between these locations was calculated using Google Earth and the data was inserted into the table. As regards the costs of the network, two already implemented networks were taken into account, the first one in Greece (Karditsa) and the second one in Cyprus (Nicosia). Examining the budget and the dimensions of each city and its network the following data were assumed for the case of Athens. The cost of establishing a station ( ) is 12,000. The cost of each slot in a station ( ) is 900. The cost of a bike ( ) is 500 and the total available budget (BDG) is 1,000,000. The value of is assumed to be 7 minutes. This means that a location with no station cannot stand off a location with a station within more than 7 minutes of walking time. The minimum and the maximum parking slots that a station can have are as many as in the Velib network (between 8 and 70 per station) (11). These are the values of and respectively. Finally, it is assumed that for every customer that has to walk from his/her location to the allocated established station the penalty cost ( ) of the network is 1 for every minute he/she walks. The second and the third term in the objective function will have the same cost, because there is no extra significance in minimizing either Demand for Pick-ups or Demand for Drop-Offs. At this project the assumed cost of unmet demand ( ) is 30, which could be the annual subscription of a customer of the network, who eventually cannot be served and does not use the network. In these two terms the values of and could be eliminated as they are not variables but 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 39

45 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT data being no subject of optimization. This means that the model is trying to minimize -BFkt and -BEkt, which means that it wants to pick up and drop off as many bikes as possible from and to the stations respectively meeting the biggest possible part of demand. The parameter is assumed to be a measure of the bike s popularity in a specific city. For example, if the citizens are keen riders, they would be willing to walk from their location to the nearest station so as to pick up a bike and use the network ( 1). However, if the bike is not a very popular means of transport in a city, then only few of the demand of a location where no station is established would be transferred to the nearest station. ( 0) Results In this paragraph, the results of 2 solved cases of the problem will be presented. In the first one it is considered that all potential customers of the locations, where no station is established, are willing to walk to the nearest allocated station and pick up or drop off a bike ( 1). In the second one it is assumed that only 50% of them would like to do this 0.5). The rest of the potential customers do not use the bikesharing network. All other parameters are the same for both cases. Table 1 depicts the results of both cases. The 50 candidate locations are grouped in the four clusters, the name of which is in the first column. The location No. and name were given by the authors of the paper. Columns 4 and 6 show the proposed bicycle parking slots (i.e. DNk ) each location should have for case 1 and 2 respectively. The locations with zero parking slots do not have an established station and their demand is transferred to the closest one. Columns 5 and 7 show the proposed bicycles (i.e. BNkt0 ) each location should have parked at the 4am in order to meet the maximum possible demand of the day. Table 1. The results of the 1 st and the 2 nd Case Case 1 (perde=1) Case 2 (perde=0.5) Cluster Housing No. of Location Location Name DNk BNkt0 DNk BNkt0 1 Ag.Paulos Axarnwn-Ioulianou Beikou Ippokratous Lofos Strefh Marasleios Mixalakopoulou Plateia Brazilias Plateia Gargaretas Plateia Lukabhttou Plateia Tasou Bourna nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 40

46 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Employment Subway Spare Time 12 Plateia Wdeiou Suggrou -Petmeza Teleferik Lukabhttou Filopappou Foithtiki Estia Foithtiki Estia Agalma Kolokotrwni Akadhmias ASOEE Axarnwn-Marni Pedion Arews Plateia Kaniggos Plateia Klauthmwnos Suggrou-Athan. Diakou Filellhnwn Akropoli Euaggelismos Thiseio Megaro Mousikhs Metaksourgeio Monasthraki Omonoia Panepistimio St. Larisis Suggrou Fiks Suntagma Aerhdes Ethniko Arxaiologiko Hrwdeio Thiseio Kafe Kudathinaiwn Panagia Kapnikarea Plateia Agoras Plateia Eleutherias Plateia Eksarxeiwn Plateia Filikis Etaireias Sthles Olumpiou Dios Stoa Athanatwn Psurrh nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 41

47 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 4. The established stations of the solution of the 1 st case categorized in clusters and with their size. (Reference: Figure 5. The established stations of the solution of the 2 nd case categorized in clusters and with their size. (Reference: Figures 4 and 5 depict the proposed established bike stations in case 1 and 2 respectively. The shape of each dot corresponds to the station s cluster, while its size represents the number of parking slots each station should have. In the first case the total number of docking stations is 34 and the number of parking slots is 517 making a mean value of 517/34=15.2 slots per station. Looking at the parking slots of each station, one can notice that the larger stations belong to the cluster 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 42

48 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Subway, which is typical of the increased demand in the metro stations. The total number of bikes in the network is 253 and their distribution over the established stations at the first time interval of the day shows that stations of the cluster Housing are nearly full of bikes in order to meet the increased Demand for Pick-Ups during the morning peak. On the other hand, the stations of the cluster Employment do not have many bikes. This results in having more free parking slots in order to meet the increased Demand for Drop-Offs during the morning peak. In the second case the established stations are 40 with a total number of 461 parking slots. This makes a mean value of 461/40= slots per station. The purchased bikes are 210 and one can make the same notes in this solution as in the previous one. Comparing the results of the two cases, it should be mentioned that there is a difference between them in the number and the size of the established stations. In the second case the model whenever it does not construct a station and serves the specific location from a nearby station, it loses 50% of its demand, which does not happen in the first case. For this reason, the second solution proposes more stations than the first one having less money to build enough parking slots and thus making them smaller. 5. SUMMARY AND CONCLUSION REMARKS Undoubtedly, public bike-sharing networks have gained a lot of attention during the last decade with a growing rate year over year. This is due to the benefits they offer as they are environmentally friendly and improve mobility in a densely populated city with traffic and pollution problems. The knowledge gained from the already implemented networks can and should be used for the design of future ones. This model reclaims the usage data from the Velib network of Paris to predict demand in Athens and designs a suitable bike-sharing network to meet that demand. In this paper a sensitivity analysis over one parameter was provided to show the changes on the solution. However, more parameters could be altered to notice how the solution changes. Such parameters could be the available budget or even the demand profiles to approximate the seasonal differences (winter-summer) or the week differences (weekdays-weekend). The larger application of the model is, finally, another work to be done concerning, for example, the whole Municipality of Athens. References (1) DeMaio, P. Bike-sharing: History, Impacts, Models of Provision, and Future. Journal of Public Transportation, Vol. 12, No 4, 2009, pp (2) Farahani, Z. R., M. Hekmatfar, B. A. Arabani, and E. Nikbakhsh. Hub location problems: A review of models, classification, solution techniques, and applications. Computers & Industrial Engineering. Vol. 64, 2013, pp (3) Shu, J., M. Chou, Q. Liu, C-P Teo and I-L Wang. Bicycle-Sharing System: Deployment, Utilization and the Value of Re-distribution. National University of Singapore, (4) Lin, J-R., T-H. Yang. Strategic design of public bicycle sharing systems with service level constraints. Transportation Research Part E, Vol. 47, 2011, pp nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 43

49 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT (5)Sayarsad, H., S. Tavassoli, F. Zhao. A multi periodic optimization formulation for bike planning and bike utilization. Applied Mathematical Modelling. Vol.36, 2011, pp (6)Martinez, M. L., L. Caetano, T. Eiro, F. Cruz. An optimization algorithm to establish the location of stations of a mixed fleet biking system: an application to the city of Lisbon. Procedia- Social and Behavioral Sciences. Vol. 54, 2012, pp (7) Garcia-Palomares, C. J., J. Gutierrez, M. Latorre. Optimizing the location of stations in bike-sharing programs: A GIS approach. Applied Geography. Vol. 35, 2012, pp (8) Juan P. Romero, Angel Ibeas, Jose L. Moura*, Juan Benavente, Borja Alonso. A simulation-optimization approach to design efficient systems of bike-sharing. Procedia Social and Behavioral Sciences. Vol. 54, 2012, pp (9)Froehlich J., J. Neumann, N. Oliver. Sensing and Predicting the Pulse of the City through Shared Bicycling. Proceedings of the 21st International Joint Conference on Artificial intelligence. USA, 2009, pp (10) Lathia N., S. Ahmed, L. Capra. Measuring the impact of opening the London shared bicycle scheme to casual users. Transportation Research Part C. Vol. 22, 2011, pp (11) Etienne C., L. Oukhellou. Model-based count series clustering for Bike-sharing system usage mining, a case study with the Velib system of Paris. Transportation Research-Part C Emerging Technologies. Vol. 22, 2012, pp. 88 (12)Institut National de la Statistique et des Études Économiques. Présentation de la région Ile-de-France. n/presentation.htm.accessed Jul. 1, 2013 (13)Hellenic Statistical Authority (EL.STAT). Population. Accessed Jul. 40 1, nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 44

50 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Research on the Development potential of Cycling routes Connecting Green Open Spaces in Athens Efthimios Bakogiannis 1, Maria Siti 2, Avgi Vassi 3, Georgia Christodoulopoulou 4 and Vasilios Eleftheriou* 5 1,2,3,4,5 Department of Geography and Regional Planning, National Technical University of Athens, GREECE. ( ebako@mail.ntua.gr, sitim.atm@gmail.com, avgi.vassi@gmail.com, geo_christ@hotmail.com, bileleutherioy@gmail.com) Abstract Cycling is constantly gaining popularity in Europe both as a leisure activity and a viable transportation mode. Development of policies and infrastructure in Greece is following a rather inconsistent way, especially regarding the integration of cycling in existing open public spaces. This paper presents a research, conducted from 2008 to 2010 in Athens, exploring the development potential of a dynamic cycling scheme which would connect existing large green parks, university campuses and other related uses. The paper is structured in three parts. The first presents a comprehensive overview of cycling promotion policies and similar schemes in major European countries and cities, such as Denmark, United Kingdom, Germany etc. The second part deals with particular infrastructure types and development attributes, and finally the third focuses on an Athens' case: the unification of Goudi metropolitan park with Police Academy park, Polytechnic and University campuses. The results and conclusions stress the key research facts and reveal the main limitations occurring by the complex ownership status and licensing procedures. Keywords: unification of green spaces, Athens, bicycle planning, green bikeways, bicycle access. Abbreviations: UoA: University of Athens NTUA: National Technical University of Athens 1. INTRODUCTION In many cases where car is seen as the trouble, bicycle is the solution. It is healthy, safe, faster than walking, cheap and friendly to other road users. Moreover, it adds on to the quality of a city and is easily accessible by almost everyone. Many cities in the world and even more in Europe, have chosen to integrate cycling as a viable transportation mode which, if combined with public transport and walking, provides the basis for sustainable mobility in urban centers. Furthermore, cities are enhancing their residents' relationship with the natural environment by developing recreational routes into urban 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 45

51 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT parks for walking and in many cases for cycling. Athens has significantly delayed to incorporate cycling in its choices for transportation and recreation, however there are encouraging signs for future development. Despite the fact that Athens is densely built, there is a number of concentrated large green open spaces in the close vicinity of the historic centre. However these spaces remain scattered, as the presence of supra-local activities and the neighboring urban highways interrupt their cohesion. In this research, the approach is focused on the development of a linkage to relate those separated green spaces, despite the challenges of the topography, through an upgraded cycling route. This scheme will increase the popularity and significance of the places as well as motivate Athens to assess the importance of greenways and cycling planning in the urban environment. 2. AIMS AND OBJECTIVES This paper aims to identify the development potential of a dynamic cycling scheme which would connect existing large green parks and university campuses in Athens, Greece. It examines all potential factors for the development of a leisure cycling network through urban parks in cities like Athens, by researching in depth similar European implementations in regard to planning principles, policy making and infrastructure attributes. The analysis is focused on the unification of greenways through cycling paths in accordance to existing facilities, geometrical characteristics, neighbouring land uses, ownership status and utility networks. The objective of the final proposal is the detailed design of a green cycling network which would serve both leisure activities for visitors and Athenians and transportation needs of the students at the UOA and NTUA. In this last part, the focus will be on detailed dimensioning of proposed paths, suggested materials, nodes and links design, particular landscape solutions, signage and parking. 3. CYCLING ROUTES IN OPEN URBAN SPACES AROUND EUROPE AND THE USA. AN EXPLORATION OF PLANNING PRINCIPLES, POLICIES AND CASES 3.1. A brief overview of National Cycling Policies in selected countries In most cities around the world bicycle planning and promotional campaigns are usually being carried out by local authorities, though there are some institutional processes in the national level, especially since 2004, when the European transportation ministers in Ljubljana signed a declaration pro national policies for bicycle promotion. This declaration set common targets and actions among the various ministries and government bodies, enhanced promotional activities in the political agenda and mainly aimed at encouraging and motivating local authorities to act in favour of cycling. Many countries worldwide have already shaped national planning principles and policies for bicycle integration in the urban environment, while others are making smaller or bigger increments in this direction. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 46

52 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT The current research looked at five of them in depth, namely Denmark, Great Britain, the Netherlands, USA and Greece, in order to determine their main characteristics and transfer knowledge and experience to the proposed scheme. The selected countries differ in numerous ways regarding cycling integration, with Denmark and the Netherlands having the largest bicycle use share worldwide (EC Eurobarometer, 2011), whereas Greece is acquiring one of the lower places in the list. In the case of Denmark, energy crisis and environmental awareness have already turned the Danes into cycling since 1975, while in the beginning of the 1980s they demanded better traffic conditions for cyclists through demonstrations and managed to introduce the National Day for Cycling. Since now, Danish national cycling strategy consists of 3 main programmes (E.C.M.T., 2004): the Cycling into the 21st Century which includes policies for promoting cycling for better cities and healthier citizens, as well as measures for the increase of bicycle use and deterioration of car travels, the Promoting Safer Cycling- A Strategy which focus on specific actions for cycling promotion (development of green zones, traffic calming, increase of bicycle infrastructure, bicycle parking etc.), the Collection of cycle concepts which aims at the dissemination of knowledge to promote cycling into local authorities. Moreover, Denmark has introduced the notion of the national cycle city (applied for Odense) and developed the national Bicycle Ideas Group. Great Britain's government was also funding and implementing innovative promotion plans for cycling already since the 1980s, though in scattered and isolated ways, in cities like Nottingham, Cambridge and Stockton. In 1996, the National Cycling Strategy report was issued by the Department of Transport, in order to raise awareness of citizens and authorities and encourage the development of policies and infrastructures that would enhance cycling use. The report set a target for 2016; to quadruple travels with bicycle in the coming 20 years. Other actions, quite at the same time, included the issue of a national study called Cycle-friendly Infrastructure: Guidelines for Planning and Design, by the Institution of Highways and Transportation, the Bicycle Association & Cyclists Touring Club and the Bicycle association. Later on, in 1998, the White Paper was published by the English Department of Transport promoting further walking, cycling and public transport infrastructure. Since then and mostly after the '00s, Great Britain's government supports widely local mobility plans, which include a number of policies and practices to benefit cycling. Regarding the Netherlands, the fact that 13 million of bicycles were in circulation in 2013, in a country with a population of 15,5 million, reveals practically the national vision pro cycling. Between 1990 and 1997, the Dutch government issued the national agenda under the name Dutch Bicycle Master Plan, and managed to complete 112 projects, which included 31 research programs and 41 pilot implementations aiming at the enhancement of infrastructure, improvement of safety issues, increase of bike parking facilities and the reduction of bike thefts. The Second Transport Structure Plan 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 47

53 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT as implemented in the 1990s also aimed at the decrease of car usage by almost 50%. A particular element in the Dutch strategy was the involvement of local authorities in the initial configuration and further support of national policies, which conduced to the actual boost in cycling. The central government transferred a number of its powers to the local units, preserving though its responsibilities as a knowledge consultant, research and infrastructure financier as well as kept assessing and monitoring all the ongoing programs. Regarding the U.S.A., national cycling policies have slowly emerged due to the realization of lack of funds and the fact that the extension of new road infrastructure was leading to further traffic jams. In the 1990s, the new legislation act Intermodal Surface Transportation Efficiency Act imposed the integration of bicycle planning in all local transportation plans, the financial support for public transport, walking and cycling infrastructure and obliged each state to appoint an expert in walking and cycling policies and plans. In the same decade, an influential new act, the Intermodal Equity Act for the 21st century defined bicycle as an equal transportation mode to cars and public transport, boosting its use in some of the states and attracting more funds. Lastly, Greece has significantly delayed to introduce and implement a national agenda for cycling integration policies, however there is a number of initiatives generated by both ministries and researchers' groups. Most of the actions so far though, had scattered approaches and dealt with cycling as a viable alternative and by no means as an equal element in transportation choices. In 2001, the Ministry of Transportation and Communication assigned at the NTUA a research program for cycling integration in 17 Greek cities and dedicated specific amounts of funds for each of them, though only 4 of them have absorbed the given resources and built part of the proposed networks. Later on, in 2003 the Ministry of Internal Affairs encouraged local municipalities to promote cycling through a funding scheme, which unfortunately was impended due to obstructionism and lack of political will Indicative review of particular planning policies and implemented plans in European cities As mentioned above, although the importance of national strategies for cycling is high, local authorities and municipal government bodies are those who implement most of the schemes and substantially drive integration attributes. The researchers looked in depth at a large number of cities in Europe, both the so called 'traditional' cycle cities and others with small bicycle share, regarding their performance in cycling promotion. The table below (Figure 1) summarizes the findings in six of the studied cities. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 48

54 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT City Population (2004) Area (km 2 ) Total Bicycle network length (km) Cars per capita ratio (passenger cars/1000 people) Trips by car (work related activities) (%) Bicycle use (%) Goals and Financing The latest plan for promoting cycling ( ), had these goals: Copenhagen % 34% Increase of bicycle use, for travels from and to work, up to 40% Increase of safety and 50% reduction of accidents possibility 10% increase of bicycle traffic speed (>5km route length) Odense % 26% 4-year project, containing 50 sub-projects. Total investment so far reaches 2.7 million Euros. Strasburg n/a 12% The aim of the city is to reach 25% in daily bicycle use share. Freiburg % 20% Total investment so far reaches 13 million Euros. Geneva n/a n/a 4% Total investment so far reaches 4 million Euros. Figure 1. Key attributes of the studied cities, regarding cycling promotion goals and results 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 49

55 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT In general, it was observed that: Copenhagen has developed extended infrastructure, integrated at a clear and readable bicycle network. It also has ongoing projects upgrading the existing substructures and constantly applies new policies and plans pro cycling. City bike, a bike sharing scheme with free bikes well distributed at numerous focal points (125 stations) in the city, was one of the exceptional paradigms in the city, dating back to Moreover, intersection improvements works are constantly developing and researches on users satisfaction are evolving. Information and awareness campaigns also increase the popularity of cycling and all its related activities and upcoming plans. Freiburg has also invested a lot on cycling upgrade and has prioritized walking and cycling in its historic centre. It is considered one of the highly committed cities to sustainable mobility approaches and is steadily restricting car usage. Geneva, although it is not considered to be a traditional cycle city, has managed to increase its daily bicycle use percentage from 2% to 4% in a 10-year period and supports further plans and campaigns which increase bike use by 0,5% each year. 4. REVIEW ON INFRASTRUCTURE TYPES AND DEVELOPMENT ATTRIBUTE The types of cycling routes in green spaces, as developed internationally, vary depending on the surroundings (rural and/ or urban), the facilities they connect and the aims they serve. Greenways are attractive recreational routes for pedestrians and cyclists and are usually found in forests, near rivers, lakes, or even seashores, close to abandoned rail lines, inside urban parks, or close to pre-existing promenades. Some of the successful common practices are redeveloped linear connections, among existing green spaces aiming at the unification of open spaces around cities, which is the case of the current scheme in Athens. These cycling routes are either exclusive bike lanes or shared spaces (with pedestrians) integrated at leisure zones. Their geometrical attributes, such as width, cross-slope/ inclination, curve radius as well as the surface materials, derive from specific regulations as defined by each country, with most of them sharing the key attributes. The fact that the typical regulations of most countries refer only to cycling paths incorporated to the street environment, allows the planner a level of freedom in designing paths inside urban parks, however some indicative features are usually followed. In this part, a quick overview of the main regulations is presented, with a differentiation for cycle lane width prerequisites in Belgium, Great Britain, France and Germany. The key planning principles are to be followed in the case of the Athenian green cycling route. Cycle lane width In Belgium the minimum width- for exclusive lanes- is 1.2meters if there is no on-street parking, which is increased at 1.75m for parallel delineation with car parking. For 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 50

56 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT recommended cycle lanes- but not exclusive (marked with white dashed lines), the minimum is 1.3m (when there is parking), and 1m when there is not. In Great Britain the minimum width for one-way lane is set to be 1.5m while regulations suggest 2m, and for two-way lanes the minimum is set at 3 meters. France sets the width range for an exclusive one- way lane at 1.2 to 2.5 meters, while considers as ideal 1.7 meters. When it comes to recommended lanes the width can be between 0.75 and 1.2m. Lately in Paris one way bike lanes have to be meters wide and two-ways must be at least 2.5m. Lastly, in Germany one-way bike lanes can be between 1.5 and 1.85m. Cycle lane cross- section The suggested cross-sections are always set in accordance to the length of upslope, the local wind conditions, and the potential momentum developed by the cyclist. Moreover, at the beginning of the cross-slope, there cannot be any intersections with traffic lights or intersections where other traffic streams are prioritized. and similarly for road sections with declivity. The maximum slope for urban cycle lanes is defined at 6%. Figure 2 below shows the eligible inclines according to specific cycle lane lengths and Figure 3 estimates the width increases to be planned in the indicative inclined section. Figure 2. Eligible inclines according to specific cycle lane lengths. Source: Vlastos& Birbili, 1999 Length of cross slope section (meters) Incline (%) 25-75m >3 & <6-20 cm 30 cm >6 & <9 20 cm 30 cm 40 cm >9 30cm 40 cm 50 cm Figure 3. Width increase according to the length of inclined section Source: Vlastos& Birbili, nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 51

57 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Curve radius in urban areas: cycle fast lanes (with a 30km/h speed limit): R > 20m secondary cycle lanes (with a 25km/h speed limit): R > 15m cycle paths (with a 20 km/h speed limit):r>10m in rural areas: cycle lanes (with a 30 km/h speed limit):r>20m It should be noted here that there is a number of cases where turns are designed with smaller curve radius in order to reduce the cyclist's speed, with the minimum radius being 4 meters. Infrastructure principles and observations According to Vlastos & Birbili (1999), the surface level of the new infrastructure, in both cycle corridors and lanes, should be at least 7.5cm above traffic level for flooding purposes. Indicative cross sections of cycle lanes at figures 4 and 5 depict some of the alternative configurations with cobbles or cast cement materials. Figure 4. Cycle lane cross section with cobble stones 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 52

58 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 5. Cycle lane cross section with cast cement Colours, materials and signage also play an important role, as they define different functions and ensure a higher level of safety. The usual materials for cycle lanes should be smooth, porous and reflective, while in new schemes a lot of attention is paid on recyclable materials. Materials vary from typical asphalt mix to concrete cement, resin mix materials, cobble stones et cetera and are chosen according to the regeneration aims, funding, weather conditions and other factors. Also, the systematic maintenance and renewal of pavements for both pedestrians and cyclists is an important matter of safety. Clear, readable and frequent signage is crucial for safety and should include both traffic signs/ labels (figure 6) and horizontal coloured stripping in order to notify all road users. The height of traffic signs is usually between 1 and 1.5meters. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 53

59 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 6. Bicycle traffic signs 5. THE PROPOSED SCHEME. STUDY AREA IDENTIFICATION AND ANALYSIS 5.1. Identification and analysis An in depth analysis was conducted, prior to the proposal, in the study area regarding previous studies, existing land uses and facilities, focal service nodes, walking and cycling paths, transportation choices, ownership status and a number of other parameters. The study area concentrates various supra-local services and includes large areas of open green spaces, which however remain scattered and underused for recreational purposes. Location, accessibility and attraction nodes The area stands in the eastern part of the Athenian basin and is surrounded by Mesogeion Ave., Katehaki Ave., Ymitou ring road, and Karamolegkou Street. The neighboring municipalities are Vyronas, Kaisariani, Papagou and Athens. Figure 7 depicts the key attributes of the area. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 54

60 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 7. Study area The area is highly accessible through metro stations (Katehaki, Megaro Mousikis, Evangelismos) and numerous bus lines. However, the presence of cars is intense and this is mostly due to the area's increased accessibility through the neighbouring urban highways. It embraces the two larger university campuses of Athens (University of Athens-UoA and National Technical University of Athens-NTUA), the Police Academy Park and Badminton facility, which is now turned into one of the most important music and sports arena. It also stands nearby essential services such as hospitals (civil and military), the Medical School, military camps, special police centres, municipal farmsteads etc. The area presents a number of advantages regarding its potential for cycling promotion, as it is close to the Athenian city centre and is still full of green, has adequate transportation features, diverse landscapes and interesting cityscapes. Indeed, a number of issues arise when it comes to activities that are fenced and inaccessible and others that operate only during the day, both of which create a sense of insecurity and solitude. Moreover, each of the afore-mentioned activities drive a number of travels. Goudi park, which includes the Police Academy park, the Stratou grove, Zografou Park as well as many of the Olympic facilities (badminton arena, riding centre, swimming centre), attracts residents from Zografou, Goudi as well as employees of the neighbouring facilities. Other Athenians do also visit the facilities in weekends and holidays. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 55

61 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT The two university campuses (NTUA and UoA) contain the educational departments, students dormitories, rectors and management buildings, libraries, sports facilities and others, and attract almost people per weekday. Existing land uses The detailed study of the existing land uses in Goudi park is depicted on the map below. Ownership status Figure 8. Existing land uses One of the most important issues while conducting this study was the identification of the ownership status in the delineation zones of the preliminary cycle lane configuration. The institutional bodies and stakeholders involved in owning and managing the area are listed below; Municipality of Athens Municipality of Zografou Municipality of Papagou Olympic Properties NTUA UoA Ministry of Public Order and Civic Protection Ministry of National Defense Ministry of Communication and Mass Media Obviously, the status is extremely complex and has a long history of concessions. According to studies, the wider study area has 53 different stakeholders and the 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 56

62 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT planning regulations vary in many ways. Statutory and implemented boundaries often differ, developing further conflicts, while also some of the plots abide in controversial legislation. Furthermore, the high presence of natural elements and upgraded landscape have led to additional protective legislation context. The different ownerships are usually separated through walls, railings, fences and other structures (figure 9) as they facilitate vulnerable uses, which is preventing further the unification of the spaces. Throughout the preliminary study of the proposed cycling route, the research team identified two of the key ambiguous land zones (blue signs at figure 13) in terms of ownership complications, with one being the transition zone between the NTUA and UoA campus and the other being the linkage of Goudi Park to Katehaki metro station through the Police Academy and Greek state broadcaster (ERT). Previous studies Figure 9. Different ownerships and structural obstacles It should be noted here that the Organization for Strategic Planning in Athens is promoting the strategy of Green Spaces Unification focusing on the whole Athenian conurbation for years, though with contentious results. Previous studies have focused on the enhancement of the wider Metropolitan Goudi park, with one of the most important being the research by the Urban Environment Lab of NTUA since This study was dealing with the planning of Goudi Metropolitan Park - an area of acres, which would integrate the pre-existing activities (education, administration, health care and sports facilities) with new uses for culture, sports and recreation. The U.E.Lab had proposed the determination of two distinctive development zones (figure 10): the central zone (1.450 acres) which would be the main regenerated public 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 57

63 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT space for recreation with cultural uses in the existing listed buildings and other facilities, supported by walking and cycling routes and the secondary development zone, expanding in an area of acres, which would incorporate the educational facilities, health care and other administration bodies that are located within its boundaries, in a unified scheme of conservation, protection and enhancement. Figure 10. U.E.Lab NTUA proposal_ Two development zones of Goudi Metropolitan Park Source: U.E.Lab NTUA The same research had also proposed traffic arrangements and specific landscape intervention among others. Moreover, it demonstrated 3 development cores, one of which will be the basis of the current proposal in this paper. This was the first linear development, starting from Mesogeion Avenue close to Katehaki metro station, crossing the Metropolitan Park in SE-NW direction and ending at the NTUA campus. In the proposal, one of the main suggested infrastructure was the opening of an underground passage linking the station to the enhanced route, allowing students and employees of the campuses to cross in a safe and upgraded way the existing large crossroads of Mesogeion and Katehaki avenues. In the above direction, there are some completed individual schemes inside the NTUA campus that will be incorporated in the current proposal, such as the policy for deterioration of through traffic and speed reduction with speed bumps as well as the completed cycle lane (figure 11 and 12) that circulates the institution. The cycle lane is configured a few meters away from the street, very close to walking paths and has a width of 2.5 meters. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 58

64 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 11. Image of the completed cycle lane inside the NTUA campus Source: Authors' archive Figure 12. Plan of the completed cycle lane inside the NTUA campus This existing cycle lane is in use since 2005 and attracts daily several students and athletes from the wider area. It is also used by residents of Zografou, Goudi and Ilisia area, however is completely segregated from the other neighbouring suburbs and facilities due to the extended presence of highways and becomes even more hostile to reach it due to the absence of proper signalling. Indeed, its spatial contiguity to Katehaki metro station as well as the planned extension of the metro line to the campus is expected to attract more cyclists in the coming years Proposal description As mentioned earlier the suggested cycling route aims at the enhancement of leisure activities in the study area, as well as the promotion of cycling culture in Athens as a viable transportation mode. The key paramaters taken into account were: 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 59

65 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT any pre-existing studies for the wider study area the enhancement of the natural environment and utilization of downgraded infrastructure the connection with main attraction nodes and transportation stations the connection between existing controversial land uses the potential linkages with the neighbouring municipalities the deterioration of through traffic links and introducing of traffic calming measures the enhancement of leisure activities the protection of green open spaces and the needed removals of fencing and walls the origin-destination survey conducted by the research team for NTUA and UOA The study of the detailed cycling route included a number of sub-studies, after the configuration of the preliminary study route. Firstly the research team went though the detailed mapping of the suggested route and produced topographic maps. The instruments used, in order to establish a geodetic control network and measure the detailed features of the area, were a G.P.S. Topcon HyperPro, and a total station Leica TCR407power. Afterwards, the study focused on the identification of 4 distinct sections (figure 13), namely the part at Goudi park, the part at NTUA, the part at the UOA and the part at Ilisia grove, as seen in the figure below. Figure 13. Cycling route proposal - 4 studied parts 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 60

66 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT The part of the cycling route at Goudi park has a length of meters and links the universities and Katehaki metro station, while it also operates autonomously inside the Police Academy. We studied three alternative scenarios for the delineation of the route for the link to the metro station, with one using the existing street network of Mesogeion Avenue, the second passing through Goudi park and the third utilizing Katehaki Avenue. Finally the first alternative was chosen and the parts inside the park were aligned to serve all the related activities equally. The final route avoids where possible the conflicts of ownership. The part of the cycling route at the NTUA campus has a length of meters. It was kept at its existing delineation and special attention was paid at the linking parts (in entrances and exits of the campus) regarding signaling and intersection improvements. The part of the cycling route at the UOA campus has a length of meters and used some of the inner-campus street networks as well as an existing long pedestrian path that runs through the center of the campus, due to the particularities of the landscape. It links educational departments, parking places, sports facilities and many more. The latter part of the cycling route which runs through the Ilisia grove has a total length of meters and was inscribed on existing paths and unpaved roads (soil), achieving to circulate the whole area of the grove. The width of the cycling lanes varies from 1.6 to 2 meters and structural details were provided for each part (through maps, cross sections etc.) which specify materials, curve radius, cross slopes and all analytic calculations for the construction of the scheme, as well as the allowed deviations from the prerequisites. The selection of surface materials was studied thoroughly and suitable materials and coatings were selected according to the specific needs. Among the various alternatives, we chose cobblestones, natural granite paving blocks, photocatalytic coatings as well as compressed earth blocks and mixes of granular materials. All materials followed the European Standards prototypes (EN1342, EN1338:2003, ΕΝ 13369, EN1342 etc.). Analytical solutions were also provided for all intersections and linkages, following national and international standards in regard to optimal conditions, safety and accurate signaling. Urban furniture proposals were demonstrated and indicative locations were provided for benches, lighting, taps, litter baskets etc., based on the development of a uniform route with wooden structures in order to enhance the existing natural environment. Moreover, places for temporary rest and recreation were located at a frequent base throughout the whole cycling route and parking places were demonstrated at focal points. Complementary to the above, spaces for automatic bike rental were indicated and details on a proposed bike sharing scheme were also determined. Lastly, the research team reviewed the construction works and set specific prerequisites for excavations, concrete and asphalt works, lighting and signaling, coatings, drainage etc. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 61

67 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT 5.3. Economic plan The study also looked into the economic plan of the suggested proposal and calculated the expenses regarding the construction and surveillance works. The total cost of the construction is estimated to reach 1,62 million Euros and the funds will be provided by the NSRF funds of Attica Prefecture. According to the NSRF, 70% of the total expenses will be invested by EU funds and the remaining 30% will be supplied by Greek national funds. 6. CONCLUSIONS AND RESEARCH LIMITATIONS Developing cycling routes in open spaces and greenways for recreational purposes is a common practice for urban areas, as it enhances the relationship of the residents with their natural environment and upgrades the urban landscape. The proposal seeks to connect supra- local activities in the Athenian urban environment, through a green cycling route subject to the European prototypes, as well as to promote cycling as a viable mode of transportation. Greece lacks official technical requirements for cycle lane construction for both the urban and rural space. Since now, the construction of cycle lanes was following the literature available by the NTUA as well as some guides by the Greek Ministry of Transportation. The latter was forced to form a committee to develop such requirements in 2013 and since then little progress has been done. As a result, the research has conducted an in depth exploration of foreign principles and planning policies and has adjusted the requirements in the Athenian context. According to the Greek legislation, after the completion of a study and before the implementation of the proposed infrastructure there is a number of procedural steps with the most important being the stage of maturation. This stage includes authorizations and audits from the local authorities and the Ministry of Transportation, environmental approvals and legitimacy controls etc., which last approximately 12 to 16 months. Due to the aforementioned procedural complexity and in order to avoid further delays arising from the ownership conflicts, the research had to explore in depth all the alternatives in cycle lane layout. The study area belongs to a number of stakeholders, such as municipalities, public authorities, military and academic institutions and consequently the authorization of all the above is difficult and many times impossible for issues of privacy and safety. Hence, the proposed layout was modified to deal with all the aforesaid matters as well as with some topography implications in the close vicinity of the university campuses. Other issues that emerged were related to the linkages with metro stations and were resolved by taking up space from traffic lanes and parking spaces. In general, the research attempted to approach the matter in a holistic way and examined similar implementations and planning principles that drive such schemes in the European context. The overview of the infrastructure elements helped at the formulation of the layout configurations and the final product is a complete and robust cycling route that is expected to unify many of the neighboring uses and regenerate the 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 62

68 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT studied open spaces. The provision of the detailed dimensioning among the analytical description of materials, signage, parking facilities et cetera can also be considered a useful precedent in similar future studies. Acknowledgement The authors would like to thank Prof. Thanos Vlastos, who was the scientific director of this research as well as Dr. Dimitris Milakis, Sofia Papastrati, Trisevgeni Papagerasimou, Ioannis Marakakis, Christos Karageorgopoulos, Angelos Liveris and Kostas Liveris for their significant contribution in the research. References City of Copenhagen, Cycle Policy , Building and Construction Administration, Roads and Parks Department City of Copenhagen, Bicycle account 2002 Department of Transport (U.K.), 1996a. Cycling in Great Britain. Transport Statistics Report, H.M.S.O., London. Department of Transport (U.K.), 1996b. The National Cycling Strategy. DoT, London. E.C.M.T., National Policies to Promote Cycling, Implementing Sustainable Urban Travel Policies: Moving Ahead, European Conference of Ministers of Transport, Organization of Economic Coordination and Development. European Commission, Eurobarometer: future of transport analytical report. Flash Eurobarometer 312 The Gallup Organization, Hungary. European Commission, Cycling: the way ahead for towns and cities, European Communities, Office of Official Publications of the European Communities, Luxemburg European Commission, The European Greenways Good Practice Guide: Examples of Actions Undertaken in Cities and the Periphery, European Greenways Association and Direction General Environment, Brussels Golias, I., Kanellaidis, G., Polyzos, I., Giannis, G., Mertzanis, F., Blana, E., Sismani, O., Vezyroglou, A. and Ravani- Melissari, A., Traffic management at technical NTUA campus. Technical report. NTUA: Department of Transportation and Transport Engineering Hopkinson P., Wardman M., Evaluating the demand for new cycle facilities, Transport Policy 3 (4): Martens K., The bicycle as feedering mode, experiences from three European countries, Transportation Research Part D 9: Mc Clintock H. (edit.), Planning for cycling: Principles, practice and solutions for urban planners, Woodhead Ministry of Transport (Denmark), Collection of Cycling Concepts, Road Directorate National Technical University of Athens, School of Architecture, Department of Urban and Spatial Planning, Laboratory of Urban Environment, Goudi Metropolitan Park- Ilisos (Research Programme). Athens: NTUA. Newman P., Kenworthy J. (and oth.), «Car Free» Copenhagen, Perspective and ideas for Reducing Car Dependence in Copenhagen, Department of Urban Design, Royal Danish academy of Fine Arts, Copenhagen, Institute for Science and Technology Policy, Murdoch University, Perth, Australia 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 63

69 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Pucher J. Komanoff C. Schimek P Bicycle Renaissance in North America? Recent trends and alternative policies to promote bicycling, Transportation Research Part A 33: Rietveld P., Non-motorized modes in transport systems: a multimodal chain perspective for The Netherlands, Transportation Research Part D 5: Sani, M., Could better roads boost commuting? Bicycle Retailer and Industry News Sustrans, Connect2 and Greenway Design Guide. UK: Sustrans. Tolley R. (ed.), The Greening of Urban Transport. Willey, Chichester. Vlastos, A. & Birbili, T., Policies and Configurations for cycling integration in the Greek city. European Union G.D. of Environment, Development Company of Athens, Strategic Plan and environmental protection organization of Athens, Athens. Vlastos, A. & Birbili, T., Developing cities for cycling. Structural and landscaping elements. Athens: Development Company of Athens, Municipality of Athens, Strategic Plan and environmental protection organization of Athens, Mbike Vlastos, A., Barbopoulos, N. and Milakis, D., Bicycle. A guide on planning and assessing bicycle networks. Athens: Technical Chambers. Vlastos, A., Barbopoulos, N. and Milakis, D., Cycling and the environment. Research on social, spatial, traffic and institutional requirements for cycling integration at sustainable mobility policies in Greece. Proceedings at Pythagoras- Day Conference for scientific research promotion at NTUA. Lesvos, Vlastos, A., Birbili, T., and Barbopoulos, N., The bicycle in Greek cities- Integration policies. Athens: YPEXODE/ Strategic Plan and environmental protection organization of Athens, Mbike. Vlastos, A., Milakis, D. and Athanasopoulos, K., The bicycle in 17 Greek cities- A study guide. Athens: OEDV Zegeer C., (ed.) F.H.W.A. Study tour for pedestrian and bicyclist safety in England, Germany and the Netherlands. US Department of Transportation, Washington, DC. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 64

70 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Shift from Car Cities to Soft Mode Cities? Learning from Best and Worst Practice? Abstract Anders Langeland PhD. University of Stavanger In this paper the Top Twenty of cycling cities in Europe is presented. Münster in Germany tops the list and is the European Bike City. In Münster 38 percent of all trips are with the bike, and 36 percent with the car. The new TEMS EPOMM database makes it easy to compare modal split in cities and thus answer questions like: Which city is most transport sustainable? Which city is the most car dependent? It is not surprising that they cycle a lot in Dutch cities, nor in Copenhagen, possibly more surprising that the bike is used extensively in Berlin? Milton Keynes has indeed become a car city, 75 percent of all trips is with a car! A line can be drawn through Europe. North of the line one find cycling countries and south of the line car countries. Belgium is divided in two. Cities in the Flemish part to the north cycle a lot, while in Valona they hardly cycle. The paper discusses some of the findings and gives tentative answers to these differences between countries and cities. Keywords: World Bike Cities, Modal Split, Sustainable Transport, Best Practice. 1. INTRODUCTION To meet the climate challenge and the urban transport problems everybody from the EU, most nations and down to most cities have set goals to reduce auto dependence and shift to public transport, cycling and walking. However the gap between the goals and the increasing car traffic seem to widen (Newman and Kenworthy, EEA). Some cities like Freiburg, Copenhagen and Amsterdam are presented as front runners in achieving the shift from car to more environmentally friendly modes. Cycling cities have become a hot topic in marketing and very often one can find lists of cities worth visiting in airflight magazines and others. This information is more and less reliable. Sandnes, Norway, was placed 7th in the world, but in fact only 4% cycle 1. This paper has used the TEMS EPOMM database to rank cities. Which is the World Bike City? Which Capital is best on cycling? Which city is most car dependent? The different rankings are discussed and some tentative answers to why countries and cities differ so much are given. The paper concludes that there is much to learn, but implementing other cities success is not straightforward. 2. METHOD There are difficulties in measuring modal shift due to lack of data and no common standard of measurement. The author has compared 4 cities in 4 countries (Kristiansand, Aalborg, Norwich and Davis). It was difficult to get hold of data and in many cases there was no information on how the data had been produced. Very often 1 Virgin Flight Magazine, spring nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 65

71 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT the data was collected for a specific purpose e.g. city branding, instead of monitoring the changes over time and evaluating the policy. (Langeland 2009) The comparisons of cities across country borders have been fraught with difficulties: lack of data, privately owned data, non-comparable data and not least political data (some glossed some hidden) used for propaganda and as strategic misrepresentation. When the aim was to sell the city, obtain money or favors from higher administrative levels, strategic misrepresentation was not uncommon. Strategic misrepresentation to support certain political goals and/or obtain political goodwill is widespread, neither is deliberate lying to support a certain project or policy uncommon (Flyvbjerg). Booth raises several questions regarding cross national comparisons: How to identify suitable countries for research? The identification of the sample cities? Data availability. Common basis for comparison? Data collection? Language problems? (Booth, 1986) The recently established database TEMS a huge step forward. TEMS The EPOMM Modal Split Tool, is a database that was made with the support of intelligent Energy Europe in the project EPOMM-Plus. Since the start of TEMS in May 2011 initiatives have been taken to collect city data. All uploaded cities are checked and located on the map by the TEMS administrator. There is still not a common standard or definition for: metropolitan area, urban area, city/municipality/district area, they differ both in size and population according to source. Trip purpose varies, some estimate all trips, some business trips, while some only work journeys. Walking is treated differently from survey to survey, and obviously wrong for many cities. Data collection methods are several: travel surveys (national or local), cordon counts, mode counts. The TEMS aim is that in the future each city will take responsibility for their modal split data and uploads and control the data themselves. TEMS is a splendid tool which hopefully will promote more data from more cities and with standardized survey methods and quality controls. Even with some shortcomings (overcome in 5-10 years?) TEMS makes it possible to compare environmentally sustainable transport development in European cities and also make inferences about countries. 3. THE GOOD THE BAD AND THE CAPITAL CITIES 3.1. The top twenty cycling cities The following table shows the top twenty Bike Cities in the world. Münster in Germany is the cycling city of the world with 38% of all trips on the bike. Four of the top cities have a cycling share over 30 percent. The population data has come from the TEMS database and might or might not be the population that the mode split data covers. The table is limited to cities in which the car share is below 50%, hence some cities with both a high bike share and high car might be exempted. Seven of the twenty cities are German, five are Dutch, three Swedish and two from Denmark. One city is included in the table from Italy, Switzerland and Belgium. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 66

72 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Table 1. The world top Bike cities City Population Walk Bike PT Car Münster Leiden Copenhagen Groningen Bolzano Freiburg Uppsala Odense Göttingen Haarlem Lund Bremen Heidelberg Antwerp Malmö Amsterdam Kiel Basel Potsdam Den Haag The table can be sorted and looked at from different angles, for example top walking cities, top public transport cities, etc. The sum of walking and cycling is presented as top environmentally sustainable cities in this paper. Figure 1. Top 20 Bike Cities 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 67

73 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 2. Top Twenty Low Car Use Cities Figure 3. Top Twenty Environmentally Sustainable Cities The above figure shows a ranking according to the highest sum of walking and cycling. All cities have more than 45% bike/walk share. In addition one should look at low car use and then Basel comes out as best with only 23% car share. Bolzano the Italian city which has shifted between nations after the first World War, is very untypical of the Italian cities coming to walking, cycling and bike use. They do cycle and walk a lot in Bolzano, as such the city resembles its neighbor to the north, Innsbruck. However, the data seem inflated and it has therefore been not been included in some of the next comparisons of good examples Münster The world bike city The mode split development in Münster is shown in the following table and figure. In 2001 there was a swing back to higher car share, which was reversed in Hence, to maintain a level of cycling at 38% will require continual effort in the next years. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 68

74 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Table 2. Münster Time series Modal Split Walk Cycle PT Car Figure 5. Münster MS When cycling increases in Münster, the walking decreases. Less than 40% car share is very good, but it raises the question if it is possible to decreas the car share further in Münster. It might be that neither public transport nor cycling are alternatives to the car for many of the trips in the outer area of the city? Figure 1. Münster Mode Split 2007 Münster is special when it comes to using the bicycle, 38 % of all trips. Other famous German cities for their transport policy and achievements like Freiburg and Karlsruhe has a smaller bike share, 19 % and 16 %. There is about 6 % less walking, 3 % less car passengers and 10 % fewer public transport passengers in Münster than the other two cities. It seems likely that the good cycle system attracts former car and public transport 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 69

75 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT passengers and some who formerly walked. It may also be that the bike system generates its own traffic. About 30 % of all trips were car driver trips. It may be that this is about a lower threshold for car usage where other modes can t compete because of travel distance, too little demand and/ public transport not viable, car used in work or combined journeys The bad ones the car only cities In the following table the bad cities when it comes to car use are listed. The list is limited to three cities from each country. In the top ten of the worst car cities there are three cities from UK, three from Belgium and three from France! Among the top twenty the Netherlands, Italy and Norway has three cities on the list. Table 3. The car only cities Country City Car share % UK Swindon 85 Belgium Charleroi 84 UK Blackpool 80 Belgium Liege 76 UK Milton Keynes 75 Italy Parma 75 France Beaujolais 74 France Vienne 74 France Étang de Berre 73 Belgium Namur 70 Norway Sandnes 70 Italy Livorno 70 Italy Verona 69 Netherlands Haarlemmermeer 65 Denmark Aalborg 64 Norway Kristiansand 63 Norway Stavanger 61 Netherlands Emmen 61 Netherlands Hertogenbosch 60 Sweden Norrköping 59 Finland Jyväskylä 58 Germany Wiesbaden 57 Austria Linz 49 At the bottom of the list one finds the worst car city in Austria with only 49% car share, and the worst in Germany, Finland and Sweden with 57, 58 and 59 per cent respectively. Even better is the Swiss cities, all with car share less than 37%. It is obvious that these countries control the use of the car in cities far better than those on the top of the list. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 70

76 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT 3.3. The Capital Cities Figure 2. High Care Share Cities The mode split in the capitals in the different countries are shown below, sorted by population high to low. Only three of the capital cities have bike share above 10%. Note that some of the data is for the metropolitan region e.g. Stockholm, while others are the city of municipality, e.g. Copenhagen. This give large differences in modal split. Table 4. Mode split in the Capital Cities City Population Walk Bike PT Car London Berlin Madrid Lisbon Paris Bucharest Stockholm Warszawa Budapest Wien Brussels Amsterdam Helsinki Oslo Copenhagen Edinburgh Zürich nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 71

77 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 3. Top Capital Bike Cities Only three of the capitals have more than ten per cent bike share. The capital city in Denmark, Copenhagen, is on top as a bike city, it is on top when walking and cycling are combined, and it has a car use of only 29%. Copenhagen has several decades had increased cycling as a key strategy. From 1995 it has published a Bicycle account measuring progress towards goals. From 1995 to 2010 the number of serious cyclist casualties fell from 231 per year to 92. Still the 2015 goal of 59 serious cyclist casualties is a bit away. The kilometres cycled in Copenhagen increased in the same period from 0,80 to 1,21 (million km per day) in % cycled to work in 2010 slightly down from 2004, but far from Copenhagen s very ambitious goal of 50% in 2015! Figure 4. Capital Cites, Best Bike and Walk 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 72

78 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Figure 10. Capital Cities Top Bike & Walk Figure 5. Capital Cities high Care share Stockholm is one of the most interesting cities in this comparison, well known for the City Plan and Transit Oriented Development from the early fifties. Stockholm is arguably the best example anywhere of coordinated planning of rail transport and urban development and Overall, experiences in greater Stockholm reveal that transit villages are not isolated islands within the larger metropolis, but rather are dependent upon each other as well as major urban centers. Clearly, jobs-housing balance and selfcontainment are not prerequisites to reducing automobile dependence. (Bernick and Cervero, 1996) Mode split in Stockholm show a low share of walking and cycling, relatively high share of public transport, probably due to the excellent metro system to and from the center, but it has also a relatively high car share, probably car travel crisscrossing the region and avoiding the city center. In the long run one probably will see that the location of jobs also will move outwards and contribute to increased car use. The Congestion Charging around the city center will support such tendencies to increased automobile dependence in the Stockholm region. The city has a long way to go before it reach a modal split as in Zürich or Vienna and should attempt to increase walking and cycling substantially. What Stockholm need is a shift from cars to walking and cycling, however possibly very difficult judged by the urban structure and location tendencies? 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 73

79 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT 3.4. National differences There are striking differences between cities across national borders and also within nations. Netherlands, Denmark and Germany are the countries with several cities that are Bike cities. Table 5. Nations with bike cities above 12% Country Bike cities Percent cycling Netherlands 26 cities above 12 percent cycling. Germany 21 cities above 12 percent cycling. Sweden 8 cities above 12 percent cycling. Finland 5 cities above 12 percent cycling. Denmark 4 cities above 12 percent cycling. Italy 4 cities above 12 percent cycling. Austria 3 cities above 12 percent cycling. Belgium 3 cities above 12 percent cycling. Switzerland 2 cities above 11 percent cycling. United Kingdom 1 city above 12 percent cycling. There are some countries hardly use the bike: Spain, France, Italy and UK Belgium Belgium is a special case. In the Flemish North, the cities of Brügge, Ghent and Antwerp are cycling cities with a high level of cycling resembling that in Dutch cities. In Valona to the south, hardly anyone cycle, just as in France further south. Figure 6. Belgian Cities Mode Split 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 74

80 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Great Britain Figure 7. British cities Mode Split Cycling has not a large modal share in Great Britain. Only Bristol resembles European cycling cities. It is the very high car share that single out Britain as being heavily car dependent. Newcastle has the lowest car share in Britain with only 30%. 4. DISCUSSION 4.1. Does mode shift matter? Does it matter to reduce car usage and transfer people from cars to public transport and bike and walking? The following calculation illustrate the effect of changing from the mode split similar to Milton Keynes to a mode split similar to Freiburg. The calculation is for work journeys only in an imaginary city with a population of about and about workers that travel on average 11 km to work, 250 days every year. The calculation is based on: Average trip length is 11km. The number of workdays is 250 per year. Average yearly travel distance per worker = 5500 km. Sum total distance work trips per year = km CO2 emissions from one worker in a car per year: 0,150 kg * 5500 = 850 kg. CO2 emissions from one worker in a bus with 15 passengers per year: 0,450 kg * 5500/15 = 165 kg. Table 6. Calculation of CO2 emissions in two imaginary cities with different MS CO2 Emissions per year. Kg Freiburg mode split Sum work trips per year, Freiburg MS. km Model city CO2 emissions with Freiburg MS. Tonnes Milton Keynes mode split Sum work trips per year, Milton Keynes MS. km Model city CO2 emissions with Milton Keynes MS. Tonnes Car PT Bike Walk nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 75

81 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT The imaginary model city with Freiburg MS has yearly CO2 emissions of tonnes, while with Milton Keynes MS the yearly CO2 emissions are tonnes, which is tonnes or 230 per cent more than in the Freiburg MS case. With other words each worker produce CO2 emissions per year respectively 1,56 tonnes with Freiburg MS, and 3,6 tonnes with Milton Keynes MS. There are substantial gains in a transport policy that reduce CO2 emissions in Car Cities and there are a lot of such cities in Europe Public attitude and image, history and culture This paper shows a marked cultural difference between the protestant North Europe and the catholic South. This divide is clearly demonstrated in Belgium where bike use in the Valona south is negligible while in the Flemish North bike use in the cities is very high and comparable to the Netherlands. The arrival of the mountain bike in the eighties suddenly changed the image of the bicycle from an outdated transport mode to an artefact for the hip professional for training as a supplement to work outs in the training studios. In Oslo, Norway a traditional cross-country skiing competition has got its summer variant on bikes Birken, which has become a key status symbol among the finance brokers both to get into the race and complete the race below the set time limit. Thus the modern lifestyle mean cycling for pleasure/competition while the car is indispensable to sustain a fast life. That was in Oslo, Amsterdam and Copenhagen certainly have a different attitude to bicycling. There it still is common to cycle to work for all income groups. In Denmark and Netherlands it also seems that biking is a way of life people in the cities will hang on to? 4.3. Income, car ownership and costs Rising incomes makes car ownership and use relatively cheaper. It is expected that real incomes in Norway will increase by 70 % in thirty years. 2 What will people do with increasing affluence? Past history shows that people use a share of it to buy more cars and use them. People will also travel more with air to major cities abroad, an increasingly unsustainable behaviour. With very fast growth in China, India and many more nations, the number of cars may reach 2 billion cars in a few decades (Dargay 2007, Sperling and Gordon, 2009.). Does this trend leave a place and space for the bicycle? Yes, the evidence in this paper shows that it is possible, especially in cities. There are many examples of cities with both a very wealthy population and a high level of bike use. The cost, travel time door to door (D2D) and the convenience of alternative modes have a crucial impact on modal choice (Pucher et al, 1999). The vast majority of households have a car, many more than one. The car is by far the fastest door to door for most trips in most cities, is also the cheapest travel compared with public transport since most drivers only compare the petrol costs with the out of pocket costs for PT. If time costs are included, the car also beat the bicycle on trips more than three to four kilometres. In addition it offers comfort, flexibility and a feeling of safety. 2 Nasjonal Transportplan Utredningsfasen, Oslo 2011:14. 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 76

82 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT 4.4. Infrastructure, climate and safety Cities with high levels of bicycle use also have a good cycle network, but it is not clear if the infrastructure or the good network caused the amount of cyclists. The infrastructure may be a response to increased cycling instead of its cause. A network of separate bike paths and bike lanes as well as general streets with low speed seems to be necessary, but not sufficient element in a Bike City. Does climate matter? Cycling is obviously affected by climate and topography, but there are questions to be raised. Why is cycling so prevalent in the wet Netherlands, while nearly non-existent in Southern France or Italy? Portland, which is both hilly and rainy, has the highest cycling rates in the USA (Pucher et al, 2009). Trondheim and Umeå both lies at the same latitude (64 degrees North), but they cycle twice as much in Umeå (19%) than in Trondheim (9%). Climate plays a role, but it is not the major barrier to increasing cycling. Cycling is dangerous is often used to explain why many persons are not cycling. It is obviously matters for some, but it is not the whole history. The Dutch culture where cycling is mixed with other traffic seem to show a more accommodating attitude between car drivers and bike users than in most other countries. If there exist a hostile attitude to cyclists on the roads, one may find that the cyclists feel more unsafe, perceive that cycling is dangerous and refrain from using the bike, irrespective of the actual risk involved Planner s vision or people s vision? After the 2 nd World War the car gradually became ubiquitous and a driving force in urban expansion. The Buchanan report Traffic in Towns asked the question: How to cope with the car? The answer was high capacity roads and Traffic Architecture to accommodate the car. Milton Keynes is one of the last New Towns to be planned from these principles. The Buchanan report became the bible for generations of transportation engineers and planners, and car modernization their leitbild. Rådberg 3 states that to understand urban change one has to understand the planners inner thoughts, ideas and visions. In the fifties and sixties, some 25 years behind USA in car ownership, the vision of the car society swept through the planners minds in Europe. The car would solve the transport problems and there was no need for trams, bikes or buses in the vision of the future. The bike disappeared as a transport mode in the planner s minds and in practise, except in Holland, Germany and to a certain extent in Denmark. Milton Keynes, one of the English New Towns is a prime example of how the planners car vision was implemented. 75 per cent of all trips in Milton Keynes are in a car. However, in many of the successful bike cities the planners car modernisation ideas met opposition from the public. The lesson to be learnt from the successful bike cities is that cities should not be planned by a small group of planners and politicians, but be developed in a dialog with the public. 3 Rådberg, Johan, 1997: Drømmen om Atlantångaren. Utopier og myter i 1900-tallets stadsbyggande. Atlantis, Stockholm 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 77

83 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT 4.6. More freedom to cities less government The main lesson from the successful bike cities is that at a certain point the car modernisation path was broken and shifted towards a bike path or a public transport path, or an eco-path, often a combination of all these. The City of Davis in California for example, adopted a bike strategy in This was expanded into more environmentally designs for neighbourhoods and by 1980, the City was praised as an Eco-City, visited by celebrities as the wife of the then president Carter and the actress Jane Fonda. Freiburg, which retained the tramway, became famous for developing the environment card a payment system for public transport. In the early nineties the design for the new urban area Vauban, became the example of how to design a sustainable city based on public transport as the main mode and with strong restrictions on cars (Langeland 2009). Another lesson from the successful bike cities is that they managed to make a coordinated action across all levels and layers. Far too often one finds that the governing structure and the incentives for the city planners and politicians work against sustainable transport in cities. The car infrastructure is expanded and the cost of using the car is going down, while the public transport suffers from decreasing quality and increased fares, and the cyclists are literally forced off the roads. This has happened with the active support of the local politicians. The consequences of a fragmented and sector organised transport policy, is that the local politicians have strong incentives to acquire state grants for road building and disincentives to promote cycling. The German system of an earmarked tax to be used for city transport and the French tax on businesses according to the number of employees, are exceptions to the common picture. In most Western European countries the nation state strongly steer the city transport system through the development of the infrastructure and the level of subsidies of the fares. The shift away from the car modernisation path seems to have released creativity and innovation in the cities. It was very much a local focus that drove change. This underline that each city is unique and instruments and measures must be adapted to the local context. Most of the public transport systems in urban areas rely on subsidies in one way or another. Hardly any system is in operation where the revenue from the ticket sales makes profit for the transport company. Hence, the public transport system is continually a headache for the politicians both regarding the financing of the infrastructure and the services. Building bicycle networks in cities experience the same barriers. Who shall pay for the infrastructure? For roads the taxation of the car and petrol tax has paid for the roads, and many places toll roads are used to finance the infrastructure. Cyclists are however not a great source to be taxed to pay for the bike net. Establishing a separate bike net never has therefore not been in the forefront of the politicians agenda. The successful bike cities have against the odds, managed to develop the cycling system. That this was possible is an important lesson to be learnt from the successful cities. 5. CONCLUSIONS An intensive bicycle policy will not persist when it does not result in increased cycling, and a high degree of bicycle use will not persist if facilities are not upgraded and 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 78

84 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT maintained. 4 TEMS with its present weaknesses has been used to make lists of best practices. The bicycling country is still the Netherlands, but the best cycling city is Münster in Germany with 38% of all trips on the bike. The German cities Freiburg and Münster are the top environmentally sustainable cities, while the Swiss cities are foremost on integrated transport planning, low car use and high use of public transport. Another lesson from Switzerland is wide public participation and the use of referendums to decide important planning issues. Practice in one country might encourage innovation in another country, it is after all the main purpose of comparison. However, different planning traditions and political, institutional and cultural circumstances might require adaption to different environments. Thus, it might not be possible to transfer the practice from one country to another, as: "the danger of proposing change in practice in the light of experience abroad is that practice may be dependent for its success upon a chain of circumstances which does not apply at home" (Booth, 1986 B). There are still good reasons to compare land use and transport in cities in different countries (Hambleton 1995): 1. Many of the social and economic problems are similar. 2. Policy responses have often been similar, but with striking differences, particular in approaches to city management. Wrong kind of government which means that the instrument used to solve collective problems and meet society s needs is outdated hence the need for reinvention. 3. Cross national comparison represents a form of natural experiment in which the results of applying differing policies to similar problems can be evaluated, and different ways of thinking about problems and policies, and different approaches. 4. Cross national comparison also contributes to system understanding. Through the contrast to other countries one can understand one s own. Areas kept off policy in one, may be very influential in others. The ranking of cities presented in this paper can give insights and inspiration, hence go to Münster, to Amsterdam, to Copenhagen and learn. But one should not only go to the best practice cities, there might be more relevant learning from a city striving to reach its goals and not succeeding, as the successful one. References Bernick, Michael and Cervero, Robert (1996) Transit Villages in the 21 st Century McGraw-Hill Booth, P. (1986) The design and implementation of cross-national research projects - introduction. in: Masser, I and Williams, R (eds) Learning from Other Countries, Geo Books, Norwich Booth, (1986 B), in Fischer, T.B. (2002). Strategic Environmental Assessment in Transport and Land Use Planning. EARTHSCAN Bristol City. Project Cycling City. Pedal powering ahead. June 2010 Cycling England. Cycling City and Towns Programme Overview. March Fietsberaad publication, number 7 2nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 79

85 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Dargay, J. et. al. (2007). Vehicle ownership and Income Growth, Worldwide: Institute for Transportation Studies, University of Leeds, January European Environment Agency, EEA (2008). Climate for transport change. Term 2007: indicators tracking transport and environment in the European Union. EEA Report 1/2008. Copenhagen: European Environment Agency. European Environment Agency, EEA (2006). Urban sprawl in Europe. The ignored challenge. EEA Report 10/2006. Copenhagen: European Environment Agency. European Environment Agency, EEA (2006). Transport and environment: facing a dilemma. Term 2005: indicators tracking transport and environment in the European Union. EEA Report 3/2006. Copenhagen: European Environment Agency. Fischer, T.B. (2002). Strategic Environmental Assessment in Transport and Land Use Planning. EARTHSCAN Flyvbjerg, Bent., Public Planning of mega-projects: overestimation of demand and underestimation of costs in Priemus, H. et al (2008): Decision-Making on Mega-Projects Edward Elgar publishing Foster, Mark S., (1983). The Automobile and the City. 1983: cited in Østby Hambleton, Robin., Thomas, Huw., eds. (1995) Urban Policy Evaluation. Paul Chapman Publishing, 1995 infas (2007). Mobilität in Deutschland 2008 MiD Bonn, 24 July infas (2011). Mobilität in Deutschland (MiD) National TransportationSurveys. Activities in Germany. TRB Annual Meeting 2012, Session 483 P Landeshauptstadt München & MVV. Mobilität in Deutschland (MiD) München und Münchener Umland. Ergebnisbericht MiD Langeland, Anders. (2009). The Quest for Environmentally Sustainable Transport Development. Land use and transport planning in 4 cities in 4 countries. Aalborg University. May,A.D., Page, M., Hull, A., (2008). Developing a set of decision-support tools for sustainable urban transport in the UK. Transport Policy 15 (2008) Norway, Ministry of Transport, Nasjonal Transportplan Utredningsfasen, Oslo 2011:14. Newman, Peter & Jeffery Kenworthy (1996) The land use-transport connection Land Use Policy, Vol. 13, No. 1 pp. 1-22, 1996 Newman, Peter & Jeffery Kenworthy (1999) Sustainability and cities: Overcoming automobile dependence. Island Press. Pharoah, Tim & Apel, Dieter (1996) Transport concepts in European cities. Aldershot, Avebury Pucher, J., Komanoff, C., Schimek, P. (1999) Bicycle renaissance in North America? Recent trends and alternative policies to promote bicycling. Transportation Research Part A, Vol. 33, Nos. 7/8 1999, pp Pucher, J., Dill, J., Handy, S. Infrastructure, programs and policies to increase bicycling: an international review. Preventive Medicine, 2009, Volume 50, pages S106-S125 Pucher, J., Buehler, R., and Seinen, M. (2009). Bicycle renaissance in North America? An update and reappraisal of cycling trends and policies. Transportation Research Part A 45 (2011) nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 80

86 SESSION: SUSTAINABILITY AND SOFT MODES OF TRANSPORT Pucher, J., Komanoff, C., and Shiemek, P. (1999). Bicycle renaissance in North America? Recent trends and policies to promote bicycling. Transportation Research Part A 33 (7/8) (1999) Sachs, Wolfgang., (1992). For the love of the automobile. University of California Press. Sperling, D., and Gordon, D. (2009). Two Billion Cars. Oxford University Press. Stadt Münster, (2008). Verkehrsverhalten und Verkehrsmittelwahl der Münsteraner. Ergebnisse einer Haushaltsbefragung im November Beiträge zur Stadforschung, 1/2008. Thomassen, Øyvind. (1997). Herlege Tider. Norsk fysisk planlegging ca Senter for teknologi og samfunn. NTNU. Trondheim Østby, Per. (1995). Flukten fra Detroit. Senter for teknologi og samfunn. NTNU. Trondheim lstage2bid/3%20european%20precedent%20study%20(separate%20volum e)% pdf TEMS The EPOMM Modal Split Tool, ( BYPAD Bicycle Policy Audit Project was funded by the European Commission STEER Programme europa.eu.int/comm/energy/intelligent/index_en.html Fietsberaad publication, number 7 Bicycle policies of the European principals: continuous and integral s.pdf MON (Mobility Survey Netherlands) oming/mobiliteitsonderzoek_nederland/ Nathanail, E., Measuring the quality of service for passengers on the hellenic railways. Transportation Research Part A: Policy and Practice, 42(1), pp nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 81

87 SESSION: PUBLIC TRANSPORT PLANNING

88 SESSION: PUBLIC TRANSPORT PLANNING The Future Search Methodology as a Tool for the Development of a Sustainable Urban Mobility Plan in the Region of Central Macedonia Afroditi Stamelou, Dr. Kostas Konstantinou, Chrysostomos Makrakis Karachalios, Irene Tsakiridou Anatoliki S.A. Development Agency of Eastern Thessaloniki s Local Authorities, Thermi, Greece Afroditi Stamelou (Corresponding author), afroditis@windowslive.com Abstract The region of Central Macedonia went through a period of rapid industrial and economic growth that shaped today s urban network, dominated by the greater Thessaloniki area and complemented by smaller urban centres. While the former has always been burdened with heavy traffic, in the last years all the urban centres in the region also experienced traffic congestions and parking problems. A sure way to deal with a lot of serious mobility issues is the creation of a Sustainable Urban Mobility Plan (SUMP). An Intelligent Energy Europe funded project, Poly-SUMP ( 2013, access 20/03/2014), aims to assist polycentric regions in order to implement their own SUMP. The development of this type of SUMP is based on the Future Search Methodology. Within the Poly-SUMP project the region of Central Macedonia organized a Local Future Search Workshop ( access 15/032014) related to sustainable mobility planning in the region. This paper describes and analyses the Local Future Search Workshop experience in polycentric urban mobility planning in the region of Central Macedonia. The list of target areas to ensure sustainable mobility in the region is presented. The priority actions that are appropriate for the region are mainly infrastructures (rail network expansion, bicycle and pedestrian network development and low traffic areas development), as well as the behavioural modification and awareness raising activities and the uniform strategic planning at regional level. A specific questionnaire about the process evaluation, comments and suggestions was distributed to the participants at the end of the workshop. The results of the questionnaire are also referred to in this paper. Keywords: Future Search Workshop, participatory approach, spider diagram, Polycentric Urban Mobility Plan, diffuse city regions. 1. INTRODUCTION The Central Macedonia Region went through a period of fast economic and industrial growth. This situation shaped today s urban network, which consists of the greater Thessaloniki area network and of the smaller surrounding urban centres networks. Thessaloniki, as well as all the urban centres in the region experience traffic congestions and parking problems. Lately, even though the high petrol prices have reduced excessive private car use, still the lack of an integrated transport policy makes the use of public transport difficult. Traffic jams, causing increased fuel consumption, air, noise and visual pollution, an increase in road accidents and time loss, also affect human health and degrade life quality. Figure 1 illustrates the region of Central Macedonia. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 82

89 SESSION: PUBLIC TRANSPORT PLANNING Figure 1. Map of the Region of Central Macedonia A sure way to deal with these mobility issues is the creation of a Sustainable Urban Mobility Plan (SUMP). This is a strategic plan designed to satisfy the mobility needs in cities and their surroundings (Rupprecht Consult, 2013). The methodology of SUMPs concentrates on a single urban area and it is more complex when operations, people and transport are scattered in different towns of polycentric regions ( diffuse city regions). A specific diagram called spider diagram is created in order to illustrate and analyze the polycentric profile of a region. For the creation of this diagram, specific indicators, considered for a comparative assessment towards polycentricity, were calculated. The most important of them are the three types of Gini coefficient, the average distance to workplace and to recreation and the modal split share of public transport and of nonmotorized modes. "Gini" coefficient is a well-established measure of statistical dispersion, firstly developed by the Italian statistician Corrado Gini in 1912 (ANATOLIKI S.A. 2, 2013). The value 1 seems that the region is near to the polycentricity and the value 0 to the monocentricity. Figure 2 demonstrates the spider diagram for the Central Macedonia Region. Figure 2. The spider diagram for the Region of Central Macedonia Greece 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 83

90 SESSION: PUBLIC TRANSPORT PLANNING 2. THE POLY-SUMP PROJECT Poly-SUMP is a European - Intelligent Energy Europe project. It started in April 2012 and runs until October The project is coordinated by Regione Marche (Italy) and is undertaken by a consortium of regional authorities, experienced consultants and research institutes in urban transport planning across Europe. The consortium of the project consists of eleven (11) partners, namely Regione Marche, Pluservice and ISIS - Institute of Studies for the Integration of Systems from Italy, ICLEI - Local Governments for Sustainability from Germany, Panteia NEA from Netherlands, Trivector from Sweden, CIMAC - Intermunicipal Community of Central Alentejo from Portugal, Development Centre of the Heart of Slovenia, BOKU - University of Natural Resources and Applied Life Sciences from Austria, Missions Publiques from France and ANATOLIKI S.A./REACM-Development Agency of Eastern Thessaloniki's Local Authorities from Greece ( access 20/03/2014). Poly-SUMP assists partner regions with developing an integrated Poly-SUMP methodology. Diffuse cities need to introduce polycentric sustainable urban mobility plans through a holistic and multi-governance approach. The overall goal of the project is to reduce energy consumption from transport in diffuse city regions through the implementation of SUMPs and to provide knowledge from this development to help other diffuse regions in their development of a Poly-SUMP. The Poly-SUMP methodology (based on application of Future Search workshops) supports the preparation of a SUMP in polycentric regions (Poly-SUMP) in connection with the existing/future SUMPs of cities within the poly-centric region. The outcomes of the project will be a detailed set of guidelines for developing and implementing the Poly- SUMP planning process. ( access 20/03/2014). Rather than relying wholly on conventional planning methodologies, the project consortium will use an innovative but increasingly well-known participatory technique, Future Search, in developing a polycentric sustainable mobility plan. This new methodology will be tested in developing Poly-SUMPs in six participating regions: These include: Marche (IT), Central Alentejo (PT), Central Macedonia (GR), Rhine Alp (AT), Heart of Slovenia (SI) and Parkstad Limburg (NL). 3. FUTURE SEARCH WORKSHOP METHODOLOGY Historically, The Future Search Workshop (FSW) originates from two different sources. The first one was the German Zukunftswerkstatt ( Workshop of the Future ), which was created and engineered at the beginning of the 80 s in order to allow citizens to participate in urban planning. The aim of this procedure was to achieve democratization from below from such processes. The second source is the North- American Future Search Conference, which was developed at the beginning of the 90 s aiming at the search of a common ground on which to build a better future. In the beginning of the new century, the two models merged into the Future Search Workshop methodology (Poly-SUMP project, 2012). A Future Search Workshop typically involves 60 to 100 people who share a common purpose and vision. The process comprises five specific type of activities spread over 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 84

91 SESSION: PUBLIC TRANSPORT PLANNING three days. During these five activities, participants review the past, explore the present, create desired future scenarios, discover common ground and make action plans (Weisbord and Janoff, 2010). Mixed groups work on the past, present situation and future actions. These are subgroups of the whole group of participating stakeholders, whose members share a common perspective on the present situation. It is important that every one of the participants actively takes part by expressing its views and suggestions. Every session of the workshop concludes with a whole-group discussion. A Future Search Workshop is typically articulated around three stages: Stage 1: Critical diagnosis, where participants analyse the past and the current situation of mobility in their region. Stage 2: Imagination and common ground, where participants have the opportunity to develop visions of their desired world and finally Stage 3: Building an action plan, where participants design actions based on the visions which they previously have developed. Future Search enables a person to learn more from the group than by individual effort. The main success of a Future Search Workshop is due to its gathering the whole system in a room. The key word is shared. It is very important to have participants from different key actors, who take part in the decision making. Moreover, the common ground and the future vision should be in the centre of the discussions and possible conflicts and problems must be maintained on the margin. The exploration of the whole context before pursuing to fix any separate part is also an important issue: Think globally, act locally. Finally, encouraging individuals to take responsibility and initiative for actions before, during, and after the Future Search is a significant aspect of the success of the Future Search approach. The region of Central Macedonia organized a Local Future Search Workshop (LFSW) ( access 15/03/2014) related to sustainable mobility planning in the region. Stakeholders from a variety of organizations were invited in order to diagnose the past and current situation of mobility in the region and agree on a common vision and an action plan. The aim of this action plan is to promote sustainability and improve the daily movements of people and the connections between the cities in the region, enhancing public transport, reducing fuel emissions and upgrading the environment and the quality of life in the region. 4. PREPARATION OF THE LOCAL FUTURE SEARCH WORKSHOP IN THE REGION OF CENTRAL MACEDONIA As part of the Poly-SUMP project ANATOLIKI S.A, (Regional Energy Agency of Central Macedonia-REACM) invited stakeholders from a variety of organisations to diagnose the past and current situation of mobility in the Region of Central Macedonia. In addition to this, stakeholders have been invited in order to agree on a vision and an action plan to guide all the parties involved, which are responsible for developing coordinated mobility options in the Region of Central Macedonia. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 85

92 SESSION: PUBLIC TRANSPORT PLANNING 4.1. Location of the workshop The Local Future Search Workshop in Central Macedonia Region was held at Agios Nikolaos, a town approximately 75 Km west of the city of Thessaloniki, from 2 nd to 3 rd of October 2013.It is important to note that the intensive interaction typical of Future Search events requires both a strict organization of the timing and a suitable location. Initially, the place of the workshop should be well connected by public transport with the main city centres, should offer several rooms where participants can work or relax during coffee breaks. It is important for the rooms to be big enough to host all participants and to avoid overblown noise during the working groups. For the workshop in Naousa small compromise was reached. Unfortunately Naousa city centre doesn t offer a hotel or a conference hall with the required characteristics to be able to host this particular kind of event (ANATOLIKI S.A. 3, 2013). For this reason, ANATOLIKI S.A. decided to select a hotel only few kilometres from the city centre of the city of Naousa, in Agios Nikolaos Park. It s a very well situated hotel near the river, which fully accomplished the workshop needs Number and type of the workshop s participants The workshop was organized for the first time in Greece with the Future Search methodology, based on the interaction of relevant actors. The workshop was based in sharing common vision and ways of implementation, which would be quite difficult to achieve by each individual entity, but could become real through the cooperation of all. Participants were invited from several stakeholders in the Region of Central Macedonia. Initially, a larger number of stakeholders had agreed to attend the workshop, but due to work commitments the number of participants eventually decreased. The final representatives were from several Municipalities and Development Agencies of the region. Furthermore, there were representatives from the Hellenic Institute of Transport, the Hellenic Institute of Transportation Engineers, the National Eurovelo Coordination Centre for Greece, the Ecological Movement of Thessaloniki, Thessaloniki s Urban Transport Authority, the Organization of Urban Transportation of Thessaloniki, the Organization of Planning and Environmental Protection of Thessaloniki, the Association of Spatial Planning and Regional Development, the Association of Hindered People "Disability Today", the Greek International Business Association, the Association for the rights of pedestrians, the Designers Society of Public Works of Central Macedonia, the Technical Chamber of Greece, the Alexander Technological Educational Institute of Thessaloniki, the Laboratory of Transportation Engineering, the Aristotle University of Thessaloniki Department of Civil Engineering and in addition, from the Local Development of ANATOLIKI S.A. (ANATOLIKI S.A ). The final workshop s participants were 37 from 25 different organizations/stakeholders Duration of the workshop As described before, the Local Future Search Workshop in the Region of Central Macedonia lasted two days. This decision was taken due to the refusal of most of the stakeholders to participate for three days in the workshop. During the first day, the participants shared their personal past experiences on mobility and continued 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 86

93 SESSION: PUBLIC TRANSPORT PLANNING discussing the current state of mobility in their Region. Subsequently, the recording of the trends that will affect mobility in the next years took place. The day ended with the creative participation of invitees to plan toward the desired future for the mobility at a regional level. The second day of the workshop, focused in the establishment of guidelines for actions to improve the transport network and to promote sustainable mobility in the Region of Central Macedonia. Participants recorded the most important actions that should be made over the next years and suggested possible ways to achieve these objectives. 5. ORGANIZATION OF THE LOCAL FUTURE SEARCH WORKSHOP IN THE REGION OF CENTRAL MACEDONIA 5.1. Schedule and procedure of the workshop implementation The two days workshop was structured in eleven sessions. The first six sessions were carried out in the first day and the remaining five in the second day. Three moderators from ANATOLIKI S.A. had the main responsibility for the workshop s success. The first moderator presented the methodology, made the rules and expectations clear before each session and highlighted the main results. The other two moderators provided further explanations whenever needed, kept the time, summarized the results and assisted the attendants. In the room there were six tables with numbers and colourful pens and post-its on them. Figure 3 shows the configuration of the room. At the beginning of the workshop participants chose a table number and sat at the corresponding table. Figure 3. Configuration of the room In the most of the sessions, participants answered individually the corresponding questions in colourful post-its. Subsequently, they shared and aggregated their opinion at their table. Finally, one member of each table presented the table s answers and put the post-its on the wall. In this way it was easy to find the similarities and the differences between participant s subgroups. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 87

94 SESSION: PUBLIC TRANSPORT PLANNING 5.2. Sessions of the workshop - Five main activities Review the past The first session included an introduction about the LFSW and the presentation of the participants in order to know each other better and break the ice between them. During the second session participants were invited to look back on mobility in their personal life, as well as in their Region, Europe and world. The purpose of this session was to have participants share their experiences and to analyse the similarities and the differences in their assessment of mobility in the past Explore the present In the third session participants were divided into institutional and interest subgroups. Only in this session participants were divided into these groups. In all other sessions, participants selected randomly their table. The groups were asked to assess the current situation of mobility and to provide good and bad practices from their region. All groups were proud for the development of the suburban railway and the international airport of Thessaloniki. The improvement of the level of service of urban public transport in the city of Thessaloniki is very important for the participants and they are very proud of this. On the other side, almost all of the participants groups reported that they are not satisfied for the lack of law enforcement in the streets. Furthermore, most of them signalized that there is lack of education on mobility issues and very high resident s dependence on their private cars. The significant issue, in this session, was that some groups classified a practice in the category proud of and some in the category regret. This fact exists due to the different approaches of each institutional and interest subgroup (ANATOLIKI S.A ). The fourth session referred to the future trends affecting the mobility in the region. During this session participants were asked to identify the main trends that could impact mobility in the next 30 to 50 years. They used colour stickers for these trends. In regional and national level, participants assessed positively the improvement of public transport systems, the alternative means of transport and the improvement of services for the disabled people. On the other hand, they assessed not so positively or negatively the lack of funding and the economic crisis. In European and global level they assessed positively the ITS systems and the development of infrastructures for rail freight and not so positively or negatively the climate change and the population movementrefugees Create desired future scenarios In the fifth session, participants were motivated to paint or to describe the perfect future scenario concerning mobility in their region. They virtually transferred to the year 2050, when the region won the Prize for being the world s most sustainable mobility region. All groups envisioned a more environmentally friendly region, which provides better quality of life for citizens. Moreover, the development of rail transport, which will connect the major cities of the region, was reported from all groups. Furthermore, 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 88

95 SESSION: PUBLIC TRANSPORT PLANNING people visionalized the use of cleaner energy sources (RES) in order to provide the required quantity of energy. On the other side, the major disagreement between the group s visions was the use or not of the private car. Some groups excluded completely private cars from the city centres, while others preferred them. They promoted more environmentally friendly vehicles, powered by alternative energy sources, such as electric vehicles, biomass, and solar energy. The other important difference between the groups was the prohibition or not of car use and residence in the city centres (ANATOLIKI S.A ) Discover a common ground In the sixth session, the final session for the first day, the participants were asked to discuss about their common ground. They were asked to find the values, the goals and the milestone activities which they have in common as a group and which they find important for 89ealizing the future they wish for (future scenarios). The moderation team wrote all the values, goals and milestones on a paper on the wall and gave orange stickers to the participants in order to vote for the most important of them (ANATOLIKI S.A ). The most important values for the participants were the quality of life and the sustainability and the most important goals were the reduction in private car s use and the opportunity of multimodal transportation. They, also, found the awareness raising and the mentality changing as significant goals. Finally, as the most important milestones, participants reported the education, the improvement of urban planning and of infrastructures of sustainable mobility, the public transport network expansion and the routes densification. As was expected, milestones were more than values and goals, because there are a lot of milestones in an effort to achieve specific goals Make action plans In the second day, the last five (5) sessions took place. In the seventh session, participants were asked to generate actions that could be started at regional level in order to open the path to the desire future.participants referred a large number of actions therefore these actions were grouped under common themes. The main thematic groups were: Policies, Funding models, Improvement of transport infrastructure, Behavioral modification and Public Transport. In this session, it was difficult for the participants to understand the way they should express the actions that need to be implemented in the region. Initially, some reported actions were quite general and it would be difficult to be implemented in the future. The moderation team was around the tables in order to support and guided the participants in the way to transform their ideas into feasible actions (ANATOLIKI S.A ). In the eighth session, the participants were divided into teams in order to work in the actions resulted from the previous session. They were asked to choose actions that they wanted to implement themselves. The total number of actions, after session seven, was twenty (20). Participants voted for the actions they would like to implement. Since there were actions in which no one of the participated stakeholders was interest to implement. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 89

96 SESSION: PUBLIC TRANSPORT PLANNING Therefore, the final number of the actions proposed for the Central Macedonia region was twelve (12) (ANATOLIKI S.A. 2, 2013). During the ninth session, participants worked in the action that they had chosen before. For each action, the work team should answer some specified questions. The main ones were: What is the title of the action? What is the goal? What is going to happen? Where does the money come from? How must does it cost? etc. Nr Table 1. Central Macedonia Region s Action Plan developed in the LFSW Action description Behavioral modification and awareness raising activities Rail network expansion and routes densification Uniform strategic planning at regional level Priority (based on the scores given) Votes after the presentation of the actions Nr Action description Bicycle network development Pedestrian network development Financing instruments for infrastructure development Priority (based on the scores given) Votes after the presentatio n of the actions Freight network improvement (city-logistics) Low-traffic areas development Establishment of an integrated regional transport authority to provide leadership in the co-ordination, planning and development of an integrated, multi-modal transportation network Bus services improvement nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 90

97 SESSION: PUBLIC TRANSPORT PLANNING 6 Updated Traffic studies and/or development of new ones Improvement of public transport services and networks for disabled people 3 4 After that, in the tenth session each team presented the corresponding action. After the presentation of all actions, orange stickers were given to the participants to vote for the action according to their importance. The results are presented in table 1 (column Votes after the presentation of the actions ). The actions in more details are presented in the D of the Poly-SUMP project, as well as in the website of the project ( The final session, No eleventh, included the conclusions and the evaluation of the workshop. The results of this evaluation are analysed in the next chapter. 6. EVALUATION OF THE PROCEDURE AT THE END OF THE WORKSHOP At the end of the Local Future Search Workshop in the Region of Central Macedonia, an evaluation questionnaire on the workshop was distributed to the participants. This questionnaire was prepared by ISIS (leader of WP5, In Built - Evaluation). The aim of this was the assessment of the organisation and the content of the workshop, as well as suggestions and recommendations about the next workshops with this methodology. Participants were asked about their expectations from the workshop, as well as their previous knowledge regarding the Future Search methodology and the SUMP. Subsequently, they evaluated the workshop sessions, the quality of the moderation and the discussions during the workshop, the place and the logistics. Participants, also, recommend topics and future improvements for the next FSW s in the Region. Most of the participants attended the first Local Future Search Workshop in Central Macedonia Region with the aim to share experiences with people interested in sustainable mobility. Moreover, they were interested in designing sustainable mobility in their region through the cooperation of relevant stakeholders and experts on mobility issues (ANATOLIKI S.A ). As we can see in Fig. 4, most of the participants declared to have no knowledge about the methodology or to have heard a little about it, but not something more in depth. The 25% of the participants had a pretty good knowledge about the methodology and the 13% of them has learned about it in the preparation of the meeting. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 91

98 SESSION: PUBLIC TRANSPORT PLANNING Figure 4. Assessment of the level of knowledge on the Future Search methodology before the LFSW Moreover, the vast majority of the participants gained knowledge during the workshop. Some of them learned a great deal and some of them learned a little. All of the participants expressed a positive view about the overall satisfaction of the workshop. Fig.5 shows the satisfaction of the participants regarding each of the workshop sessions. Summing up, all sessions were awarded a score greater than or equal to 4, with an average value of 4,2 out of 5,0. The session more appreciated was a shared diagnostic - Future trends (day 1) with an average vote of 4,6. Participants expressed positive view about the innovative Future Search methodology. Moreover, they were also impressed by the exchange of views with representatives of different stakeholders. They liked the creative participation through the interactive process of the future search methodology and the share of experiences. Similarly, they enjoyed the interactive way of the workshop and the fact that all of the participants could express their own view. Figure 5. Overview of each Future Search Workshop sessions satisfaction 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 92

99 SESSION: PUBLIC TRANSPORT PLANNING On the other hand, participants didn t like the constant pressure of time in order to complete each session. Some of them suggested that the duration of the workshop could be three days. The participants from smaller cities expressed dissatisfaction about the dominance of the priorities of the metropolitan area of Thessaloniki in the discussions. Subsequently, participants were asked what they would have done differently if they had been involved in the workshop organization. A lot of them suggested the workshop to last more days, because they got tired from the whole demanding process. In addition to this, they desired to have more representatives from municipalities outside of the metropolitan area of Thessaloniki. In order to achieve this, they suggested operating a parallel online workshop with citizens from the municipalities, which were not able to attend the workshop. Finally, participants were asked to leave comments regarding recommendations for future editions of the Local Future Search to deal with new sustainable mobility or other planning topics in their region. They suggested organizing future search workshop both in fields related to the transport sector or not. The recommendation about the transport sector was the organization of a Future Search workshop on freight or public transport. Moreover, the participants recommended fields such as waste management, rural development, environmental protection, renewable energy sources and energy saving (ANATOLIKI S.A ). Another questionnaire regarding the outcomes (actions) of the LFSW in the Region of Central Macedonia was sent to the participants via . This procedure is on-going, therefore the results cannot yet be presented. 7. CONCLUSIONS The development of a SUMP and its subsequent application greatly improves the mobility of the citizens, in terms of their accessibility to home, work or leisure. The aim of the LFSW was to formulate specific actions with the purpose of developing a SUMP for the Region of Central Macedonia. At the end of the workshop, a list consists of twelve (12) of priorities actions was formulated. The main ones are behavioral modification and awareness raising activities, the rail network expansion and routes densification, as well as the development of low traffic areas, pedestrian zones and bicycle lanes. In order for these actions to take effect, it is necessary that expert stakeholders become involved in a common place so that all sides participate and differences are resolved. Following the first evaluation questionnaire circulated at the workshop, a second questionnaire related to the workshop s outcomes has already been sent to the LFSW s participants. Through this questionnaire, additional information related to the most appropriate key actors who could potentially play a role in designing and implementing the actions will be given. After that, meetings with these key actors will follow for further development of the Action Plan. Stakeholders will consider in which of the actions they can contribute and what should be the next steps towards sustainable mobility in the region. The resources for financing these actions, as well as the main risks and barriers will be discussed in these meetings in order to be reach specific objectives. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 93

100 SESSION: PUBLIC TRANSPORT PLANNING The transferability of the Poly-SUMP methodology in other regions is another important element. Each of the Poly-SUMP countries selected a twinning region with the aim to transfer this methodology. The Poly-SUMP approach compared to the traditional way of making a SUMP is a bottom-up approach and creates a better understanding of regional transport needs. Moreover, it is very time efficient and fosters the creativity of the participants, while it addresses the most relevant key issues. Close coordination among all key actors that will have a role in developing and implementing of the actions is required. It is expected to engage all pertinent stakeholders and make an integrated approach with the purpose of implementing Poly- SUMP in the Region of Central Macedonia in the next months. References ANATOLIKI S.A. 1 Development Agency of Eastern Thessaloniki s Local Authorities/REACM (2013) D.2.4 Comparative assessment - Central Macedonia, Greece ANATOLIKI S.A. 2 Development Agency of Eastern Thessaloniki s Local Authorities/REACM (2013) D Local Future Search Workshop Report, Greece ANATOLIKI S.A. 3 Development Agency of Eastern Thessaloniki s Local Authorities/REACM (2013) D.5.2 Evaluation of the Local Future Search Workshop Process, Greece Poly-SUMP project (2012) D3.1 Report of the 1 st Future Search workshop, France Rupprecht Consult-Forschung und Beratung GmbH: Sebastian Bόhrmann, Frank Wefering, Siegfried Rupprecht, Susanne Böhler-Baedeker (2014) Guidelines. Developing and Implementing a Sustainable Urban Mobility Plan, Germany Weisbord.M and Janoff S. (2010) Future Search-Getting the whole system in the room for vision, commitment and action, Third edition, Berett-Koehler Publishers, Inc., San Francisco Last accessed 15/03/ Last accessed 20/03/ nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 94

101 SESSION: PUBLIC TRANSPORT PLANNING Integrating Urban Vehicle Characteristics into Sustainable Transportation Planning Lambros K. Mitropoulos 1, Panos D. Prevedouros 2 1 Corresponding Author: Hellenic Institute of Transport, 52 Aigialeias Street, Marousi, Greece, Tel , , Fax , lmit@certh.gr. 2 University of Hawaii at Manoa, Department of Civil and Environmental Engineering, 2540 Dole Street, Honolulu, HI 96822, pdp@hawaii.edu. Abstract Environmental assessments are on the critical path for the development of land, infrastructure and transportation systems. These assessments are based on planning methods which, in turn, are subject to continuous enhancement. The substantial impacts of transportation on environment, society and economy strongly urge the incorporation of sustainability into transportation planning. The increase in the number of hybrid electric vehicles and car-sharing users in the U.S. in the past decade urge the incorporation of vehicle characteristics in traditional transportation planning and sustainability assessment. In this study a sustainability framework is developed that enables assessment of transportation vehicle characteristics. Identified indicators are grouped in five sustainability dimensions (Environment, Technology, Energy, Economy and Users). Our methodology joins life cycle impacts and a set of quantified indicators to assess the sustainability performance of six highway vehicles: an internal combustion engine (ICEV), a hybrid electric vehicle (HEV), a car-sharing program with ICEV, a car-sharing program with HEV, a diesel bus, and a hybrid diesel electric bus. The three travel combinations were developed into scenarios: (a) passenger vehicle only, (b) passenger vehicle and public bus, and (c) car-sharing and public bus. Results show that having car-sharing in the travel mix is the best alternative. The most sustainable vehicle relative to the other five is found to be car-sharing with HEV. The superior performance of car-sharing over transit bus reveals the potential of new policies towards sustainable transportation. Keywords: (B) Transportation; Sustainability assessment; Car sharing; Public transit; Indicators. 1. INTRODUCTION Interest in sustainable development and sustainable transportation has grown rapidly in the first ten years of the millennium due to the environmental, social and economic impacts that these sectors have on communities, regions and Earth. Transportation activities accounted for 33% of CO2 emissions from fossil fuel combustion in 2009 (EPA 2011). On the other hand, improvements in vehicle efficiency and changes in vehicle travel have likely contributed to some pollutant reduction: An 11% decrease of CO2 emissions was observed between 2004 and 2010 (ORNL 2012). Hybrid electric vehicle (HEV) sales in the U.S. increased by 30% between 2005 and 2010 (DOE 2013). Between 2000 and 2012 there was an exponential increase in carsharing (CS) from a few hundred to about 800,000 (Shaheen 2013). CS is a service that 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 95

102 SESSION: PUBLIC TRANSPORT PLANNING provides members access to a fleet of vehicles with low cost based on time or distance. CS enables a more efficient use of vehicles and has resulted in decrease of car use and traveled distance. In North America, studies have shown that users that have joined CS programs have reduced the vehicle miles traveled by their members by an average 44% (Shaheen et al. 2006). When benefits of CS programs are aggregated there is a reduction in pollution, energy consumption, congestion, and parking space requirements. CS results vary, for example, a CS vehicle reduces the need for 4 to 10 privately owned vehicles in Europe and 6 to 23 privately owned vehicles in N. America (Shaheen 2013). Well-established CS programs can be found in Canada, Germany, Singapore, Sweden, Switzerland, United Kingdom and the U.S. CS fleets include gasoline, HEVs and plug-in electric vehicles (Shaheen et al. 2006). Car-sharing systems combined with highly efficient cars are proposed as a potential pathway for future transportation in European Union. The FUture prospects in Transport evolution and for the competitiveness of Europe Project conducted under the FP7 suggests that transportation priorities should focus on integrated solutions of public transportation, car-sharing and cycling, and such solutions will support policy making and sustainable transportation systems (FUTRE 2014). There is a small but increasing body of knowledge relating to alternative fuel and propulsion vehicles and their effects in transportation sustainability are summarized in our past work (Mitropoulos and Prevedouros 2013). However, there is no sustainability study of CS in the literature. Attempts to incorporate sustainability into transportation planning have resulted in the development of indicators representing elements of sustainability (Jeon et al. 2008; Maoh and Kanaroglou 2009; Zietman et al. 2003). Transportation indicators that measure impacts on mobility and environmental effects are applied mainly to the operational stage of the transportation system. Past assessments of transportation sustainability considered only personal vehicles or all modes present on a section of a network. The vehicles considered in those studies were assumed to be gasoline or diesel and to be propelled by an internal combustion engine (ICE). This assumption was valid ten years ago when the share of HEV and CS users in the US was minimal, but approximately since year 2010 these vehicles and modes require explicit accounting in methods and models. Additionally, the aggregation of transportation performance measures restricts one of sustainability s primary roles in transportation planning, which is to assist agencies in evaluating new and proposed transportation modes. The sustainability framework used in this study sets the direction for an assessment in transportation that focuses specifically on vehicles, which together with infrastructure are the two key components of transportation modes. Six vehicle types and propulsion options examined herein include an internal combustion engine vehicle (ICEV) and a hybrid electric vehicle (HEV) in three travel modes: private car, car-sharing and public bus. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 96

103 SESSION: PUBLIC TRANSPORT PLANNING 2. SUSTAINABILITY DIMENSIONS AND INDICATORS The sustainability of a transportation system can be assessed if the parameters that compose it can be defined and measured. To a large extent, available data determine the parameters to be used in a sustainability assessment. A framework is necessary to define what to measure, what to expect from measurement and what indicators to use (Pintér et al. 2005). There is no universally accepted framework for assessing sustainability. Usually, a sustainability framework is divided in different dimensions of sustainability and a set of indicators is defined for each dimension. Several studies in the literature utilize three sustainability dimensions, Environment, Society and Economy, for developing indicators to assess transportation systems (Zietman et al. 2003). Maoh and Kanaroglou (2009) developed a tool as an add-on module in an integrated model for assessing urban sustainability. Their indicators reflected aspects of Environment, Society and Economy. Renne (2009) evaluated Transit Oriented Development sustainability by using indicators based on six categories. He argued that since it is difficult to categorize indicators using the three basic categories of sustainable development (environment, society and economy) and many indicators cross boundaries, six different categories had to be selected. Travel behavior, local economy, natural environment, built environment, social environment and policy context. Jeon et al. (2008) developed indicators for sustainable transportation assessment and grouped them under transportation system performance, environment, society, and economy. Past studies on CS evaluated changes in ownership (Stasko et al. 2012), parking demand (Martin and Shaheen 2011) and decrease in GHG emissions as a result of change in travel behavior among CS users (Shaheen et al. 2006; Martin and Shaheen 2011). CS has also been approached as an example of how life cycle analysis is applied in sustainable consumption patterns (Briceno et al. 2005). Bevan et al. (2008) grouped projects into five major objectives which are related to (i) energy reduction, (ii) materials resource reduction, (iii) environmental impacts reduction, (iv) urban communities support, and (v) sustainability support during implementation at the local level. Building on this background, our sustainability framework consists of five dimensions as follows. 3. METHODOLOGY 3.1. Sustainability framework The sustainability framework consists of five dimensions that are captured by the proposed goals governing transportation systems: Environment, Technology, Energy, Economy, and Users. The goals of the framework are to: Minimize environmental impact. Minimize energy consumption. Maximize and support a vibrant economy. Maximize user and community satisfaction. Maximize technology performance to help a community meet its needs. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 97

104 SESSION: PUBLIC TRANSPORT PLANNING The five dimensions are described below. 1. Environment. The environmental impact of transportation vehicles begins when raw materials to manufacture components are extracted and ends when the vehicle is disposed or recycled. Lighter materials, fuel efficiency and alternative fuels are three primary areas among several that vehicle manufacturers pursue in order to minimize overall environmental impact of vehicle manufacture and usage. 2. Technology Performance refers to the performance of all components of transportation systems made by humans to meet their needs. Weight reduction, high strength materials, engine and aerodynamic enhancements, and alternative propulsion systems are some of the technologies that contribute to sustainable transportation. Performance measures in this sustainability dimension ensure that improvements in design and technology are reflected in the appraisal methodology and capture the full range of sustainable transportation concerns (i.e., balance mobility needs while meeting long term environmental, social and economic goals). 3. Energy is a major component of transportation and is directly connected with the environment and economy. Energy availability, demand, price and actual consumption have short and long term impacts. Consumption of non-renewable energy sources generates emissions that are harmful to humans, and deprives energy from future generations. Globally, vast amounts of energy are needed for transportation infrastructure development, vehicle manufacture and transportation operations. 4. Economy. The creation of a sustainable economy requires the disciplined use of energy and technology. An unsustainable economy results in destruction of environment, has a multitude of social impacts, and results in suboptimal transportation services. In this context, a sustainable economy facilitates a lower cost for urban mobility by assessing vehicle costs including purchase, registration, insurance, operation, parking and fuel costs, and promoting vehicle types and technologies that minimize total cost. 5. Users represent a large set of stakeholders including individuals. User perceptions and preferences vary, hence vehicle and mode choices also vary. Vehicles with performance deficiencies are less attractive to users and become unsustainable in the long term. The Technology Performance and Users dimensions are used in this sustainability framework instead of the traditional sustainability dimension of Society. The Technology Performance dimension takes into account the capabilities and limitations of technology in transportation explicitly. The Users dimension takes into account the preferences and restrictions of system users and other transportation stakeholders explicitly. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 98

105 SESSION: PUBLIC TRANSPORT PLANNING Table 1. Sustainable Transportation Indicators Goal Indicator Ref. Indicator Description and Quantifying Conditions Economy a Energy Technology Performance Environment Minimize environmental impact Maximize technology performance to help people meet their needs Minimize energy consumption Maximize and suppose a vibrant economy Carbon dioxide (CO 2 ), Methane (CH 4 ),Nitrous oxide(n 2 O), GHG, Volatile organic compound (VOC), Carbon monoxide (CO), Nitrogen oxides(no x ), Particle matter (PM 10 ), Sulphur oxides (SO x ) Noise Fuel frequency Maintenance frequency Space occupied (Chester et al. 2008; Jeon et al. 2008) (Dobransky te Niskota et al. 2007) (Anielski 2001) (Anielski 2001) (Cambridge Systematics 2009) Engine power (EEA 2002) Manufacturing energy Fueling energy Operation energy Maintenance energy Manufacturing cost Operation cost Maintenance cost Parking cost (Chester et al. 2008) (Jeon et al. 2008) (Jeon et al. 2008) (Jeon et al. 2008) (Institute of Transportat ion Engineers 1999). Emissions are an outcome of all life cycle stages of a vehicle including manufacture, fueling, operation, maintenance and disposal. Environment indicators were quantified as described in section 3.3. Noise is representative of average urban speeds at a distance of 50 ft. At speeds greater than 30 mph vehicles with advanced propulsion offer negligible noise benefits. Noise levels were used from the literature (Cowan 1994). Time required to fuel a vehicle; significant for short ranging modes. It was estimated by dividing the lifetime miles of a vehicle by the product of fuel tank capacity and fuel efficiency (EPA 2006). ICEV is the base vehicle. It is required to be maintained 22 times in its lifetime; its owner spends two hours each time to drop off and retrieve the vehicle. Maintenance intervals for HEV were estimated to 20 per lifetime based on its mechanical components (Edmunds 2011). For the CS it is estimated that repair and inspection occur 9 and 7 times per year for the ICEV and the HEV, respectively. For transit buses it was assumed that each one requires an average of 260 hours per year for maintenance (Chandler and Walkowicz 2006). Its value reveals that space occupation per passenger can be decreased for different vehicle types or programs. When not in use space is a fundamental requirement effected by parking stalls, garages, depots, etc. It refers to the physical characteristics of a vehicle. Maximization of vehicle power It is estimated as the ratio of torque to vehicle weight. This is a significant technology indicator as it shows how technological advances reduce the vehicle weight (i.e., material quantities and types) and maintain or increase vehicle power that improves vehicle performance. Energy is required for all vehicle life cycle stages; as described for the indicator emissions Represents the invoice price of a vehicle. The invoice price is the price a car dealer pays the manufacturer. For public transit buses, the invoice price was 90% of the manufacturer suggested retail price to account for typical block orders by transit systems (Edmunds 2011). The ICEV is the base vehicle, insurance, license, registration and tax costs for HEV were estimated based on its weight. License, registration and taxes costs were set at $0.052 per mile (AAA 2012). The insurance cost for an ICEV was set at $0.085 per mile (AAA 2012). Passenger fares for buses were estimated to be $1.50 for a 3.9 miles trip (APTA 2009). For CS, an average of 5.5 miles per hour of city usage and an hourly average cost of $8.25 were assumed (Zipcar 2012). The fixed costs of annual fee and application fee are spread over a 3,850 annual CS mileage (TCRP 2005) and 10 years of membership was assumed. Gas and insurance are included in the membership. Includes the cost for maintenance and tires replacement. The maintenance costs of the passenger vehicles were estimated based on ICEV maintenance cost of $ per mile. It is estimated that its maintenance cost is $ per mile based on the maintenance schedule and costs relative to an ICEV (35). Tire cost for ICEV and HEV is $ per mile (2011$) (AAA 2012). The maintenance cost for transit buses was estimated to be $ per mile including tires (Chandler and Walkowicz 2006). Monthly expenses for parking the vehicle (national average). For ICEV and HEV parking price is estimated based on the U.S. national average (Duvall 2002). For CS users, it was assumed that free parking is offered in designated areas. Indirect costs such as city taxes subsidize the free stalls for alternative fuel vehicles; these costs were not included. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 99

106 SESSION: PUBLIC TRANSPORT PLANNING Users Maximize user and community satisfaction Global availability (It is expressed as an annual percentage) Reasonable availability (It is expressed as an annual percentage) Passenger/Leg room, Cargo space Access time (Jeon et al. 2008) (Jeon et al. 2008) (Cambridge Systematics 2009). (Jeon et al. 2008) Time during which a vehicle is not available to its users during a day. It is estimated by dividing the total hours when a vehicle is unavailable per year by the total number of hours in a year. The unavailable hours for vehicles were estimated by multiplying the time it takes to fuel a vehicle times the fueling frequency per year. We assumed that transit buses are not in operation for five hours per day (from midnight to 5 am). Time during which a vehicle is not available to its potential users during the 19 hours (5 am to 12 am) per day when 98.8% of total trips occur. Reasonable availability: Time during which a vehicle is not available to its potential users during the 19 hours from 5 a.m. to midnight. Public transit buses are assumed to be fully fueled upon start of service and do not require fueling until the end of their shift. Physical vehicle characteristics which maximize user comfort and convenience. Passenger space, cargo space, leg room space available to each passenger. For transit buses it was assumed that the space under seats is the cargo space available to each passenger (Zimmerman and Levinson 2004). The time that is required for a user to reach the vehicle and start the trip. Includes walking and waiting time. Indicator values can be replaced with local data Access time is the time spent to reach the vehicle and start the trip. For bus, it is estimated to be 12.3 minutes based on bus stop spacing and mean headway. Access time for CS is estimated to be 5.0 minutes. Fueling opportunities (Cambridge Systematics 2009). Available locations for fueling or charging a vehicle (regional planning). Indicator is not applicable to public transit modes. It is expressed by the number of gas stations (DOE 2011). a Economic sustainability indicators focus on affordable transportation systems, therefore user out-of-pocket costs are used. The indicator values stated above can be replaced with local data. Several performance measures used for evaluating the sustainability dimensions were collected from the literature and are shown in Table 1 along with their sources. These measures have been modified to address objectives by identifying individual vehicle features that improve transportation sustainability. Indicators including, emissions, energy, trip cost, fuel cost, or trip time that usually apply to vehicle operation only; they are generalized over a vehicle s life cycle. Additional indices and measures can be incorporated in the proposed sustainability framework and Indicator values can be replaced with local/regional data Selected urban transportation vehicles This assessment focuses on road-going vehicles and the component of highway infrastructure is not included in the sustainability assessment because it is common to all vehicles assessed. For each vehicle type a representative vehicle was chosen based on highest volume of sales (Edmunds 2011). The six urban road vehicles examined were: 1) Internal Combustion Engine Vehicle (ICEV), 2) Hybrid Electric Vehicle (HEV), 3) Car-Sharing program with ICEV (CS ICEV), 4) Car-Sharing program with HEV (CS HEV), 5) Diesel Bus (DB), and 6) Hybrid Diesel Electric Bus (HDEB). The CS companies tend to take CS vehicles out of service after one or two years in service, thus for this study it is assumed that the average lifetime for a vehicle in a CS program is two years. After that time the vehicle is assumed to be in operation as a private vehicle for 8.6 years, which is the difference between a passenger car s lifetime and years in CS service. For CS programs the majority of vehicles are driven for 18,000 miles per year (CST 2010). To estimate the impact of CS on passenger miles traveled (PMT) it is assumed that each CS vehicle reduces the need for four privately owned vehicles with average 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 100

107 SESSION: PUBLIC TRANSPORT PLANNING occupancy of 1.15 per vehicle. Therefore, each CS vehicle has an equivalent occupancy of The insurance cost of a CS vehicle ranges from $4,800 to $6,000 per vehicle per year (CST 2010); we used the average value of $5,400 per vehicle year. Table 2. Vehicle Specifications (APTA 2011; Edmunds 2012; EIA 2012; TCRP 2005) ICEV HEV CS ICEV CS HEV DB HDEB Weight lbs 3,307 3,042 3,307 3,042 26,000 28,500 Average lifetime years Average annual usage miles 11,300 11,300 18,000 18,000 41,667 41,667 Fuel economy mpg (urban) Fuel price (Jan. 2012) $ per U.S. gallon Quantification of sustainability performance measures The indicators for each sustainability dimension were quantified based on the conditions described in Table 2. For the quantification of emission and energy the Greenhouse Gases, Regulated Emissions and Energy Use in Transportation (GREET) models, the MOBILE 6.2 model and the Economic Input-Output Life Cycle Assessment (EIO-LCA) were used (CTR 2005; EPA 2003; Hendrickson et al. 2006). The details and assumptions for the quantification of indicators have been covered in detail in Assessment of Sustainability for Transportation Vehicles (Mitropoulos and Prevedouros 2013) and are summarized in Table 1 and below. Manufacturing. Manufacturing emissions and energy in GREET 2.7 include vehicle materials, batteries, fluids and vehicle assembly. The weight and battery properties of each vehicle are input data along with GREET s material percentage composition of each vehicle component (e.g., body, powertrain, chassis, transmission, generator, etc.) Manufacturing emissions and energy inventory for transit buses are estimated with EIO-LCA. Fueling. GREET is used for the life cycle analysis of fuel. The model estimates the emissions and energy associated with primary energy production (feedstock recovery), transportation and storage, and with fuel production, transportation, storage and distribution. The fuel production option for conventional gasoline and low sulfur diesel is petroleum. Operation and Idling. MOBILE 6.2 was used to estimate the emissions generated from gasoline vehicles. Urban average speeds of 28 and 12 miles per hour were used for passenger vehicles, and transit buses, respectively. Idling emissions were estimated based on the assumption that the 2.5 mph emission factors can be applied to the entire idling time. Our study assumed that passenger vehicles and a transit buses idle for 7.5 and 35 minutes per day, respectively. Maintenance. Vehicle maintenance includes the maintenance and disposal of vehicle parts. GREET examines the emissions and energy associated with vehicle maintenance. EIO-LCA was used to estimate the emissions and energy inventory associated with automotive mechanical repair and maintenance and the tire manufacturing services based on maintenance costs. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 101

108 SESSION: PUBLIC TRANSPORT PLANNING 4. AGGREGATION OF INDICATORS IN SUSTAINABILITY INDEX The summation of sustainability assessment from each indicator to a sustainability dimension index, and the development of an overall sustainability index for each urban transportation vehicle is a major task. The proposed sustainability indicators have a positive or a negative impact. A larger absolute value of each indicator indicates a more positive or negative impact on sustainability. The addition among indicators with different units is performed only after the different measurement units are normalized into a dimensionless scale as detailed in (Jeon et al. 2008; Maoh and Kanaroglou 2009). The normalized values are dimensionless and range from 0 to 1. Problems that involve multiple criteria and alternatives, such as the ones associated with sustainable transportation, are defined as Multi Criteria Decision Making (MCDM) problems. In a MCDM problem, weights (w1, w2,, wn) are assigned to alternatives to account for their relative importance. Weights can be assigned directly by the decision maker, or by a group of experts (e.g., Delphi method), or can be determined by a methodology, such as cluster or factor analysis. Other methodologies, such the Bayesian decision theory or fuzzy logic can be used to account for non-linear, interrelated and stochastic aspects of transportation (Paz et al. 2012). Utilization of equal weights to minimize bias is less challenging and may be a preferred base assessment. In this paper, aggregation of normalized indicators into an index per vehicle type was performed by using the weighted sum method (Yoon and Hwang 1995). The commonly used weighted sum model (WSM) was employed in this study to aggregate sustainability indicators. WSM was used by Maoh and Kanaroglou (2009) and Jeon et al. (2008) to aggregate normalized values of criteria into sustainability dimension indices and an overall sustainability index per studied scenario. Addition among criteria with different units is performed only after the different measurement units are normalized into a dimensionless scale. The utility Vi for each alternative is estimated as follows. 1,,. 1 1 Where wj is the assigned weight for each indicator j for alternative i, Nij is the normalized value of indicator j for alternative i. 5. RESULTS AND COMPARISON The sustainability dimension index and the overall sustainability index per vehicle type summarize the sustainability performance for each vehicle type. The indicator values were quantified and weighted per PMT. The outcomes are presented in Table 3 and provide a comparison for estimating the total impact of any fleet mix scenario containing our six vehicle types. As expected, the results vary substantially: Using PMT as the base, the hybrid bus ranks third from top. CS ranks first. Among the six vehicles types examined, the most sustainable vehicle relative to the other five is found to be CS with HEV. The 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 102

109 SESSION: PUBLIC TRANSPORT PLANNING configuration of car-sharing with a hybrid vehicle receives near perfect relative scores for Technology and Energy; and respectively. The superior performance of CS over transit bus reveals the important role of policies towards sustainable transportation. Privately owned vehicles that are ranked in the bottom two positions are propelled the top two positions when used in a CS program. Comparing a CS and a HDEB, the environmental index improves by 47% if a CS vehicle reduces the need for four privately owned vehicles. Using the cited upper values of auto ownership substitution of six and 23 for North America, the environmental index improves by 57% and 75%, respectively. Propulsion systems that depend partially on electric drive did better than the traditional ICE technology when bus and passenger cars are compared. Environment Technology Performance Energy Economy a Users Table 3. Vehicle Sustainability Indicators and Relative Indices Indicators Code Units ICEV HEV CS ICEV CS HEV DB HDEB CO 2 grams/ PMT CH 4 grams/pmt N 2 O grams/pmt GHGs grams/pmt VOC grams/pmt CO grams/pmt NO x grams/pmt PM 10 grams/pmt SO x grams/pmt Average noise level db Fuel frequency hours/passenger NA NA Maintenance frequency hours/passenger Space occupied m 2 /passenger Engine power + lb.ft./lb Manufacturing energy Mjoule/ PMT Fueling energy Mjoule/ PMT Operation energy Mjoule/ PMT Maintenance energy Mjoule/ PMT Manufacturing cost $/PMT Operation cost (user) $/PMT Maintenance cost $/pass./year Parking price $/passenger NA NA NA NA hours of down Global availability time or not 0.03% 0.02% 0.05% 0.04% 20.83% 20.83% operable per Reasonable availability year expressed 0.04% 0.03% 0.00% 0.00% 0.00% 0.00% as an annual % Passenger space + cu.ft/passenger Goods carrying (cargo) space + cu.ft/passenger Leg room front + inches Access time minutes number of Fueling opportunities + stations in 121, , , ,446 NA NA operation Environment Sustainability Vehicle Index Technology Sustainability Vehicle Index Energy Sustainability Vehicle Index Economy Sustainability Vehicle Index Users Sustainability Vehicle Index Overall Sustainability Vehicle Index (PMT) nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 103

110 SESSION: PUBLIC TRANSPORT PLANNING a All costs are converted in 2011$. All Economy indicators are assumed to have negative impact to sustainability. Indicators are perceived from users point of view; therefore they reveal how vehicle monetary parameters may affect vehicle utilization and make sustainable or unsustainable a transportation vehicle for a chosen network. 6. SUSTAINABILITY FRAMEWORK IN A TRAVEL SCENARIO APPLICATION Car-sharing vehicles can be used by both car owners and non owners; they can also be used exclusively or for part of a trip together with another vehicle such as bus, bike or train. A case with three scenarios is developed to show that our sustainability assessment framework represent more realistic situations. The annual distance traveled travelled is considered to be 33,004 PMT (19,850 VMT) per household (FWHA 2011). In this application, shown in Table 4, annual traveled distance per household changes for public bus and CS users to reflect travel behavior changes when different transportation modes are chosen. The sustainability assessment framework should be sensitive to this change. Three scenarios of vehicle options were analyzed: (a) passenger vehicle only, (b) passenger vehicle and public bus, and (c) car-sharing and public bus. The National Household Travel Survey (FWHA 2011) provides data on annual PMT and VMT by trip purpose. Based on this information and the survey on CS trip purposes (TCRP 2005) the usage share per vehicle type was estimated. Prior to examining the results we should note that some indicators may be removed from the assessment. For example, in this application fueling and maintenance frequency indicators for buses are not applicable because the assessment is performed from the user perspective and such indicators do not affect a user s trip. For Scenario 1, the private car is used for all activities. For Scenario 2, the public bus is used for to or from work activities and the private car is used for the rest of the activities, including errands, shopping and recreation. For Scenario 3, distance traveled by bus is the same as for Scenario 2, and CS substitutes for the private car. Typically the annual distance traveled by car owners, CS subscribers, and bus users is different. CS and bus users reduce their annual mileage by 40% and 6%, as suggested by Shaheen et al. (2006) and Briceno et al. (2005), respectively. Therefore the annual PMT for Scenario 1 remains the same whereas annual PMT for Scenarios 2 and 3 are reduced by 6% and 40%, respectively. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 104

111 SESSION: PUBLIC TRANSPORT PLANNING Table 4. Sustainability Indices for Travel Scenario Application Scenario 1 Scenario 2 Scenario 3 DB HDEB DB HDEB ICEV HEV ICEV HEV ICEV HEV CS ICEV CS HEV CS ICEV CS HEV Distance/car (VMT) 19,850 19,850 14,337 14,337 14,337 14, Distance/bus (PMT) a 0 0 7,186 7,186 7,186 7,186 5,500 5,500 5,500 5,500 Distance/CS (VMT) ,602 8,602 8,602 8,602 Total distance travel (PMT) 33,004 33,004 31,024 31,024 31,024 31,024 19,802 19,802 19,802 19,802 Environment Technology Energy Economy Users OSI Ranking a The travel distance for public bus is in units of PMT; for private car and CS is in units of VMT. Distance traveled per activity in each scenario was reduced by the same percentage. Annual distance traveled for Scenario 2 is 31,024 PMT. It is 19,802 PMT for Scenario 3. The scenario results are visualized in Figure 1. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 105

112 SESSION: PUBLIC TRANSPORT PLANNING Environment Technology 1 0,75 0,5 0,25 0 HDEB_CS HEV HDEB_CS ICEV 0,75 0,5 0,25 ICEV 1 0 HEV DB_ICEV DB_CS HEV DB_HEV DB_CS ICEV HDEB_ICEV HDEB_HEV Energy Economy HDEB_CS HEV HDEB_CS ICEV 0,75 0,5 0,25 ICEV 1 0 HEV DB_ICEV HDEB_CS HEV HDEB_CS ICEV 0,75 0,5 0,25 ICEV 1 0 HEV DB_ICEV DB_CS HEV DB_HEV DB_CS HEV DB_HEV DB_CS ICEV HDEB_ICEV DB_CS ICEV HDEB_ICEV HDEB_HEV HDEB_HEV Users HDEB_CS HEV HDEB_CS ICEV DB_CS HEV 0,75 0,5 0,25 ICEV 1 0 HEV DB_ICEV DB_HEV DB_CS ICEV HDEB_HEV HDEB_ICEV Figure 1. Visual Representation of Vehicle Indices per Sustainability Dimension for the Three Scenarios CO2 and GHG emissions decrease by 42% and 41%, respectively relative to the case that travelled distance reduction was not occurred. This is consistent with Shaheen et al. (2006) who reported that car-sharing lowers average user CO2 emissions by 40% to 50%. The results in Table 4 show that the highest overall sustainability indices are attributed to Scenario 3 for the CS HEV when mixed with a HDEV and a DB. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 106

113 SESSION: PUBLIC TRANSPORT PLANNING Based on the Overall Sustainability Index (OSI) our estimates indicate that for an ICEV-and-public bus mix to be more sustainable than a HEV private car, the annual distance traveled in Scenario 2 should be reduced by an additional 20%. Clearly having CS in the mix is the most sustainable option examined. Changes in travel behavior are very significant in planning of a sustainable transportation system and our sustainability assessment framework is a sensitive and practical tool for assessment based on highway modes with conventional and alternative powerplants and fuels used. 7. CONCLUSION Incorporation of sustainability into transportation planning process was explored by the development of a sustainability framework that disaggregates vehicle characteristics by technology and fuel type and accounts for evolving travel regimes. The sustainability framework was used to assess car-sharing a new transportation mode with existing urban vehicles. The vehicle sustainability assessment revealed that car-sharing with hybrid-electric vehicles yield a better performance over private vehicles and transit bus, conventional of hybrid. This outcome suggests an important role of policies towards sustainable transportation. Private owned vehicles which are found to have the worst sustainability performance shift to being the best vehicle, relative to the rest of the vehicles in this assessment, when they are used in a car-sharing plan. Sustainability performance of transportation systems can be improved in the short term without relying solely on advanced vehicle technologies such as electric and fuel cell vehicles. Implementing policies such as car-sharing and carpooling, are some of the immediate measures that can be used to improve the sustainability performance of urban transportation systems. Car-sharing in this application was found to have the highest sustainability index when it is combined with travel behavior changes, such as reduced distance traveled. Additionally, hybrid technologies and alternatives to driving should be promoted by policies such as free entry to congestion zones and HOV lanes, as well as telecommuting in order to improve transportation system sustainability. Upfront membership fees joining for car-sharing might prevent low income groups from using them. Minimization or incorporation of membership fee in the rates per mile or hour might work as in incentive for using a car-sharing vehicle. In addition, the inclusion of the supporting infrastructure in the assessment will provide a complete picture of the sustainability performance of any transportation mode over its life cycle. Infrastructure and land use planning are major components of transportation. This is a necessary expansion for the comprehensive sustainability assessment of both highway and guideway urban modes and would enable to assess the service of mass transit for physically, economically and socially disadvantaged people. Overall, this study updated the state of the art in three main areas: (i) employed a life cycle approach of vehicles instead of focusing only on the operation of modes; (ii) disaggregated vehicles by type instead of assuming a uniform light-duty vehicle fleet; and (iii) assessed conventional and hybrid technologies explicitly for three different travel regimes instead of assuming only fossil fuel powered vehicles and private cars. Acknowledgements: The research presented in the present paper has been conducted within the context of the FUTRE project FUture prospects on TRansport evolution 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 107

114 SESSION: PUBLIC TRANSPORT PLANNING and innovation challenges for the competitiveness of Europe, funded by the Research Directorate General of the European Union under the Seventh Framework Programme for Research and Technological Development. References AAA - American Automobile Association. (2012). Your driving fixed costs. Accessed June Anielski, M. (2001). Alberta GPI blueprint report. Pembina Institute Accessed 2 September APTA - American Public Transportation Association (2009). Public transportation fact book, 60th Edition. Bevan, T.A., Donna, L., Senner S., Seskin, S. (2008). Planning for sustainability: planning for sustainable transportation infrastructure. Canadian Institute of Transportation Engineers. Briceno, T., Peters, G., Solli, C., Hertwich, E. (2005). Using life cycle approaches to evaluate sustainable consumption programs, car-sharing. Industrial Ecology Programme. Cambridge Systematics. (2009). Performance measurement framework for highway capacity decision making. Strategic Highway Research Program. Chandler, K., Walkowicz, N. (2006). King county metro transit hybrid articulated buses: Final evaluation results. National Renewable Energy Laboratory. Chester, M.V., Horvath, A. (2009). Environmental assessment of passenger transportation should include infrastructure and supply chains. Environmental Research Letters, 4. Accessed June Cowan, P. (1994). Handbook of environmental acoustics. John Wiley & Sons. CST - Car Sharing Tips. (2010). Guide to used car auto auctions. Hhtp://carbuyingtips.com/auto-auctions.htm. Accessed January CTR - Center for Transportation Research. (2005). Operating manual for GREET - Version 1.7. Argonne National Laboratory. DOE - Department of Energy. (2013). Data, analysis and trends. Hybrid Electric vehicles sales by model. Accessed 10 May DOE - U.S. Department of Energy. (2011). Alternative and advanced fuels - Energy efficiency and renewable energy. Accessed 30 December Dobranskyte-Niskota, A., Perujo, A., Pregl, M. (2007). Indicators to assess sustainability of transport activities. European Commission, Joint Research Centre, Institute for Environment and Sustainability. Duvall,. M. (2002). Comparing the benefits and impacts of hybrid vehicle options for compact sedan and sport utility vehicles. EPRI. EEA - European Environment Agency. (2002). Paving the way for EU enlargement Indicators of transport and environment integration TERM Accessed August Edmunds. (2011). Accessed December EIA - U.S. Energy Information Administration. (2012). Gasoline and diesel fuel update. Accessed March nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 108

115 SESSION: PUBLIC TRANSPORT PLANNING EPA - Environmental Protection Agency. (2011) Inventory of U.S. greenhouse gas emissions and sinks: , 430-R , Washington D.C. EPA - U.S. Environmental Protection Agency. (2006). The EPA 10 gallon per minute fuel dispensing limit. Accessed 15 August EPA - U.S. Environmental Protection Agency. (2003). User s guide to MOBILE6.1 and MOBILE6.2 mobile source emission factor model. FUTRE - FUture prospects in Transport evolution and for the competitiveness of Europe. (2014). Deliverable 3.2, Long-term future analysis on transport demand market and drivers. Accessed April FWHA Federal Highway Administration. (2011). U.S. National Household Travel Survey Hendrickson CT, Lave LB, Matthews SH. (2006). Environmental life cycle assessment of goods and services, an input-output approach. RFF Press Book. ITE - Institute of Transportation Engineers. (1999). Transportation planning handbook. 2nd edition. Washington, DC. Jeon, C.M., Amekudzi, A., Guensler, R.. (2008). Sustainability assessment at the transportation planning level: Performances and measures and indexes. Transportation research board annual conference. CD-ROM, January 13-17, 2008 in Washington D.C. Maoh, H, Kanaroglou, P. (2009). A tool for evaluating urban sustainability via integrated transportation and land use simulation models. Urban Environment, 3 : Martin, E.W., and Shaheen, S. (2011). Greenhouse gas emission Impacts of carsharing in North America. IEEE Transactions on Intelligent Transportation Systems, 12 (4). Mitropoulos, L.K., Prevedouros, P.D. (2013). Assessment of sustainability for transportation vehicles. Journal of the transportation research board, Transportation research board of the national academies, Washington, D.C., No. 2344: ORNL - Oak Ridge National Laboratory. (2012) Vehicle technologies market report. Paz, A., Maheshwari, P., Kachroo, P. (2012). Estimation of performance indices for the planning of sustainable transportation systems. Transportation Research Board, 91st Annual Meeting, Compendium of Papers, Washington, D.C. Pintér, L., Hardi, P., Bartelmus, P. (2005). Sustainable development indicators, proposal for a way forward. UN Division on Sustainable Development, Accessed 10 July Renne, J.L. (2009). Evaluating transit-oriented development using a sustainability framework: Lessons from Perth s network city. Planning sustainable communities: Diversity of approaches and implementation challenges, University of Calgary: Shaheen, S., Cohen, A.P. (2013). Worldwide carsharing growth: An international comparison. Accessed March Shaheen, S., Cohen, A.P., Roberts, J.D. (2006). Carsharing in North America: Market growth, current developments, and future potential. Journal of the transportation research board, Transportation research board of the national academies, Washington, D.C., No. 1986: nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 109

116 SESSION: PUBLIC TRANSPORT PLANNING Stasko, T.H., Buck, A.B., Gao, H.O. (2012). Impacts of carsharing in a University Setting: Changes in vehicle ownership, parking demand and mobility in Ithaca, NY. Transportation Research Board Annual Conference. CD-ROM, January 13-17, Washington D.C. TCRP - Transit Cooperative Research Program. (2005). Car sharing: Where and how it succeeds. Transportation Research Board of the National Academies, Report 108. Yoon, K.P., Hwang CL. (1995). Multiple attribute decision making, an introduction. Sage University paper. Quantitative applications in the social science series. Zietsman, J., Rilett, L.R., Kim, S.J. (2003). Sustainable transportation performance measures for developing communities. Texas Transportation Institute. Accessed May Zimmerman, S.L., Levinson, H. (2004). Vehicle selection for BRT: Issued and options. Journal of Public Transportation, 7 (1): Zipcar. (2012). Accessed June nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 110

117 SESSION: PUBLIC TRANSPORT PLANNING Public Transport Emissions Measurement and Reactive Measures Maria Boile 1, Afroditi Anagnostopoulou 1, Eleftherios Sdoukopoulos 1, Georgios Papapanagiotakis 2 Abstract 1 Centre for Research and Technology Hellas / Hellenic Institute of Transport 52 Egialias Str Marousi, Athens, Greece boile@certh.gr, a.anagnostopoulou@certh.gr, sdouk@certh.gr Emphasis Telematics 2 Mesogion Av. 123 and Krimeas 1 Str Athens, Greece gpap@emphasisnet.gr The use of technology and telecommunication developments could contribute to the sustainability of the cities minimizing the environmental footprint and improving the operational costs caused by the public transport sector. Smart ICT tools provide the relevant and appropriate information that constitute the basis for more effective decisions according to the current circumstances and conditions. To this end, a potential solution scheme proposed to rationalize and improve the efficiency of public transport and to promote reduced energy consumption and associated greenhouse gas (GHG) emissions while promoting economic savings and social benefits. Keywords: ICT tools, public transport, GHG emissions 1. INTRODUCTION The study of the public transport sector has gained significant attention over the last decades since the environmental externalities in urban areas have significantly increased and public authorities have already started to develop master plans for improving the sustainability of the cities. Traffic management is at the core of the activities that have been performed by transport authorities aiming to promote the goals of public transport policy, which mainly focuses on improving the efficiency of public transport and reducing the resulting environmental impact. In order to achieve these goals, traffic authorities have implemented various traffic management measures including public transport vehicle priority, dedicated HOV lanes etc. Monitoring and evaluating the level of success and the impact of such strategies over time proves to be an essential factor supporting the decision making process. Capitalizing upon such capabilities, an essential goal would be to rationalize and improve the efficiency of the public transport system, promoting the reduction of the associated energy consumption and greenhouse gas (GHG) emissions, realizing economic savings and other social benefits. For these reasons, the current technological advances can be utilized for having an efficient public transport system that regional authorities can use and afford. The use of smart Information and Communication Technologies (ICT) in transport operations could contribute, to a large extent, to the reduction of the environmental pollution in cities in terms of carbon emissions, fuel and energy consumption. In an effort to support sustainability, ICT solutions may be integrated in public transport services with the aim to minimize the resulting carbon footprint in cities considering 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 111

118 SESSION: PUBLIC TRANSPORT PLANNING the recent prediction of the Climate Group 1 that reveals a 75% increase of Greenhouse gas by ICT tools constitute innovative and feasible approaches beyond the common practices for facilitating better coordination, communication and planning for green public transport in cities. Although several ICT public transport solutions have already been applied within the European context, real-life implementations targeting the environmental effectiveness of the public transport system are still limited. The above considerations form the background of this paper, which aims to present trends in the field of public transport, the ICT tools that have been developed worldwide for improving the energy efficiency of public transport in cities and provide recommendations for future research and development, well aligned with the general objective of promoting innovative, energy efficient and green solutions for cities that will help improve the current knowledge of public and private actors thus meeting the sustainability goals. The rest of the paper is structured as follows: Section 2 provides a comprehensive description of available ICT tools providing information about fuel consumption and traffic movements that reveal their utility as well as the potential benefits that can be derived. Taking a step further, Section 3 summarizes basic observations about the use of ICT tools in public transport management and planning providing also recommendations for future developments in smart urban areas while the last section (Section 4) draws the main conclusions. 2. RECENT DEVELOPMENTS Due to the wide applicability and high complexity that characterizes public trans- port planning, significant developments have been made focusing on management policies, decision support systems and ICT solutions. Simple rationing policies have been introduced at first (Filippi 1996) focusing on restricting the access of different types of vehicles to particular areas, especially during rush hours, promoting in that way the use of public transport. These policies were more recently advanced to pricing policies as presented by Gentile (2005) for managing travel demand based on simplied schemes. Meignan (2007) developed a bus-network simulation tool to analyze and evaluate a bus-network adopting a multi-agent approach in an effort to describe the interactions among travelers and buses within the public transport network. Similarly, Cats (2011) provides a dynamic transit analysis and evaluation tool for network level and Cats (2013) proposes a framework of a transit system with multi-agent operations and an assignment model to capture uncertainties and user decisions. In the field of the energy savings in public transport, Barrero (2008) studied the use and the impacts of electric vehicles (battery electric buses, trolley buses, trams, metro, light rail) in public transport systems, and Zhang (2013) discusses possible strategies for improving energy consumption considering technological solutions, infrastructure improvement, management schemes and travel modes. The introduction of ICT to the public transport system provides the opportunity to transcend time and space by having more accurate access anytime and anywhere. 1 The Climate Group - Smart 2020 Report, June Available online at: 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 112

119 SESSION: PUBLIC TRANSPORT PLANNING Important factors such as the distance traveled, traffic conditions and driver profile, are essential for analyzing factors that affect energy consumption and environmental impact as well as for determining the parameters of the emissions factors based on realtime information and developing tools to implement economic and environmental beneficial management decisions. There are several innovative telematics systems, which measure and evaluate corporate emissions related to transportation activity, including fuel monitoring and control and advanced fleet management systems similar to the ones that are briefly described below: Fuel monitoring systems such as the e-fuel tool 2 enable remote fuel control offering the opportunity to public authorities to distantly monitor and get fuelrelated information. These tools measure fuel in the tank and automatically transmit the measurement with no human intervention, providing a fuel/km report. Specifically, e-fuel is comprised of (a) a telematics vehicle device which is installed on the vehicle and constantly records the fuel quantity in the tank, the kilometers traveled, and the vehicles location, and (b) communications including the GPRS network that provides the opportunity to be updated online regarding the fuel quantity, the kilometers traveled, and the position of the vehicle(s). The center of administration of the fleet (base station) receives analytic reports fostering the real-time control of fuel costs and the operations of the vehicles (fillings, drainages, consumption in relation to kilometers). The tool can operate at all gas station and refueling points and it can be installed in all types of vehicles. In combination with the kilometers and the speed of the vehicle monitored, fuel consumption can be calculated with accuracy offering real-time information so that fuel losses can be avoided due to faulty invoicing, tank drainage, excessive consumption. Fleet management systems such as the e-track tool 3 are used for managing a fleet of vehicles using advanced and sophisticated communications, logistics and IT technologies. Such tools monitor, simplify, optimize and support management and control of the vehicles using the GPRS network offering immediate cost benefits as well as public transport strategy benefits. e-track for example, consists of (a) the vehicle unit: the electronic device installed in the vehicle that is based on the GPS (Global Positioning System) satellite tracking system and constantly records the vehicles position as well as other important data, (b) telecommunications that optimally utilise the GPRS network offering on-line notification regarding the position and status of your corporate / public vehicles, and (c) the base station that receives the required information, either through direct communication with the vehicles or over the Internet (by connecting to the e-track Server). Such systems present examples of technological investment which may produce immediate financial and strategic benefits for reducing fleet operating costs, supporting strategic decisions and processes by providing reliable information about routes, traveling speed and stop duration. In the dynamic environment of public transport planning, the aforementioned innovative technologies are able to provide accurate real time traffic information and fuel consumption data for buses and other public transport vehicles in an at- tempt to provide competitive traffic management decisions as well as improvements in terms of ibid 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 113

120 SESSION: PUBLIC TRANSPORT PLANNING public transport services quality, energy consumption and pollutant emissions. Public transport operations could be improved by selecting appropriate vehicle technology, evaluating driver profiles, providing specialized training to drivers and optimizing scheduling. 3. MOTIVATIONS FOR FUTURE RESEARCH In today s economy, cities tend to expand at a quick pace highlighting the need for an efficient and sustainable public transport system, which is vital for economic success and the provision of high level of services. Environmental concerns and management issues related to the flows of passengers have become critical concerns for modern cities seeking to reduce energy consumption and the related production of greenhouse gas (GHG) emissions considering to this end the application of more effective and successful traffic management strategies. Capitalizing upon such capabilities, an ICT solution framework should be introduced to utilize current technological advances for efficient public transport in order to provide competitive traffic management decisions as well as improvements in terms of energy consumption and pollutant emissions. More specifically, the development of a module generating reliable network traffic and vehicle emissions maps as well as a set of criteria and performance indicators to assist in assessing traffic management decision tools prove to be necessary for implementing traffic management strategies and realizing significant economic and environmental benefits. A promising research pointer suggests a decision support tool in order to minimize the energy consumption of public transport vehicles resulting in less pollutant emissions as well as in improved operating costs. For this purpose, ICT are necessary to provide the relevant and appropriate information that constitute the basis for more effective decisions according to the current circumstances and conditions. To this end, a potential solution scheme aims to rationalize and improve the efficiency of public transport and to promote reduced energy consumption and associated greenhouse gas (GHG) emissions while promoting economic savings and social benefits. It should be an attempt to utilize the current technological advances for efficient public transport that regional authorities can use in order to provide competitive management decisions as well as improvements in terms of energy consumption and pollutant emissions. A future solution approach may utilize FCD (Floating Car Data) as a basis for determining traffic flow conditions and calculating aggregate vehicle emissions at a network level using existing estimation models and in return, emissions would be calibrated using real information at a vehicle level. In such a system public authorities (such that urban transport organizations) use both network traffic information (at a network level) and real data at a vehicle level. This ensures product evolution using gradually improved and more accurate traffic models for decision making as it is depicted in 1 below. Hence, urban transport organizations would be able to assess the overall traffic and environmental improvements at a network level but also monitor the impact of measures at individual vehicle or corporate level. In particular, FCD could serve as the basis for determining traffic volumes and traffic composition. The traffic data, along with statistical fleet composition, na- tional and local data, would then be used for the calculation of vehicle-generated emissions at the network level using existing emissions models and providing. Vehicle traffic 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 114

121 SESSION: PUBLIC TRANSPORT PLANNING parameters and data provided by vehicles equipped with monitoring devices would serve as input to the traffic model to generate the environmental footprint at the vehicle level. Supporting the evaluation process of various traffic management strategies would be achieved by the development of appropriate comparative performance indicators (CPIs), reflecting both traffic efficiency aspects and environmental aspects enabling a before and after evaluation. CPIs would be developed for the evaluation of traffic management strategies implemented by transportation authorities and the dimensions incorporated into their design would be derived from the nature of traffic management decision making processes. The spatial-temporal nature of the transportation system as well as the transport modes would be considered for the development of the CPIs in order to enable summation across various components as well as to avoid situations in which certain components dominate others. Designing the CPIs in this manner will enable to compare various traffic management strategies based on a broad view, and at the same time disaggregate the overall CPIs into their components as a basis for identifying success and failure factors. In terms of innovative technologies, a possible development could be a tool to measure and evaluate transit agency fleet emissions related to transportation activity including the following components: An on board system: a microcontroller based platform which includes GPS for obtaining information regarding the location of the vehicle and consists of a hardware interface for the interconnection between the on board system and interfaces of the vehicles manufacturer as well as an embedded Software protocol for the interconnection between the on board system and the CAN FMS protocol for the buses as well as the ODB2 protocol for taxes. It also includes acquisition of vehicle operations data, additional sensors to approve emissions measurements and a GPRS modem for data transfer from the system to the central database. Emissions analysis tool: a tool that receives emissions relative data from the on board units and provides analysis of the measured emissions data. Emissions are categorized according to the business activity taking into account various dimensions of the business activity such that assigned transportation tasks, traffic conditions, driver profile, vehicle type and status. This tool is based on web software technologies and includes reporting, notifications and user role functionality. It could assist transit operators to make assessments on the fuel consumption and emissions and evaluate new transportation strategies. Figure 1. Solution Cycle 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 115

122 SESSION: PUBLIC TRANSPORT PLANNING 4. CONCLUSIONS In order to achieve the environmental and energy goals set by the European Commission, several European cities have implemented or are in the process of investigating different strategies that can contribute to meeting sustainability targets. Public transport systems play a significant role in the urban transport network and several policies have been introduced favoring the use of public transport. However, there is still room for improvement and ICT tools present a good opportunity for realizing significant environmental and energy savings. Having real-time information regarding the movement of public transport vehicles and their fuel consumption enables the investigation of different solutions (e.g. eco-driving, etc.) that can impose a significant impact on the environment leading to more sustainable cities. References Barrero R Van Mierlo J Tackoen X (2008) Energy savings in public transport. Veh Technol Mag, IEEE 3(3):26 36 Cats O Koutsopoulos H Wilco B Toledo T (2011) Effect of real-time transit information on dynamic path choice of passengers. Working papers in Transport Economics 2013:28, CTS - Centre for Transport Studies Stockholm (KTH and VTI) Cats O (2013) Multi-agent transit operations and assignment model. Procedia Comput Sci 19: Filippi F Persia L et al (1996) Public transport prioritization. Transport Res APAS Urban Transport Report DG VII, Office for Official Publications of the European Commission, Luxembourg Gentile G Papola N Persia L (2005) Advanced pricing and rationing policies for large scale multimodal networks. Transportation Res A: Policy and Practice 39(7-9): Meignan D Simonin O Koukam A (2007) Simulation and evaluation of urban busnetworks using a multiagent approach. Simul Model Pract and Theory 15(6): Zhang L Zhang Y He Y Hong Y Xin R (2013) The behavior strategy in energy saving and emission reduction of transportation. Adv Mater Res 718: nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 116

123 SESSION: PUBLIC TRANSPORT PLANNING Designing Sustainable Urban Transport Interchanges Giannis Adamos 1,2*, Eftihia Nathanail 1,2 & Maria Tsami 1,2 1 University of Thessaly, Department of Civil Engineering, Transportation Engineering Laboratory Pedion Areos, 38334, Volos, Greece 2 Centre for Research and Technology Hellas (CERTH) / Hellenic Institute of Transport (HIT), 6 th km Charilaou - Thermi Rd., P.O. Box: 361, P.S.:57001, Thessaloniki, Greece *Corresponding author: giadamos@civ.uth.gr, Tel.: , Fax: Abstract Urban transport interchanges are considered as a crucial determinant of public transport networks and operations, since they facilitate the links among different transportation modes. When properly designed and managed, they have significant benefits, among which, time saving, better use of waiting time and urban integration. The scope of this paper is to capture the viewpoint and preferences of travellers on different aspects and elements for defining a smart and efficient interchange. Towards this direction, an on-site face to face questionnaire survey was conducted at the intercity bus station and the railway station in Volos, Greece. Useful information was gathered about travellers habits, preferences and satisfaction, and in total thirty indicators were rated by respondents, including issues such as travel information, wayfinding information, access, time and movement, image and attractiveness, comfort and convenience, and emergency situations. Results showed that, although the intercity bus station had a higher score in the majority of the indicators assessed, the railway station seems to satisfy slightly more the users, since it received higher rating in the overall satisfaction level that provides to travellers. This outcome may be explained by the fact that there are specific characteristics of the station and the relevant services offered, which strongly affect travellers opinion and consequently modal choice. Keywords: Urban transport; sustainability; interchanges; public transportation. 1. INTRODUCTION During the last decades, social and economic opportunities have been underpinned by mobility growth, resulting in significant increase of congestion in urban areas and environmental deterioration. Transport started to pose a threat for the modern societies, while satisfying the demand for mobility becomes a key determinant for citizens quality of life. Although transport and mobility are two different notions, they are strongly interrelated; transport, including infrastructure and services, is the means to achieve the requisite level of mobility, while mobility is the behavioral outcome of peoples need to travel, in order to participate in several activities, or to consume products, goods and services. And, where is the balance between what transport can provide and how the requested mobility level is achieved? Sustainable mobility forms the obvious answer, but still, the 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 117

124 SESSION: PUBLIC TRANSPORT PLANNING pursuit of sustainability presents a challenge. The solution is widely perceived to lay in an integrated approach, addressing and combining joint initiatives and integrated policies in the fields of transport, environment, land use, economic instruments, regulations, new technologies and other actions. In an era of transformations, like the current, both citizens and decision makers need to address contradictory situations; they have the technological advances and instruments to improve daily lives, but, on the other hand, the economic crisis restricts the potential use of these instruments through a mechanism of mobility management, i.e. fewer movements as a measure for savings. Digital society in combination with strategies that promote sustainability, enhance the promotion of soft transport modes, such as walking and cycling, and public transport. However, although national governments and local authorities make efforts to persuade travellers to switch mode, it seems that public transport still cannot be capable of competing with private car (Grotenhuis et al., 2007). From the customers perspective, the quality of services provided by public transport is not at a satisfying level that would urge them to replace their car with other modes, i.e. bus, train, tram or combination of them. As the world becomes more urbanized, there is a strong need that urban public transport provides a viable alternative to individual car transport. However, the availability of opportunities for direct journeys when using public transport is limited, and for this reason the majority of trips require interchange zones, thus areas which encompass one or more interchange facilities and public spaces used for access and/or transfer (Allen et al., 2013). Urban transport interchanges are considered as a crucial determinant of public transport networks and operations, since they facilitate the links among different transportation modes, and particularly the connection of bus services to the subway and metropolitan railway system (Vuchic, 2005). When properly designed and managed, the interchanges have significant benefits, among which, time saving, better use of waiting time and urban integration. The main purposes of urban public interchanges may be highlighted to the following (Di Ciommo, 2002): Facilitation of users transfer between two or more public transport modes by reducing the transfer time and increasing their convenience; Coordination of public transport services, through information services, provided at the interchange facilities; Efficient exploitation of the urban space, improving at the same time the image of the urban area and promoting the development of local businesses. On the other hand, in order to ensure the integrated and efficient transport of passengers between different transportation modes and between several routes, an urban transport interchange should (Pitsiava-Latinopoulou & Iordanopoulos, 2012; Allen et al., 2013): Provide reliable, adequate and direct level of the services offered, i.e. information, ticketing, etc.; Develop satisfactory facilities serving the transfer, i.e. service areas, waiting areas/platforms, amenities, internet access, comfort, etc.; Provide adequate accessibility of the site for all users (especially the disabled); 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 118

125 SESSION: PUBLIC TRANSPORT PLANNING Afford assistance to travellers with navigating aids, so that they can find their way from where they are to where they wish to go, both within the interchange, as well as to and from the local vicinage (wayfinding); Offer easy and seamless navigation and movement of users, improving also their understanding, enjoyment and experience (legibility); and Allow users to move around the interchange under several alternatives, providing at the same time clear connections to existing routes, facilities and services (permeability). Travellers switch from one mode to another when there is no direct connection with urban public transportation from their origin to their destination, or when they wish to reduce their total travel cost (DTO, 2000). In addition, person travel needs and perceptions affect the mode(s) choice, and subsequently the use of an urban transport interchange (TOOLQIT, 2007; KITE, 2009). Towards this direction, understanding passengers needs and opinions is an issue of high importance that should be considered by local authorities, transport providers and operators, and decision makers. The aim of this paper is to capture the viewpoint and preferences of travellers on different aspects and elements for defining a smart and efficient interchange. Also, the study aims at understanding travellers perceptions in issues such as Is the interchange environment convenient?, Is the interchange modern and dynamic? and How do you spend your time when not travelling?. Lastly, a cross-site comparison between the Intercity bus station of Magnesia and Volos Railway provides useful information about travellers habits, preferences and satisfaction. 2. METHOD AND DATA COLLECTION 2.1. Design and realization of the survey In order to achieve the above aims, an on site face-to-face questionnaire survey was conducted in the above interchanges in Volos, Greece. The survey ran from the middle of December 2013 till the middle of January 2014, and was separated into three main parts: Part A: Travellers satisfaction; Part B: Travel habits; and Part C: Socio-economic information. The first part of the survey aimed at understanding travellers views and preferences on various aspects and elements of an urban transport interchange, and the respective level of their satisfaction. This part contained twenty nine items related to various aspects of an interchange, grouped into seven main categories: travel information, wayfinding information, access, time and movement, image and attractiveness, comfort and convenience, and emergency situations. An additional item regarding the overall satisfaction was also included (Table 1). Each respondent stated his/her level of satisfaction for each of the twenty nine items and the overall satisfaction using the Likert scale ranging from 1 (strongly dissatisfied) to 5 (strongly satisfied). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 119

126 SESSION: PUBLIC TRANSPORT PLANNING In the second part of the survey, referring to the travel habits of the users, information was collected about the trip purpose, the chosen transportation mode, from origin to the interchange, and from the interchange to the destination, the respective time to/from/inside the interchange, the ticket type of public transport, etc. In the last part of the survey, socio-economic characteristics of the respondents were recorded, including gender, age, employment status, household income, etc. Table 1. List of items classified into eight main categories Main categories Items Availability and ease of travel information at the interchange Availability of travel information before the trip Travel information Accuracy and reliability of travel information displays Ticket purchase (ticket offices, automatic machines, etc.) Signposting to different facilities and services Wayfinding information Signposting to transfer between transport modes Information and assistance provided by staff Access Ease of access to the interchange Distances between transport modes Coordination between different transport operators or transport services Time and movement Use of time at the interchange Distance between the facilities and services Ease of movement due to number of people inside the interchange Surrounding area Image and attractiveness Internal design of the interchange External design of the interchange General cleanliness of the interchange Temperature, shelter from air and wind, etc. General level of noise of the interchange Air quality, pollution, i.e. from vehicles Number and variety of shops Comfort and convenience Number and variety of coffee shops and restaurants Availability of cash machines Availability of seating Availability of telephone signal and Wi-Fi General comfort Information to improve the sense of security Emergency situations Signposting to emergency exits Location of emergency exits in case of fire Overall satisfaction Level of services provided by the interchange 2.2. Participants The sample was composed of 210 travellers, 105 of which were interviewed in the intercity bus station and the rest 105 in the railway station. In the railway station, 54.3% of the participants were men and 44.8% women, while in the bus station the respective percentages were 42.9% and 57.1%. The majority of the respondents (42.9%) in the railway station were between years old, the 37.1% of them between 26-40, the 14.3% between 41-65, the 4.8% older than 65, and the rest 0.9% younger than 17 years old. In the case of the bus station, the majority of users (55.5%) were also between years old, while the same percentage, equal to 23.8%, was noticed in the age groups 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 120

127 SESSION: PUBLIC TRANSPORT PLANNING of and Low percentages were indicated in the peak and trough age groups, thus 1% in participants older than 65 (1%) and younger than 17 (0.9%) years old. Regarding the employment status of the participants, in the railway station, it was observed that the majority (41%) of travellers were students, the 37.1% employed, the 10.5% unemployed, the 8.6% retired and the rest 2.8% housewives/-men. In the bus station, it was indicated that most of travellers were employed (44.8%), the 29.5% were students, the 13.3% unemployed, the 4.8% housewives/-men, the 3.8% retired and the rest 3.8% stated other status. Lastly, as far as the education level of the travellers is concerned, the analysis of the results showed that the majority (53.3%) of users in the railway station was highly-educated, the 39% held a university degree and the rest 7.7% were primarily-educated. Similar results were observed in the case of the bus station; specifically the 71.4% of users were highly-educated, the 26.7% held a university degree and the rest 1.9% were primarily-educated Case studies Volos Railway Station The Railway station of Volos was developed in 1884, and is located in the urban area of the city of Volos. The station is close to the Intercity Bus Station of Magnesia, approximately at 1.5 kilometers, and very close to the port, approximately at 500 meters. In the surrounding area of the interchange, there are catering/coffee shops, restaurants, super markets, and public services facilities (i.e. revenue office, social security organization). Housing development is rather intense in the vicinity of the station. The interchange has a significant role in the overall transport network, although it connects Volos with only one city, thus Larisa. Still, travellers choose this transportation mode, even though they have to transfer in Larisa in order to reach their final destination. The connectivity provided among different modes at the terminal may be considered as good, since taxis, and car and bicycle parking areas are available outside the main building of the interchange. However, the connectivity of the station with the local buses needs to be improved, i.e. bus stops should be constructed nearest at the station, etc Intercity Bus Station of Magnesia The Intercity bus station of Magnesia was opened in the decade of 1970 and an entirely redevelopment took place in 1990s, including the refurbishment of the waiting and ticketing area, the storage and the offices. In addition, the roof of the bus area was replaced and the building was repainted. The station is located in the urban area of Volos, and is very close (50 meters) to the terminal local buses station and the Polytechnic campuses of the city s university (100 meters). The railway station and the port are also close enough at and kilometres, respectively. In the surrounding area of the interchange, apart from the university campuses, catering and coffee shops, and restaurants street markets are met. On the other hand, housing development is not intense. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 121

128 SESSION: PUBLIC TRANSPORT PLANNING The interchange has an ascendant role in the overall transport network, since it provides travelling services for 9 big cities (destinations out of Magnesia), 36 destinations in Magnesia, and routes from and to the airport of Volos. The connectivity provided among different modes at the terminal is considered as adequate, since the local buses station is situated at 50 meters away from the interchange, while taxis are available right outside the main building (10-15 meters). Car and bicycle parking space is also available right outside the main building of the interchange (approximately meters). 3. DATA ANALYSIS AND RESULTS In this chapter, the results of the questionnaire survey are presented, separated into two sections. The first one regards the travel habits of the interchanges users, while the in the second part the results of the travellers preferences are depicted, based on the 30 items of Table 1. For the data analysis, both descriptive and inferential statistics were used; specifically in order to record users satisfaction and assess potential statistically significant differences in the values of the items between the two interchanges, the statistical analysis of the responses was carried out using non-parametric tests, which are considered as powerful for analyzing data collected through questionnaire surveys (Siegel & Castellan, 1988). The normality of the data was estimated through the onesample Kolmogorov-Smirnov test, while the Mann-Whitney-Wilcoxon test was performed to assess differences between the responses of the railway station and intercity bus station users Travel habits Figure 1 shows the trip status of travellers, at the time they were interviewed. In the railway station, the majority of users (69.5%) were at the beginning of their trip, the 29,5% in the end, and the rest 1% of them were transferring. Similar results were indicated in the intercity bus station, where the 56.2% of users were interviewed before they started their trip, the 29.5% in the end of their route and the rest 14.3% were transferring (Figure 1). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 122

129 SESSION: PUBLIC TRANSPORT PLANNING Figure 1. Travellers trip status Regarding the trip purpose, leisure-visit and work were the main purposes for all of the trips recorded in the two case studies (Figure 2). Particularly, in the railway station, the 61.9% of users were travelling for leisure or visit, the 16.2% for work, the 13.3% for other reasons, and the rest 8.6% for education purposes. In the case of the intercity bus station, the majority of users (41.9%) were travelling for leisure or visit, the 26.7% for work, the 17.1% for education and the rest 14.3% for other reasons. Figure 2. Travellers trip purpose Since both stations cover intercity travel needs, it was expected that the interchange travellers participating in the survey, would not be habitual users. Indeed, the majority of respondents in the railway station travel few times a month (38%) or less frequently (37%), while the 12% use the station once or twice a week, the 9% 3-4 times a week, and only the 4% daily. Similar results were observed in the intercity bus station, where the 30% of users travel few times a month, while the majority of them (51%) less frequently. The 8% of respondents use the bus station once or twice a week, the 7% daily and the 4% 3-4 times a week (Figure 3). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 123

130 SESSION: PUBLIC TRANSPORT PLANNING Figure 3. Frequency using each station 3.2. Users satisfaction level: Cross-site comparison A cross-site comparison was conducted in order to identify the differences in travellers satisfaction across the two interchanges, and the results are presented in Table 2, in which the average rating for each item per station is included, as well as the p-value and the test parameters relation. Each user stated his/her level of satisfaction by rating from 1 (strongly dissatisfied) to 5 (strongly satisfied). Results showed that the average rating of the overall satisfaction of users was slightly higher in the railway station (3.19) in comparison to the intercity bus station (3.15). However this difference was not statistically significant (p-value=0.6). The provision of information in the two stations was assessed in two dimensions. The first one regarded travel information and four items were used to evaluate travellers satisfaction. It was observed that the rating of these items was higher in the intercity bus station compared to the railway station. However, statistically significant difference was indicated only in the accuracy and reliability of travel information displays, and the relevant scores were 3.04 in the railway station and 3.41 in the intercity bus station (p-value=0.005). The second dimension referred to the wayfinding information, and in this case, results showed that travellers were more satisfied in the railway station in terms of signposting to different facilities and services (3.14 vs. 3.07, p-value=0.368), and information and assistance provided by staff (3.30 vs. 3.03, p-value=0.05), while users stated their satisfaction more strongly in the intercity bus station when they evaluated signposting to transfer between transport modes (3.10 vs. 2.89, p- value=0.17). The ease of accessing the interchange was the indicator for the evaluation of access in the two stations, and results showed in this case that travellers were more satisfied in the intercity bus station (3.71) in comparison to the railway station (3.05). This difference was statistically significant (p-value=0). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 124

131 SESSION: PUBLIC TRANSPORT PLANNING In terms of time and movement, five items were used for the assessment of the level of satisfaction of travellers, and it was observed that the intercity bus station was higher rated in all cases. In fact, statistically significant differences were indicated in four out of five items, while the item that received the highest rating was the distances between transport modes (3.71 vs. 3.05, p-value=0). Table 2. Average rating and summary of test results for comparison between the two case studies Items Average rating p-value Test parameters R B R vs. B relation Travel information Availability and ease of travel information at the interchange r R < r B Availability of travel information before the trip r R < r B Accuracy and reliability of travel information displays * r R < r B Ticket purchase (ticket offices, automatic machines, etc.) r R < r B Wayfinding information Signposting to different facilities and services r R > r B Signposting to transfer between transport modes r R < r B Information and assistance provided by staff * r R > r B Access Ease of access to the interchange * r R < r B Time and movement Distances between transport modes * r R < r B Coordination between different transport operators or transport services r R < r B Use of time at the interchange * r R < r B Distance between the facilities and services * r R < r B Ease of movement due to number of people inside the interchange * r R < r B Image and attractiveness Surrounding area * r R < r B Internal design of the interchange * r R > r B External design of the interchange * r R > r B Comfort and convenience General cleanliness of the interchange r R > r B Temperature, shelter from air and wind, etc r R > r B General level of noise of the interchange * r R > r B Air quality, pollution, i.e. from vehicles * r R > r B Number and variety of shops r R < r B Number and variety of coffee shops and restaurants * r R < r B Availability of cash machines * r R < r B Availability of seating * r R < r B Availability of telephone signal and Wi-Fi * r R < r B General comfort r R < r B Emergency situations Information to improve the sense of security r R > r B Signposting to emergency exits r R < r B Location of emergency exits in case of fire * r R < r B Overall satisfaction Level of services provided by the interchange r R > r B R: Railway, B:Intercity buses; r: mean rank * Significant at confidence level 95% and confidence interval 5% 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 125

132 SESSION: PUBLIC TRANSPORT PLANNING Statistically significant differences between the two stations were also indicated when travellers rated image and attractiveness, and results showed that the internal design and the external design were more satisfactory in the case of the railway station, 3.88 vs. 2.86, p-value=0 and 3.55 vs. 2.78, p-value=0, respectively. On the other hand, the surrounding area was higher rated by travellers in the case of the intercity bus station (3.11) in comparison to the railway (2.78) (p-value=0.02). Ten items were used for the evaluation of comfort and convenience of the two stations. Findings showed that in four out of ten items, i.e. general cleanliness, temperature, shelter from air and wind, general level of noise, and air quality and pollution, the users of the railway station were more satisfied compared to the intercity bus station. Statistically significant differences were observed in the general level of noise (3.3 vs. 2.87, p-value=0.002) and in the air quality and pollution (3.15 vs. 2.65, p-value=0). As it was expected, since the intercity bus station accommodates more facilities compared to the railway station, users rated higher the bus station when they were asked to evaluate the number and variety of shops (2.37 vs. 2.12, p-value=0.26), the number and variety of shops and restaurants (2.32 vs. 1.93, p-value=0.009), the availability of cash machines (2.77 vs. 2.28, p-value=0.002), the availability of seating (2.85 vs. 2.36, p- value=0.001), the availability of telephone signal and Wi-Fi (3.36 vs. 2.08, p-value=0), and the general comfort (3.04 vs. 2.76, p-value=0.11). The last category that was evaluated was emergency situations, and in this case three items were used. The first item regarded the information for the improvement of the sense of security, which was higher rated in the railway station (2.81 vs. 2.79, p- value=0.56), the second one referred to the signposting to emergency exits, which was more satisfactory for the users of the intercity bus station (2.71 vs. 2.65, p-value=0.83), and the last one regarded the location of emergency exits in case of fire, which received higher score from the intercity bus station users (2.77 vs. 2.26, p-value=0.002). 4. CONCLUSIONS The scope of this paper was to capture the viewpoint and preferences of travellers on different aspects that define sustainable urban transport interchanges. For this reason, an on-site face to face questionnaire survey was conducted at the intercity bus station and the railway station in Volos, Greece. Useful information was gathered about travellers habits, preferences and satisfaction. In total thirty items were rated by respondents, and one of the main conclusions arising from users overall satisfaction ratings was that, although the intercity bus station had a higher score in the majority of the indicators assessed, the railway station seems to satisfy slightly more the users. This can be explained by the fact that there are specific characteristics of the station and the relevant services provided, which strongly affect travellers opinion and consequently modal choice. Among the items that were highest evaluated were the external design of the railway station, the distances between transport modes in the case of the intercity bus station, and the easiness of ticket purchase also in the bus station. On the other hand, the number and variety of coffee shops, restaurants and shops in the railway station were the items that received the lowest rating. In addition, findings showed that travellers require that 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 126

133 SESSION: PUBLIC TRANSPORT PLANNING both interchanges should increase seating, improve comfort and convenience, and lastly, improve telephone signal and Wi-Fi. Next steps of this research is to look into the viewpoint and perspective of local authorities and operators, and consequently make an effort to open ground to an integrated design and operation process, able to meet the expectations and real needs of users and decision makers, towards sustainable urban transport interchange design. Acknowledgments The present paper is based on the methodology that was developed in the framework of the City-HUB project ( which is co-funded by the European Commission within the 7 th Framework Programme. The authors would like to thank both the consortium of the project and the European Commission. Also, the authors would like to thank the Laboratory of Transportation Engineering of the University of Thessaly for permission to use in this study data collected during the questionnaire surveys in the two case studies. References Allen, H. Harmer, C. Millard, K., Palmer, D., Monzon, A., Hernandez, S., DiCiommo, F., de Oña, R., Nathanail, E. and Adamos, G., Guide for Smart and Efficient Design. Deliverable D3.2, City-HUB Project. Di Ciommo, F., L accessibilité: l enjeu prioritaire de la nouvelle politique des transports publics à Naples, in Bernard Jouve, Les politiques de déplacements urbains en Europe, L Harmattan, pp DTO, Dublin Transportation Office, Advice note on public transport interchange. Available through: [accessed on May 2014]. Grotenhuis, J.W., W.W. Bart and P. Rietveld (2007). The desired quality of integrated multimodal travel information in public transport: Customer needs for time and effort saving. Transport Policy, Vol. 14, pp KITE, Sammer, G., Stark, J., Uhlmann, T. (2009). KITE project, Deliverable D14 Guidelines for seamless intermodal interchanges. Pitsiava-Latinopoulou, M.& Iordanopoulos, P., Intermodal Passengers Terminals: Design standards for better level of service. Procedia Social and Behavioral Sciences 48 (2012) Siegel, S., and Castellan, J. (1988). Non parametric statistics for the behavioral sciences, 2nd ed. McGraw - Hill, New York. TOOLQIT, Tools for the assessment of level and quality of services across different transport market segments, Institute of Transport Economics for EC DG TREN Vuchic, V.R. (2005) Urban Transit: Operations, Planning and Economics. John Wiley & Sons, New Jersey. United States of America. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 127

134 SESSION: MODELLING PUBLIC TRANSPORT

135 SESSION: MODELLING PUBLIC TRANSPORT Applications of Traffic Distribution using the Software VISUM and more Environmental Capacity Utilization Stella Ch. Strataki 1, Christos Taxiltaris 2 1 Department of Transportation and Hydraulic Engineering, School of Rural and Surveying Engineering, Aristotle University of Thessaloniki, GREECE ( sstrataki@gmail.com) 2 Department of Transportation and Hydraulic Engineering, School of Rural and Surveying Engineering, Aristotle University of Thessaloniki, GREECE ( chtaxilt@auth.gr) Abstract The paper focuses on the examination of the level of service provided, in relation to capacity and demand and its contribution in the decision making process. The macroscopic model VISUM is applied on two road networks, one test network, Sioux Falls, and one real, the road network of Eastern Thessaloniki. With the use of the specialized software, the planning of an urban transportation network under a macroscopic scale is examined in order to define the trips distribution. A sensitivity analysis is performed on the changes of demand and capacity and their impact to the level of service is investigated. The One-factor-at-time (OFAT) sensitivity analysis method was selected and the results of the application are presented. The transport network and infrastructure in Greece should be improved in order to achieve sustainable development. Therefore, the scientific results of researches should be taken into account. The different contribution of the capacity and demand is noted. The implementation of environmentally friendly solutions is proposed as well. Finally, the ultimate aim is to improve the quality of peoples everyday life. Keywords: environmentally friendly functions, Network of Eastern Thessaloniki, sensitivity analysis One-factor-at-time (OFAT), software VISUM, Test Network Sioux Falls, traditional four-step transportation forecasting model. 1. INTRODUCTION Transport expresses the daily need of people to move in an effort to conduct their everyday activities. It is therefore considered more of a service, than a commodity. The purpose of transportation planning is to satisfy the demand for transport of people and goods, with different purposes, during different periods of time, with various means of transport, in a given transport system, which has a specific functional capacity, infrastructure, and management. (Cohen, S. 1990) Transportation planning is a complex area, which aims at proposing interventions or extensions of the transport system, in order to ensure the satisfaction of estimated future demand, taking into account the available resources. (Ortuzar, J. de D. & Willumsen, L.G. 1995, Williams, H.C.D.L. 1977). Multiple factors are involved in this process. The relationships and interactions between them are various and multifactorial. However, the latest technological development of programming tools, through computer systems, contributes to the improvement of models and makes their use more efficient and effective. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 128

136 SESSION: MODELLING PUBLIC TRANSPORT The development of models aims at a better representation of the current situation and a reliable prediction, according to the impacts of the proposed interventions. Many variables are co-calculated, with better distribution of their prices, based on assessment of their evolution in time. Thus, the transportation models are considered as a useful tool for predicting future demand and evaluating alternative options. Furthermore, it is widely accepted that transport projects require large investments. The cost of infrastructure and maintenance is high. Choosing the optimal solution can prove to be a demanding task. Apart from the cost-benefit analysis, it should convince the stakeholders (state, private investors, users and the local community) for its accuracy. On the contrary models can be misleading, inaccurate and difficult to understand. So it s better to state both approaches and far and foremost refer to a scientific source for such a significant conclusion. The outline of this paper is as follows. It is presented the transport planning analysis, the sensitivity analysis and we describe the road networks in Sect. 2. Results concerning the ratio of traffic volume to capacity and the ratio of the free flow travel time (to) to the loaded network travel time (tcur) in Sect. 3 are provided. Section 3 contains figures with charts and relevant mathematical expressions. Concluding remarks and suggestions are provided in Sect METHODOLOGY The traditional four-step transport model is applied: a) trip generation, b) trip distribution in the network, c) modal choice d) traffic assignment. The first two steps deal with models attempting to represent the travel demand in the study area, while the third step, modal choice, refers to the distribution of travel demand among modes. The fourth step deals with the assignment of travel demand, based on the supply characteristics of a given network (e.g. capacity). (Ortuzar, J. de D. & Willumsen, L.G. 1995). The macroscopic model VISUM was used. Software Planning Transport «VISUM» has developed and distributed by the company PTV AG. The VISUM incorporates the traditional four-step transportation forecasting model. It is software which examines a macroscopic transport system and attempts the trip distribution in the network paths Transport planning analysis The transport planning analysis was performed on two road networks, one test network, Sioux Falls, and one real, the road network of Eastern Thessaloniki. The free flow speed is considered equal to 60 km/h in the Sioux Falls network and equal to 90km/h in the network of Thessaloniki. The impedance transit route is considered 100 * tcur, where tcur (sec), is an indicator of travel time in the loaded network. The characteristics of the transport system includes the offer for travel by public and private means, as defined by the system capacity, and the demand for travel. Macroscopic traffic models are based on the fundamental relationships between traffic speed, traffic flow and traffic density. The traffic speed is defined as a decreasing function of density. While the network is loaded, the distance between vehicles 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 129

137 SESSION: MODELLING PUBLIC TRANSPORT decreases and consequently the speed is reduced. As the model deals with groups of vehicles, all individual vehicles are considered to move at the same speed for a given density. This method improves the computational performance, but reduces the level of simulation s detail Sensitivity analysis For the sensitivity analysis, the one-factor-at-time (OFAT) method was used. (Cruz, J. B. and Perkins, W.R., 1964, Fassò A. 2007, Fassò A., Perri P.F. 2002, Taleb, N. N., 2007). The OFAT method was applied on both networks in order to determine the sensitivity of results, as the factors of capacity and demand are changed. The model was solved twice for each network, one with changes in demand and another with changes in capacity. Despite the fact that this method does not take into account the effect of the simultaneous change of several sets of data in the results, interesting conclusions are obtained. The network of the study area was modeled by introducing real data of capacity and demand. The route assignment during the morning rush hour was thoroughly investigated The road networks Traffic was assigned in the Sioux Falls network based on the Least User Cost Equilibrium (LUCE) algorithm. The use of this algorithm is appropriate when demand is amended. According to this model, the probability Po for a traveler to select a route Χo between k paths Χk, each of which is characterized by the corresponding relative utility uk, is assigned to the following accounts expression (Taxiltaris X., 1994): While, for the network of Eastern Thessaloniki, the model Intersection Capacity Analysis ICA, was used. This algorithm is considered as ideal for the network of Eastern Thessaloniki as it is used when there are signalized junctions, and is based on an iterative process to optimize the result with respect to the traveler, using delay-load functions. The Test Network Sioux Falls consists of: 24 Nodes 76 Links 24 Zones 48 Centroids The real Network of Eastern Thessaloniki consists of: 1145 nodes, of which: the 94 are signalized, the 489 are two way stop, the 528 are nodes only with signaling and the 34 are contributions of roads at roundabouts 3432 Links 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 130

138 SESSION: MODELLING PUBLIC TRANSPORT 2.4. Visum The software has the ability to identify the matrix Origin Destination and to estimate the available paths for each transport mean and for each group population, considering that its members behave with great homogeneity. VISUM is a software system that enables integration of all public and private transport means in a single model. It manages data related to transportation and planned transport systems. In addition, it can create and utilize complex relations between one or more transport systems. Moreover, it can identify the effects of existing, or planned capacity of private and public transport means. The description of the transport system includes the ridership of traveling by public and private means, as defined by the system capacity, and the travel demand. Regarding the research field, the effects of applying the static algorithm, which is provided by VISUM, were studied in an effort to find which model depicts in the best way the traffic conditions of the study area. 3. RESULTS 3.1. The ratio of traffic volume to capacity The relationships between the changes of the ratio of traffic volume to capacity were investigated, as demand and capacity are being changed. The relationship which links the change in traffic volume with the change in demand is linear, as shown in Figure 1. Subsequently, there is a proportional relationship between them. The congestion (Volume/Capacity = 1) is observed at 2.70 of the given demand. The index of sensitivity in traffic volume against changes in demand is very high. Figure 1. Changing the variable Volume: Volume/Capacity increasing function - Thessaloniki In contrast, the relationship which links the change in traffic volume with the change in capacity is hyperbolic, as indicated in Figure 2. The congestion (V/C = 1) is at 0.35 of the given capacity. The index of sensitivity in traffic volume against changes in capacity is very low. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 131

139 SESSION: MODELLING PUBLIC TRANSPORT Figure 2. Changing the variable Capacity: Volume/Capacity decreasing function Thessaloniki The graph Volume/Capacity when capacity is kept constant and demand varies, is directly related to the graph Volume/Capacity when demand is kept constant and capacity varies. In the first case, the resulting function is linear by varying the numerator, namely the demand, while the denominator, namely the supply, remains constant. Vice versa, in the second case, the function obtained is hyperbolic as the numerator remains constant, namely the demand, and the denominator is changed, namely the capacity. The mathematical expression of the above is: According to the above observations, the change in the level of service, changing demand and maintaining steady capacity, is described by a linear function. On the contrary, the change of capacity, keeping demand stable, affects to a small extent the level of service. Therefore, according to the above graphs, it is observed that if the current capacity reduces by 40%, the level of service remains the same. That means that the level of service with LOS = A with price ration = 0,348, after reducing the capacity remains the same, LOS = A now with price ratio = 0,431. While, if the original capacity reduces by 50% the level of service results to become, LOS = C with rate ratio = 0,71. This sudden change in the level of service between -40% to -50% of the original capacity shows that between these values there are optimal conditions. Optimal conditions are those where any increase in supply does not lead to any increase in traffic, which the specific road section serves. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 132

140 SESSION: MODELLING PUBLIC TRANSPORT 3.2. The ratio of the free flow travel time to loaded network travel time The relationships between the changes of the ratio of the free flow travel time (to) to the loaded network travel time (tcur) were investigated, as demand and capacity are changing. As journey time and speed are correlated variables similar relationships occur on the change of the ratio of the free flow speed (vo) to the loaded network speed (vcur). Figure 3. Changing the variable Volume: t0 / tcur decreasing function - Thessaloniki The graph to/tcur, when capacity is kept constant and demand varies, can be expressed as a decreasing function as demand increases, as shown in Figure 3. The slope of the curve is smoother on the real network. The increase of demand has great influence on journey time and speed. Since the free-flow speed is constant, the t0 for each path remains stable. Therefore, the sensitivity index of the denominator (tcur), when demand is changing, is high. Figure 4. Changing the variable Capacity: t0 / tcur increasing function - Thessaloniki Thus, the graph to/tcur, when demand is kept constant and capacity varies, can be expressed as an ascending parabolic function with increasing capacity, Figure 4. The slope of the curve is now softer on the theoretical network. The increase of the capacity 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 133

141 SESSION: MODELLING PUBLIC TRANSPORT has very little effect on journey time and speed. The sensitivity index of these variables to changes in capacity is very small. 4. CONCLUSIONS According to the above observations we conclude that the change in the level of service, as the demand is changing, maintaining steady occupancy, is influenced at a great point by varying demand with a significant effect on loaded network travel time. Therefore, the changes in demand should be checked carefully. It is recommended to promote the use of public transport, possibly with special rates for use during rush hours. In contrast, the change in capacity, keeping demand stable, affects to a limited extent the level of service and the journey time and speed. The increase in supply does not significantly improve the operation of a road network. On the roads of Eastern Thessaloniki it has been concluded, that by decreasing the current capacity by 40%, the level of service remains the same. This means that there is the possibility of restriction of the motorized traffic in some links of Thessaloniki s road network. These sections should be defined certainly after the appropriate transportation study. In that way, part of the road section could be used for more environmentally friendly functions. According to the study s results, creation of walkways, increase of the city green areas and expansion of the existing cycling network are recommended. The use of the above conclusions from the city of Thessaloniki would result in improving citizens everyday life, upgrading their standard of living, enhancing the sense of neighborhood, stimulating the local market, without additional journey time to the residents trips. References Cohen, S. 1990, Ingénierie du trafic routier, Ponts et Chaussées Cruz, J. B. and Perkins, W.R., 1964, A New Approach to the Sensitivity Problem in Multivariable Feedback System Design, IEEE TAC, Vol. 9, Fassò A Statistical sensitivity analysis and water quality. In Wymer L. Ed, Statistical Framework for Water Quality Criteria and Monitoring. Wiley, New York. Fassò A., Perri P.F Sensitivity Analysis. In Abdel H. El-Shaarawi and Walter W. Piegorsch (eds) Encyclopedia of Environmetrics, Volume 4, pp , Wiley. Ortuzar J. de D., Willumsen L.G.,1995, «Modelling Transport», Wiley Patriksson M.,2003 The traffic Assignment problem Models and methods, Sweden Sheffi, Y., Urban Transportation Networks- Equilibrium Analysis with Mathematical Programming methods, MIT Taleb, N. N., 2007 The Black Swan: The Impact of the Highly Improbable, Random House Williams, H.C.D.L Environment and planning 9A(3), Willson, A.G Urban and regional models in Geography and Planning, Clarendon Press: Oxford 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 134

142 SESSION: MODELLING PUBLIC TRANSPORT Foundation and Implementation of a New Composite Accessibility Measure in Individual Level F. Moustou 1 *, Y. Photis 2, V. Vasiloglou 3 1 Engineer of Planning and Regional Development, Phd 2 Associate Professor, National Technical University of Athens 3 Civil Engineer, Phd * Corresponding author: moufotini@yahoo.gr, Tel Abstract The present paper proposes and implements a new composite accessibility measure. Knowing that the zonal approach can hide or even change the accessibility levels in certain areas, the accessibility is proposed to be determined in individual level. The new accessibility measure encompasses all activities of the individual in a day. In the proposed measure, individual activity space, temporal and travel constraints are combined successfully. The proposed measure is applied for the determination of the university students accessibility in the city of Volos. Particularly, data of students mobility and all outdoor activities in a day are imported in the new measure aiming to determine the accessibility of every student during a typical day. Also, the GIS functions contribute significantly to the application of this new measure. Keywords: Accessibility Measures; Activity Space; Temporal and Travel Constraints, GIS functions. 1. INTRODUCTION Accessibility is being investigated in many scientific fields, such as geography, transportation research and urban planning. Accessibility is the key element in analyzing the efficiency of transportation systems, in predicting travel demand, in programming transportation investments, and in evaluating planning policies in the urban transportation planning process (Koening 1980; Handy and Niemeier 1997; O'Sullivan et al. 2000). Accessibility has been defined in several different ways (Hansen 1959; Hägerstrand 1970; Ingram 1971; Dalvi and Martin 1976; Burn 1979; Ben-Akiva and Lerman 1979; Tagore and Sikdar 1995; Geurs and Van Wee 2004). Accessibility measures are used to evaluate the ease with which the individual, population or community can have access to one or more services at home or other locations, using different ways of movement. Various ways for categorizing the accessibility measures have been developed (Giannopoulos and Boulougaris, 1989; Geurs and van Wee, 2004; Dong et al., 2006; Joly, 1999). In this research, the accessibility measures are categorized into two large groups of traditional measures (the simple distance measure, the cumulative measure, the gravity measures) and modern measures (the activity-based measures, the space - time accessibility measures). The degree of difficulty of the computation, the simplicity and usability of the measures are the main factors of the above distinction. Specifically, the traditional measures are simpler, more usable and they need less computation than the modern measures. Contrary to the modern measures, traditional measures have been criticized for problems related to functionality. Also, traditional measures have presented lack of 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 135

143 SESSION: MODELLING PUBLIC TRANSPORT temporal constraints and they are unable to capture the complex movement behaviour (Handy and Niemeier 1997; Kwan and Weber 2003). It was observed that the variety of definitions and applications of accessibility led to different spatial levels of analysis. In particular, the accessibility analysis started with the accessibility at zonal level. The zonal accessibility analysis has the disadvantage of being influenced negatively by the size and the form of zones, and by the data aggregation. For this reason, the accessibility analysis in zonal level was considered a relatively insufficient way of presentation of accessibility and thus the accessibility analysis in individual level was imported (Ben - Akiva and Lerman 1979; Kwan 1998; Kwan 1999; Kwan and Weber 2008). This paper studies the accessibility in individual level. Geographic information systems are suitable for analysis and measurement of accessibility because they have the ability to display the attributes of transport systems and activities. The basic functions that are used by a standard geographic information system are the preparation of a database and the algorithm that measures the physical accessibility (Arentze et al. 1994; Kwan 1998; O ' Sullivan et al. 2000; Shen 1998). So, GIS functions are used for the application of a new accessibility measure. In the present paper, a new accessibility measure is created. The proposed accessibility measure is influenced not only by usual factors such as the places of activities and the travel time, but also by the individual activity space and temporal constraints as the participation in the daily activities. The incorporation of the individual activity space in an accessibility measure imposes the effect of the daily spatial movement of a human, which shows the individual's perception of accessibility. The activity space with the proposed form is imported to the proposed accessibility measure (Moustou 2013). The suggested measure is applied to a certain social team in the city of Volos. Particularly, the university s students of city of Volos are selected because they represent the most active social team in the city and they have a continuously increasing mobility in the city of Volos. In this research, information is collected with regard to the location of students residence, the individual outdoor activities (recording the individual activity schedule) and the daily travel from their residence to their daily outdoor activities. This article is composed of four units. In the second unit, the proposed measure is developed. In the third unit, the proposed measure is implemented. In the fourth unit, conclusions are presented. 2. THE CREATION OF THE PROPOSED MEASURE The suggested measure is an outcome of the triple relation (spatial, temporal and travel components) which were imported by Burns (1979) and Odoki (1992). Also, the proposed measure uses the base rules of the composite measures of the Bhat et al. (2002), which use aggregation techniques (Moustou 2013). Following the general structure of the composite measures of Bhat et al (2002), the suggested measure is founded in two steps: 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 136

144 SESSION: MODELLING PUBLIC TRANSPORT 1. In the first step, a new accessibility measure has to be created for each activity (fixed and flexible) of the daily individual activity schedule. 2. In the second step, the composite accessibility measure has to be created for all daily individual activities. An important factor for the construction of the proposed measure is the activity space. In particular, activity space represents the spatial movement as part of a human daily experience. This experience of place shows the relationship of distance and dispersal of individual activities and so it represents the individual's perception of accessibility (Golledge and Stimson 1987). Activity space is a measure of individual spatial behaviour that theoretically accounts for these individual and environmental differences and offers an alternative approach for studying geographic accessibility. Activity space has been formulated as a standard deviational ellipse (SDE), a road network buffer and a polygon of standard or relative travel (Sherman et al. 2005). In the present research, it is selected the suitable formulation for the depiction and the determination of individual activity space First step A new accessibility measure from origin to different activity of the daily activity schedule has to be created. The present investigation proposes to use a modified gravity measure, combining two parameters. The first parameter is the spatial element (which is named as SEij) and the second parameter is the travel element (which is named as TrEij) (Eq. 1) (Moustou 2013). Aij SE ij TrE ij The spatial element (SEij) is a parameter which shows the spatial importance of each daily individual activity. For the determination of the proposed spatial element of each activity j (SEij) is incorporated the effect of the factor, which shows the importance per activity j (ωij) for every individual i weighted with the individual activity space (Actspacei). The importance of activity j for the individual i (ωij) is defined as a factor that is related with the adjacency of activity j to the residence of the individual i. This factor is normalised in a scale 0-1 (the zero mean small importance, while values near the one declare high importance of activity) (Eq. 2) (Moustou 2013). (1) max j dist ij dist ij ij max dist min dist j ij j ij (2) distij: is the distance from residence i to activity j, maxj distij: is the maximum distance from residence i to the total individual activities minj distij: is the minimum distance from residence i to the total individual activities Also, the weighed factor of activity space of each individual (Actspacei) is formed by the place of residence and the daily individual activities. The activity space has been occasionally measured as deviational ellipses, as a road network buffer and as a polygon with predetermined or relative time of traveling using GIS functions. In the present research, the convex polygon is selected for the depiction of individual activity space. The area of the convex polygon that is shaped by the place of residence and individual activities constitutes the basic parameter of determination of individual activity space. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 137

145 SESSION: MODELLING PUBLIC TRANSPORT The calculated area of each individual activity space is normalised and expressed by a factor that takes values between 0 and 1 (Eq. 3) (Moustou 2013). max Ar i Ar Actspace i i max Ar i min Ar i (3) Ari: is the area of individual activity space i, maxari: is the area of individual activity space i with the bigger polygon in the set of individuals and minari: is the area of individual activity space i with the smaller polygon in the set of individuals. Travel element (TrEij) is a function, which incorporates the travel cost. The most representative function for the travel behaviour is the negative exponential function (Handy 1992; Levinson 1998; Shen 1998; Kwan 1998). A major drawback of the exponential function of travel cost is that accessibility falls significantly as an individual moves away from the center of city. This sudden change in levels of accessibility dealt with the Gaussian function (Iacono et al. 2008; Litman 2008; Lee and Goulias 1997; Guy 1983; Ingram 1971). So, Travel Element (TrEij) is proposed to be estimated according to the equation below (gaussian type) (Eq. 4): TrE ij f tij ( 2t ( cij ) e (4) t: is the mean travel time from the summary of daily individual activities tij: is the travel time from origin of individual i to destination-activities j The equation below (Eq. 5) depicts the proposed measure of the accessibility of individual i from its place of residence to each activity-destination j (Moustou 2013): 2 2 ) t 2 ij 2t 2 Aij SE ij f ( c ij ) Actspace i e ij (5) (SEij): The parameter of spatial importance is identified with the created term of Spatial Element per daily activity j Actspacei: factor of individual activity space ωij: factor of importance per activity (TrEij=f(cij)) and the impedance function is the proposed function of Travel Element per daily activity j Second final step The final composite individual accessibility measure is defined as the summation of individual ability to have an access to all activities of its daily personal schedule weighted by the factor of temporal element of each activity j (αij). The proposed temporal index will be incorporated in the final measure as the attendance of individual in each activity in the day. More analytically, attendance of an individual in each activity in the day is defined as the total daily hours (hij) of the individual in the activity at the day. Consequently, temporal element (TEij) is determined according to the 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 138

146 SESSION: MODELLING PUBLIC TRANSPORT equation below (Eq. 6) (Moustou 2013): TE ij a ij hij (maxh ij ) (6) Where hij: is the total individual daily hours for every activity maxhij: is the maximum time attendance of the total individual activities The final form of proposed individual accessibility measure (ACCi) is the summation of different daily activities (fixed and flexible) accessibility weighting exponentially by the factor of temporal element. Specifically, the proposed compose individual accessibility measure (ACCi) is determined according to the equation below (Eq. 7) (Moustou 2013): ACC i j ( e j a ij e A a ij ) Actspacei ij j ( e j h ij ij max h e h ij max h ij e ij 2 tij 2t In the following unit the proposed accessibility measure is implemented. 3. IMPLEMENTATION OF THE PROPOSED MEASURE As it is mentioned, the suggested measure is applied to a certain social team in the city of Volos. Particularly, the university s students of city of Volos are selected because they represent the most active social team in the city and they have a continuously increasing mobility in the city of Volos. In the following sections, it is presented the methodology, the data and the results of the application of the new individual accessibility measure for the university s students Methodology of the application The calculation of the proposed individual accessibility of every student is fulfilled in three steps: 2 ) (7) In the first step, the necessary data for the group of students are collected. The collection of information is done by using questionnaires. There are four parts of the questionnaires: o The first part of the questionnaire consists general questions for the target team of the students (such as gender, year of introduction, school attendance etc). o The second part of the questionnaire aims to determine both the precise location of the student's residence in the city of Volos and its features. o The third part is associated with the daily student destinations activities. In particular, a typical day in the week is chosen and a daily activity diary is built. o The fourth part includes questions related to daily movements of student activities. Specifically, the daily way of student travelling and the time they spend to reach out to their activities are recorded. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 139

147 SESSION: MODELLING PUBLIC TRANSPORT In the second step, the collected data are processed and the proposed measure is applied. GIS techniques and statistical analysis programs are used. For the determination of individual accessibility are used the following GIS functions: o The data preparation: The places of origins and destinations, the characteristics of activities and the characteristics of travel are required for the data preparation. The origins and destinations are portrayed in the road network with the function of geocoding. o The algorithm of accessibility: The algorithm that measures the natural accessibility between origin and destination finds the nearest path on the network. The GIS function that is used is the Network analysis. Finally, in the third and last step, the results of the analysis are displayed and the proposed measure is evaluated. The GIS functions are used for the depiction and the evaluation of the results. o At first, it is depicted the results of the implementation of the proposed measure for each activity and for the total activities of the daily schedule on the point entity of the student residence. The estimated accessibility values are converted to the whole area of the city. However, this conversion can hide the real level of accessibility, so a way to avoid this problem is to be selected the proper scale of the generalisation. A proper entity and scale for the depiction of accessibility is the raster entity with the size of 100*100 meter per pixel (Wagner and Wegener 2007; Wagner and Wegener 2011; Cervero and Kockelman 1997; Munshi, et.al. 2009). So, the Kriging function is selected by the GIS functions. This technique products raster entities and it is used to predict the variance of the proposed composite accessibility to the whole area of the city Data of the reasearch The data which was used for the determination of individual accessibility are: The location of residences, The daily individual schedule of out of home activities The location of out of home activities, The travel to activities, including the time and the mode of travelling. The locations of origins are the first important elements for the determination of accessibility. Origins are defined as the students residences. The locations of destinations are the second important elements. Destinations are defined as the students out of home activities. According to the daily student activity schedule, students activities are divided in four categories: University departments Market Entertainment Public Services According to the frequency of visiting of daily activities, the activities are divided in two kinds of activities (the fixed and the flexible activities). Specifically, the university 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 140

148 SESSION: MODELLING PUBLIC TRANSPORT departments are the fixed activity and the markets, the entertainment and the public services are the flexible activities. The research was conducted on the basic areas of the University (the faculty of Engineering, the faculty of Humanities and the faculty of Agriculture). The faculties of the university are scattered sparsely to the whole area of the city. The flexible activities, such as shops, public services and entertainment, are mostly located to the central area of the city and the fixed activity of the university faculties is located to the different areas (for example the faculty of Engineering and the faculty of Agriculture are located to the distant areas and the faculty of Humanities is located to the centre of the city). The sample was set as the 5% of the total population of the students. Analyzing the students sample answers were found that: the 55,5% of the interviewed students were women and the 44.5% were men, the 80% of the sample ranged between 19 and 24 years old, the 50% of the students resided in the areas of St. Nicholas and Metamorphosis, their residences have high accessibility levels at stops of public transports and the students can caught more than one of a bus stop in a walking distance of 400 meter from their residences, the activity with the biggest participation was the school (5,3 hours per day), second was the entertainment activity (3 hours per day), third was the shopping activity (0,8 hours per day) and the public services had the smallest participation in the day of the students (0,13 hours per day), the 43% of movements of students sample to their activities were on foot, the 16% of movements were by bus, the 20% of movements were by car and the 21% of movements were by bicycle, the average travel time of students in the total daily activities ranged among 7 minutes for the destination of school, 5 minutes for the market, 6 minutes for the entertainment and 8 minutes for the public services Results of the implementation of the proposed measure to university students The GIS functions play significant role for the depiction of the proposed accessibility. The following figures (1-6) illustrate the accessibility values of the proposed measure, which are normalized to a scale of 0-1. Zero corresponds to the minimum value of accessibility and one to the maximum accessibility value. Particularly, accessibility values are separated in four quadrants: High accessibility level (1-0.75) Satisfactory accessibility level ( ) Low accessibility level, ( ) Very low-limited accessibility (0.24-0). In particular, students accessibility from the place of residence to school is depicted to the Figure 1. The high and the satisfied accessibility levels are scattered sparsely in different locations in the city of Volos due to the fact that many faculties of the University are scattered in distant areas of the city of Volos, contrary to the locations 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 141

149 SESSION: MODELLING PUBLIC TRANSPORT of the students residences. So, students accessibility from their residences to the fixed activity of school is limited. Students accessibility from the place of residence to the place of entertainment is represented to the Figure 2. In addition, students accessibility from the place of residence to the market activity is depicted in the Figure 3. Also, students accessibility from the place of residence to public services is represented in Figure 4. The high and the satisfied accessibility levels to the flexible activities cover a large part of the city of Volos and extend from the centre to the interior of the city. The market is the destination with the best proximity from the place of residences. The following table 1 depicts the percentage of the different level of student accessibility when the proposed accessibility measure for each activity (Eq. 5, Aij) and for all activities of the day (Eq. 7, ACCi) is implemented. Table 1. Student accessibility percentage for any level of the proposed accessibility measure Accessibility to each student activity Levels of Fixed activity Flexible activities Composite accessibility University Public individual Entertainment Market departments services accessibility High 7% 34% 17% 48% 22% Satisfactory 10% 26% 28% 31% 51% Low 14% 15% 25% 11% 22% Very low 69% 25% 30% 10% 5% Total 100% 100% 100% 100% 100% Figure 1. Student accessibility to university departments 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 142

150 SESSION: MODELLING PUBLIC TRANSPORT Figure 2. Student accessibility to entertainment Figure 3. Student accessibility to market Figure 4. Student accessibility to public services 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 143

151 SESSION: MODELLING PUBLIC TRANSPORT The final proposed composite individual accessibility measure, which encompasses all the daily outdoor activities of every student (Eq. 7, ACCi) is depicted to the Figure 5. High accessibility level has the 22% of the sample of students, satisfactory accessibility level has the 51%, low level has the 22% and very low-restricted accessibility level has the 5% of students sample. Consequently, university students have very good access from their location of residence to their daily outdoor activities. Students accessibility in daily activities outside residence in the city of Volos is high in central areas and decreases as the distance from the center of the town is increased. So, students who choose to stay in central areas of the city of Volos have better accessibility levels to daily activities than students who reside in distant areas. Figure 5. Student accessibility to all daily outdoor activities As it is mentioned, the next step is the conversion of the above composite individual accessibility values to the zones, which cover the whole area of the city. Particularly, the composite individual accessibility is converted from the point entity of students residences to the raster entity. For this reason, it is implemented the technique of Kriging for the point entity of residence, weighted by the values of the estimated compose individual accessibility measure, so as to create the raster entity with the size of 100*100 meter per pixel. The results of Kriging analysis (figure 6) shows that high levels of accessibility of students identified in the downtown and correspond to 20% of the area of the city. Satisfactory accessibility levels covers to 33% of the area of the city of Volos. Low and very low levels of students accessibility cover the remote areas from the centre of the city and represent 47% of the area of the Volos. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 144

152 SESSION: MODELLING PUBLIC TRANSPORT Figure 6. Predicted student accessibility to the city of Volos The proposed measure is compared with some knowns accessibility measures in order to assess its effectiveness. In particular, the proposed measure is compared with a n gravity measure ( i O j exp( tij ) ), a cumulative measure ( A i O jt ) and the ji j separation measure of distance. More analytically, the above measures are implemented with the data of the university students so as to be calculated the values of the student accessibility level to all daily destinations. In addition, the calculated values are normalised in a scale of 0-1 in order to be compared with the proposed measure. Table 2 illustrates the relationship of these measures with the proposed measure and shows the existence of satisfactory correlation with existing measures. The proposed measure is associated with the separation measure at 63%, with the gravity measure at 40%, while there is a lower level of correlation with the cumulative measure at 29%. Table 2. Correlation matrix between proposed measure and existing measures Gravity measure Cumulative measure Separation measure Proposed measure Gravity measure Cumulative measure Separation measure Proposed measure Pearson Corr Sig Pearson Corr Sig Pearson Corr Sig Pearson Corr Sig Finally, the application of the proposed measure for student s accessibility for the city of Volos shows that this particular social group has very good access from the locations of their residences in total daily individual activities. In the next section, the main conclusions of this research are presented. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 145

153 SESSION: MODELLING PUBLIC TRANSPORT 4. CONCLUSIONS The main scientific contribution is the development of a new composite individual accessibility measure, which is founded in the base of the triple relation of spatial, temporal and travel constraints, using aggregation techniques. All outdoor activities (fixed and flexible activities) of the individual in a day are imported in the new measure aiming to determine the overall individual accessibility during a typical day. The proposed measure has many advantages and confronts many drawbacks of the traditional and modern accessibility measures. At first, the incorporation of the individual daily activity space in an accessibility measure imposes the effect of the individual daily travel pattern to the activities, which is not found in the traditional measures. A second advantage is that this measure incorporates the effect of temporal factor, which is not applied in the existing traditional accessibility measures, while this factor has functional and application problems in modern accessibility measures. Concerning with the factor of travel constraint, the incorporation of Gaussian function into the new measure confronts the sudden change in the levels of accessibility, which is a drawback of the traditional gravity measures. At the same time, an additional advantage of this proposed measure is the functionality of this approach. The suggested measure can be used effectively in practice and the techniques which applied are relatively simple. The proposed measure is simple and it is quantified in a scale of 0-1, where zero (0) is defined to indicate the low levels of individual accessibility and 1 the high levels. To sum up, the suggested accessibility measure introduces a new aspect of accessibility at individual level, which combines successfully spatial, travel and temporal data. References Arentze T.A, Borgers A.W.J., Timmermans H.J.P. (1994) Multistop-based measurements of accessibility in a GIS environment. International Journal of Geographical Information Systems, Vol 8, Ben- Akiva M. and Lerman S. (1979) Disaggregate Travel and Mobility Choice Models and Measures of Accessibility. Behavioral Travel Modeling, eds. Hensher, D. and Stopher, P., London: Croom Helm. Bhat C., Handy S., Kockelman K., Mahmassani H., Gopal A., Srour I., Weston L. (2002) Development of an Urban Accessibility Index: Formulations, Aggregation and Application. Center for Transportation Research, University of Texas, Austin. Burns L.D. (1979) Transportation, Temporal and Spatial Components of Accessibility. Lexington Books D.C. Heath and Company Lexington, Cervero R. and Kockelman K. (1997) Travel demand and 3D s:density, diversity, and design, Transport Research Digest, Dalvi Q. and Martin K. (1976) The Measurement of accessibility: preliminary results, Transportation, nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 146

154 SESSION: MODELLING PUBLIC TRANSPORT Dong X. Ben-Akiva M. E. Bowman J. L. Walker J. (2006). Moving from Trip-Based to Activity-Based Measures of Accessibility, Transportation Research Part A 40, Geurs K.T. and van Wee B. (2004) Land use transport interaction models as tools for sustainability impact assessments of transport investments. European Journal of Transport and Infrastructure Research. Giannopoulos G.A., Boulougaris G. A. (1989) Definition of Accessibility for Railway Stations and its Impact on Railway Passenger Demand, Transportation Planning and Technology 13(1), Golledge R.G., Stimson RJ. (1987) Analytical Behavioural Geography. New York, Croom Helm. Hägerstrand T. (1970) What about people in regional science? Papers of the Regional Science Association, Vol 24, Hansen W.G. (1959) How accessibility shapes land use. Journal of the American Institute of Planners, Vol 25, Handy S. (1992) A Cycle of dependence: Automobiles, accessibility and the evolution of the transportation and retail hierarchies. The Berkeley Planning Journal, Vol. 9, Handy S.L., Niemeier D.A. (1997) Measuring Accessibility: An Exploration of Issues and Alternatives, Environment and Planning A 29, Iacono M., Krizek K., El-Geneidy A. (2008) Access to Destinations: How Close is Close Enough? Minnesota Department of Transportation Research. Ingram D.R. (1971) The Concept of Accessibility: A Search for an Operational Form. Regional Studies, Vol 5, Joly, O. (1999) Geographical Position: State of French Art of Spatial Accessibility Indicators, SPESD - France, (accessed www. Nordregio.se/spespn/Files/1.1.annex5.pdf). Koening J.G. (1980) Indicators of urban accessibility: theory and application. Transportation 9, Kwan M.P. (1998) Space-time and Integral Measures of Individual Accessibility: A Comparative Analysis Using a Point-based Framework. Geographical Analysis, Vol 30(3): Kwan M.P. (1999) Gender and individual access to urban opportunities: A study using space-time measures. Professional Geographer, Vol 51(2), Kwan M., Weber J. (2008) Scale and accessibility: implications for the analysis of land use travel interaction. Applied Geography, Vol 28 (2008) Kwan M.P., Weber J. (2003). Individual Accessibility Revisited: Implications for Geographical Analysis in the Twenty-First Century, Geographical Analysis 35(4). Lee M S., Goulias K. G. (1997) Accessibility indicators for transportation planning using GIS. 76th Annual Transportation Research Board meeting. Levinson D.M. (1998) Accessibility and the journey to work. Journal of Transport Geography, Vol 6(1), Litman T. (2008) Evaluating accessibility for transportation planning. 87th Transportation Research Board Annual Meeting, January Moustou F. (2013) Foundation and implementation of a composite individual accessibility measure using mobility data: Implementation in the city of Volos, Phd Thesis, University of Thessaly, Polytechnic of Volos, Department of Planning and Regional Development. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 147

155 SESSION: MODELLING PUBLIC TRANSPORT Munshi T. Brusel M. Zuidgeest M. (2009) Developing a Geo-Spatial Urban Form- Travel Behaviour Model for the City of Ahmedabad, India, (accessed urban-form-travel-behaviour-model-for-the-city-of-ahmedabat-india-talat- MUNSHI-Mark-BRUSSELS-Mark-ZUIDGEEST.pdf). Odoki J. (1992) Accessibility-Benefits Analysis as a Tool for Transportation Planning in Developing Countries. PhD Thesis, University of Trieste, Polytechnic of Milan, Italy. Odoki J., Kerali H.R.G., and Santorini F. (2000) An integrated model for quantifying accessibility-benefits in developing countries. Transportation Research Part A., Vol 35/7, O'Sullivan D., Morrison A., Shearer J. (2000) Using desktop GIS for the investigation of accessibility by public transport: An isochrone approach. International Journal of Geographical Information Systems, Vol 14, Shen Q. (1998) Location characteristics of inner-city neighborhoods and employment accessibility of low-wage workers. Environment and Planning B: Planning and Design, Vol 25, Sherman J., Spencer J., Preisser J., Gesler W. and Arcury T. (2005) A suite of methods for representing Activity Space in a healthcare accessibility study. Int J Health Geogr. Vol 4. Tagore M.R. and Sikdar P.K. (1995) A new accessibility measure accounting mobility parameters, 7th World Conference on Transport Research, The University of New South Wales, Wagner P. and Wegener M. (2007) Urban land use, transport and environment models, DisP, Wagner P. and Wegener M. (2011) From macro to micro How much micro is too much? Transport Reviews, 31(2) nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 148

156 SESSION: MODELLING PUBLIC TRANSPORT A Descriptive Study on Public Transport User Behaviour from Live Bus Arrivals Mauro Dell Amico 1, Selini Hadjidimitriou 1, Ioannis Kaparias 2 1 Department of Sciences and Methods for Engineering, University of Modena and Reggio Emilia, Pad. Morselli, via Amendola 2, Reggio Emilia (IT) ( dellamico@unimore.it, selini@unimore.it) 2 City University London, Northampton Square EC1V 0HB London (UK) ( kaparias@city.ac.uk) Abstract In order to offer public transport that meet citizens needs for transport and further increase the use of bus services, Public Authorities need to analyse and understand travellers behaviour. Automatic Vehicle Location (AVL) data provide information on the observed time of arrival and departure of a bus at each stop. These data are fed into an algorithm to provide information to users on the expected time of arrival at the bus stop by an on-line service. In the city of London this service is called Live Bus Arrivals. This work describes the general behaviour of Live Bus Arrivals users by analysing the type of requests, localising them and compare them in different days of the week and time ranges. The objective is to identify some of the main passengers origin, destination and interchanges behaviour that could be of value to decision-makers and planners. Keywords: public transport, real-time travellers information, travellers behavior. 1. INTRODUCTION The Live Bus Arrivals relies on the ibus AVL system by Transport for London (TfL) and provides information on the real time arrival of a bus line at stop in the London network. The user accesses the TfL website, using a web browser on a PC or a dedicated app for smart phones. Requests can be of four types: The user types the bus number and, if there are two possible route directions, chooses one. The output is a list of stops names and the user needs to select one; The user types the stop number and gets a list of buses approaching the stop with their corresponding time of arrival; The user types the name of the stop and will get the same output as above; The user writes the number of bus line and gets a map of the route. The aim of this work is to identify some of the main behaviours of the Live Bus Arrivals users, to make considerations on possible new functionalities for real time information systems and to outline the possibility for Public Authorities and decision makers to take advantage of the information that the Live Bus Arrivals data can provide. Specifically, this study analyses the potential capability of the data to provide information on transport flows between origins and destinations or in case of transfers between bus 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 149

157 SESSION: MODELLING PUBLIC TRANSPORT routes. The analysis will focus on the description of the requests for departures, arrivals and for bus-to-bus interchange by identifying which type of data can provide information on passengers behaviour. The final objective is to identify a set of procedures that use Live Bus Arrivals data that can be developed in future works to provide information on transport flows or a description of the selection of different stops as alternatives. 2. LITERATURE REVIEW There are many studies on the impact of real time information on public transport travellers behaviour and on the use of Intelligent Transport System (ITS) data to derive information on transport flows or users behaviour. For instance Dziekan et al. [1] studied the effects of real time information displays at stops on waiting time, on easiness to use the system, on willingness-to-pay, on mode choice and on customer satisfaction. They found that the provision of real time information at stop has the capability to reducing perceived waiting time. Additionally, they found that it is five time cheaper to introduce real time information systems with the aim to reduce perceived waiting time than increasing the frequency of a public service. Finally, they discover that people react at real time information by running when they see there are only few minutes left until the train departs. Watkins et al. [2] carried out a study on waiting time perception for users with and without real-time information. They discovered that perceived waiting time was 30% less for users of real time information. Additionally they found that mobile real time information reduce the actual waiting time because users arrive at the stop closer to the bus arrival instead of being there well in advance. Tang et al. [3] analysed longitudinal data on bus ridership and related them to employment rate, gas prices, weather conditions and other socioeconomic characteristics to study the impact of real time information on public transport usage. They found variations over time on bus ridership due to the installation of real time information but they were not able to provide conclusion on geographical variations. One of the first work on transit passengers and real time information was proposed by Hickman [4] who described a path choice model by incorporating real time information. Specifically, they introduced in the path choice model stochastic and time-dependent travel time. Their objective was to take into account the influence of a real time information system on alternative origin to destination choices. However the results showed that real time information does not provide significant benefits for transit passengers path choices in terms of reduction of travel time. In a recent work, Cats et al. [5] propose a dynamic transit path choice model and evidence that potential time savings are associated with the provision of real time information. Trozzi et al. [6] developed a time dependent route choice model to determine the impact of the count down system under overcrowding scenarios. They found that live information do not lead to a reduction of travel time but it changes travel behaviour of passengers who prefer less congested interchange stations. There are some studies aimed at estimating passenger flows using data obtained by smart ticketing systems. For instance Munizaga et al. [7] estimated a multimodal public transport OD matrix from SmartCard data and found that their method allowed to estimate 80% of the total boardings. Although the dataset did not included information 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 150

158 SESSION: MODELLING PUBLIC TRANSPORT on the destination, the alighting station has been estimated by looking at the passengers position and the time of the next boarding. Other studies focus on the development of algorithms and interfaces for real time transit planner. For instance Jariya- sunant et al. [8] developed a mobile real time system that implements a K-shortest path algorithm using bus arrivals predictions. The user types the origin and the destination and obtains a personalised shortest path on a map. The results of performance analysis show only a marginal improvement of travel time. Finally, Hardy [9] analysed Live Bus Arrivals requests by time of the day, location and type of channel. The user requests information from home, using a PC, a smart phone or the SMS service. The study showed that the demand for real time information was higher during peak hours meaning that the service is mostly used when waiting time is longer. In this study we will describe Live Bus Arrivals data by focusing on specific examples, without stating general as- sumptions on users behaviour. The description of the database will allow to determine its use in relation to the possibility to contribute to existing studies on the use of real time information to estimate waiting time, on the possibility to estimate public transport OD matrices and on the analysis of passengers route choices. 3. THE DATASET The data analysed in this work refers to five days between the 16th and 20th July The week under examination covers the final stages of preparation for the Olympic Games that were held in London between the 27th July and 12th August The period corresponds also to the last week before most schools closed for summer holidays. Week-ends and night hours are not included in the analysis which focuses on working days (Mon-Fri). The original file consists of about 4 million rows for each observed day. Each string includes the IP address which usually refers to one or a set of requests sent by the same user, the time and the day of the request, the bus and/or the stop number, the cellphone model and the browser installed. An example of string is shown in Figure 1. The format of the file includes univocal keywords and non structured system calls from which it is necessary to identify and extract the needed information such as stop codes and line buses. Figure 1. Original data The example shows a request for real time arrivals at Chatteris Avenue (stop number 72979). The output provided to the user consists of information related to three bus routes due to arrive at Langbourne Place (72979): the 256, 499 and 174. This single string does not provide any information on which bus routes the user is waiting for. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 151

159 SESSION: MODELLING PUBLIC TRANSPORT To analyse bus passengers behaviour using Live Bus Arrivals data, the first step was to extract the number of stops and/or bus routes from each string. As a general rule, buses and stops are always preceded by one of the keywords listed in Table 1. Table 1. Keywords Type of string 1 /arrivals/ 2 /stopsnearlocation/ 3 stop= 4 /mystops/ 5 searchterm= 6 showjourneypattern/ The resulting file was then imported into the relational database Postgresql 9.3 for data cleaning and elaboration. The output is a list of requests with several empty rows, bus routes and stop codes either in one of the last two columns of the table as reported in Figure 2. In the second phase, bus routes and stop codes have been positioned in their corresponding columns and duplicates or empty rows have been eliminated. A screenshot of cleaned data is shown in Figure 3 in which five different users have requested either for a bus line or a stop arrival. Two requests for time of arrival at Cloister Gardens (48215) were asked by the first user ( ) at 6:41 and 6:45. Another couple of requests are associated to the same IP address and have been sent at 18:02 for stop Oakmead Gardens (50049). Although the two stops are very close to each other, and the two IP addresses are equal it is not possible to know with certainty if the two requests have been sent by the same user. The Mobile Network Operator assigns different IP addresses to mobile users especially when they access the web or when they change their position. Since the same IP address can be assigned to different users, it cannot be deployed as unique id identifier. Nevertheless the IP address can help to identify multiple consecutive requests sent by the same user in a short period of time. To reduce data misinterpretation, in this work the IP address will be used to identify a single user only when the requests are sent within a hour. With reference to the other users of Figure 3, multiple requests were sent within one or fifteen minutes. Specifically, user by sending multiple requests for the same stop and bus route in a time range of about fifteen minutes has probably boarded at Clamp Hill (57184) on bus 258. The time elapsed between the last and the first request can be used as an indicator of user s waiting time at stop. Or as an indicator of perceived waiting time longer than expected. Finally a further data elaboration was required to help in the analysis. First of all requests in each hourly time range have been grouped and georeferenced. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 152

160 SESSION: MODELLING PUBLIC TRANSPORT Figure 2. Extracted data Figure 3. Cleaned data Moreover, for each hourly range and for each user, repeated requests for a stop and/or for a bus have been counted. This information is included in an additional column named count. 4. CLASSIFICATION OF LIVE BUS ARRIVALS USERS BEHAVIOUR This section describes user behaviour by the identification of few main groups of customary users of the Live Bus Arrivals service. These groups belong to two main categories based on the information Live Bus Arrivals users provide by sending their requests for time of arrival at one or more stops or for a bus route. These information can be complete, partial or on bus-to-bus transfer. The possibility to have complete information on passengers trips depends on two factors: 1. the accuracy of the information requested by the passenger and 2. if the time of arrivals is asked for more than one stop, the possibility to unequivocally identify the bus route passing by the requested stops. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 153

161 SESSION: MODELLING PUBLIC TRANSPORT 4.1. Complete information In case of complete information, the most accurate data that a user can provide consists of the code number of boarding and alighting stops and the bus route. Some variants are also possible which provide complete information such as passengers asking information not only for bus arrivals at origin and destination but also at intermediate stops. Complete information can be obtained if a passenger specifies the stop of origin and destination and there is only one line passing by both stops. Figure 4 displays the table with the list of requests for real time arrivals of a bus route (496 to Harold Wood), two stops, Harold Wood Station (51285) and The Brewery (52277), and a graphical representation of the re- quests. These types of requests allow unambiguous interpretation and can be used as disaggregated data on passengers origin and destination Partial information Figure 4. Complete information Partial information on trips is obtained when a passenger asks about the time of arrival of a bus at the stop of boarding or indicates both the stop number and the bus route. This is the case in which a user accesses the Live Bus Arrivals system by typing the number of bus and selects one of the stops from the list. Although the destination is not known, these data can provide information on stop departures and/or on route s flows. Repeated requests. During the week under observation, the 10,2% on average of the Live Bus Arrivals users have sent more than 5 repeated requests between 7 and 8 a.m. When asking for real time arrivals, users continuously update the system, probably until the bus arrives. Figures 5 shows the average number of requests sent between 7 and 8 a.m. during Monday-Friday. Large circles indicate high averages of repeated requests (more than 20 requests within a hour for arrivals at the same bus stop). High average values are mainly located at stops outside the city centre. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 154

162 SESSION: MODELLING PUBLIC TRANSPORT Figure 5. Average repeated requests (7-8 a.m) Monday, 16/07/ Friday, 20/07/2012 Requests for live arrivals of a bus route at a boarding stop. When passengers ask for live arrival of a bus at a specific stop, the information is partial because there are no data on the destination but only on the stop of departure, on the bus route and its direction. For instance Figure 6 shows that the user has asked for bus W8 arrivals at Browning stop (47051), the route direction is Herefield Close. In this example even if the destination is unknown, there are only five stops until the end of the line. Figure 6. Request for live arrivals of a bus route at a stop Requests of a bus route. Sometimes users request information on a bus line to know the route direction or to have an overview of the list of stops. In this case no real time information on arrivals at stop is requested so that no data are available on the trip. However this information can be used to know how popular a particular route may be. Requests at near stops. Furthermore information is partial when a user enquires about the time of arrival at near stops. When a user is interested on arrivals at neighbouring stops, there could be three possible reasons: a. the same bus is passing by the neighbouring stop and the user intends to walk there without missing the bus; b. the user is checking for information on another stop to transfer or board another bus; c. the user wants to use the bus only for very few stops. Figure 7 shows a request for information on two stops located near each other. In the example it is not possible to know which of the two bus lines, the 34 and 102, the passenger will board. Thus there 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 155

163 SESSION: MODELLING PUBLIC TRANSPORT is no information on the final destination. This data should be grouped together with the Repeated requests and should be handled accordingly to avoid misinterpretation (i.e. by considering the two stops as the origin and the destination). Figure 7. Requests for near stops Requests for same stop but opposite route direction. In some cases users look for Live Bus Arrivals at both directions of the same stop. Therefore it may be that the user cannot easily identify which is correct route direction. Bus route directions are usually identified with the stops at the end of the line. However public transport passengers probably know better about their stop of arrival instead of the end of the line. A new real time functionality could include the possibility to type the stop of arrival and, based on that indication, the system could provide some indication on which is the bus stop the user has to board. Indication of stops direction could consists in colours or letters applied at the stop, near the stop code. So that when a user types his origin and destination, the system could provide the indication on the correct stop to board. Bus-to-Bus transfer. Passengers often use the Live Bus Arrivals service to get information on arrivals at the bus stop they want to transfer. For instance, Figure 8 displays the set of requests of a passenger who has boarded at Kingsbury station (72740). Six bus routes by Kingsbury station but the user has most likely boarded bus route 183 to Pinner Station. This is in fact the only bus route that allows to arrive close to where the next two requested stops (72141 and 59160) are located. The interpretation of user behaviour in the second part of the trip is complex. The two stops for which the user has requested real time arrivals are in reality the same stop (Hunters Grove) where bus routes transit with opposite directions. From these stops several bus routes transit: the 114, H9, H19, H10, H18. All of them have transfer points along route 183. Possible explanations on why the user has asked arrivals information at Hunters Grove are: 1. the user did not know he could transfer along line 183 to one of the buses passing by Hunters Grove so that has caught the bus from another stop; 2. the user knew that there was an easier transfer point but, based on real time of arrivals, has decided to get off one stop earlier and to walk to the transfer stop. The first hypothesis is less probable because the passenger has on-line information. Furthermore this behaviour and other examples that will be described in the next paragraphs suggest that public transport passengers make use of real time information when boarding from one bus line to another. So that a specific functionality aimed to quick and easily provide such type of information could be well accepted by the users. Another example of bus-to-bus transfer which is easy to interpret is shown in Figure 9 where a passenger has asked for real time arrivals of bus route 309 to Bethnal Green at Devons Roads stop. The same user has then looked for departures of bus route number 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 156

164 SESSION: MODELLING PUBLIC TRANSPORT 323 to Mill End firstly from Devons Road (DLR), then to a closer stop, St. Pauls Way School stop. Generally the interpretation of Live Bus Arrivals data becomes complex when users look at two or more travel options. Figure 8. Bus-to-Bus transfer: data interpretation Figure 9. Bus-to-Bus transfer at near stops Figure 10 shows that twelve repeated requests have been sent to get live arrivals at Worsley Bridge Road to Catford and four requests have been sent for the same stop but for the opposite direction. Worsley Bridge Road is connected to Catford Road by bus route 181, while Beckenham Hill is connected to Catford by bus 54 or by rail. The graphical representation of the requests allows to understand that the user has looked at two possible routes to arrive at destination (Catford Road). The decision on which route to choose has been probably influenced by real time arrivals. So that the user has boarded either the 181 to Catford or the 181 to Beckenham Hill and then bus route 54 to Catford. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 157

165 SESSION: MODELLING PUBLIC TRANSPORT Figure 10. Bus-to-Bus transfer: alternative routes Figure 11 shows two types of requests for real time arrivals: the first at bus stop Leonard Avenue (51795), the second for bus route 174 at Reinham Road North (56184). In this case there are two optional bus routes that allow to arrive near to Reinham Road North from Leonard Avenue: bus routes 175 and 103. The most probable option is that the user has caught bus 103 to arrive close to Reinham Road North, walked a few hundred meters and transferred to the southbound bus 174. Figure 11. Bus-to-Bus transfer: first route identification Figure 12 shows an easy interpretation of bus-to-bus transfer. Firstly the user has asked information on arrivals at Whitworth Road to Upper Norwood and then at Springfield Road to West Croydon. The two lines intersect each other even if there is no stop in common. Given the direction of the bus lines which are known thanks to the stops number, it is possible that the user has transferred from bus 468 to nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 158

166 SESSION: MODELLING PUBLIC TRANSPORT Figure 12. Bus-to-Bus transfer Finally Figure 13 shows the last example of bus-to-bus transfer. In this case the traveller has firstly asked for information on arrivals at Tottenham Court Road Station (51056). Since from that station is possible to board many bus lines, it is necessary to look at the location of the second request which is Ebury Bridge (47551). The only bus route from Tottenham Court to a stop closed to Ebury Bridge is bus route 73 to Victoria. From Ebury Bridge it is only possible to board bus route C10 to Canada Water. It has to be noticed that 73 to C10 transfer was already possible at Victoria. Therefore this is another example of a user who has boarded from another stop instead of the one located very close to the one of the previous bus line (bus 73). A possible explanation is that, using real time information, users prefer to walk instead of waiting at the stop for the bus Outliers Figure 13. Bus-to-Bus transfer Not all data can be used for the analysis of travel behaviour. For instance, some users ask for arrivals at two opposite direction stops. In this case is it not possible to define from which stop the passenger boards. Furthermore it is necessary to exclude users with behaviours that are not clearly defined for instance when many requests are sent during a hour for any stop around the city such as in Figure 14. There are other types of requests which are not possible to interpret. For instance users of the Live Bus Arrivals service who request information on live arrivals at multiple stops within 10 km where several bus routes serve the requested stops. The intent of this type of request is probably to gather information on all bus routes transiting near by. However these data should be eliminated from the analysis because they cannot provide any precise information on user behaviour. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 159

167 SESSION: MODELLING PUBLIC TRANSPORT Figure 14. Outliers 5. CONCLUSIONS AND FUTURE RESEARCH DIRECTIONS The main conclusions on traveler behaviour as arising from the study of the Live Bus Arrivals data can be summarised as follows: The 10% of users frequently update their request for live arrivals at a bus stop. A real time information system could include the possibility to keep the system updated, for instance with a clock which the user can activate or disactivate. Furthermore the information gathered from repeated requests could allow to develop an indicator of perceived waiting time that depends on the number of requests sent by each user. For instance if the time elapsed between the first and the last request is greater than 10/15 minutes and the user has updated the system every second, the time elapsed between the first and the last request could be used as an estimation of waiting time at stop. The value of the indicators can be then related to weather and traffic conditions or bus frequencies at stops to detect if there are variables that have an influence on perceived waiting time. To this aim two different time periods could be compared: a period of regular traffic and an overcrowding period such as the one during which the Olympic Games took place. In some cases public transport passengers check for arrivals at both directions of the same stop. This probably means that it is not always clear to them which bus route direction they should board to arrive at destination. A real time public transport information system could include an indication referred to signals applied to the stop that allows to instantly recognise the bus route direction the user is looking for. In order to be able to indicate the route direction, the system could require the user to type the stop of departure and arrival. Therefore information on origin and destination would be provided directly by the user thus allowing to estimate the public transport OD matrix. Users sometimes request for arrivals at a bus stop and for a bus line. This type of information allows to determine the popularity of a bus route or stop especially with reference to the use of the Live Bus Arrivals service or to determine if users prefer less crowded stops. Moreover the description of some examples of bus-to-bus transfer has allowed to have an idea on how passengers use the Live Bus Arrivals service to gather information on bus arrivals at the transfer stop. They probably adjust their travel plan while they are travelling and choose the route or the alighting stops for transfer basing on real time information. For instance it seems from the examples on bus-to-bus transfer that passengers sometimes walk to another stop instead of boarding at the interchange stop. Hence, following Trozzi et al. (2013) results, future work will concentrate on how 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 160

168 SESSION: MODELLING PUBLIC TRANSPORT live information could influence equally attractive stops and lines upon departures. For instance given two equally attractive stops, the study will explore how users route choice change based on LiveBus Arrivals information. Acknowledgements: This work was partially supported by the COST Action TU1004: Modelling Public Transport Passenger Flows in the Era of Intelligent Transport Systems. We would like to acknowledge networking support by Cost Action TU1004 and Transport for London Buses for providing the data. References Katrin Dziekan, Karl Kottenhoff, Dynamic at-stop real-time information displays for public transport: effects on customers, Transportation Re- search Part A, 41, , (2007) Kari Edison Watkins, Brian Ferris, Alan Borning, G. Scott Rutherford, David Layton, Where Is My Bus? Impact of mobile real-time information on the perceived and actual wait time of transit riders, Transportation Research Part A: Policy and Practice, Volume 45, Issue 8, Pages (2011) Mark D. Hickman, Nigel H.M. Wilson, Passenger travel time and path choice implications of real-time transit information, Transportation Research Part C: Emerging Technologies, 3, 4, (1995) Oded Cats, Wilco Burghout, Tomer Toledo, Haris N. Koutsopoulos, Effect of Real- Time Transit Information on Dynamic Passenger Path Choice, Transportation Research Record, 2217, (2011) Valentina Trozzi, Guido Gentile, Michael G. H. Bell, Ioannis Kaparias, Effects of countdown displays in public transport route choice under severe overcrowding, Networks and Spatial Economics, 01 (2013) Marcela A. Munizaga, Carolina Palma, Estimation of a disaggregate multimodal public transport OriginDestination matrix from passive smartcard data from Santiago, Chile, Transportation Research Part C: Emerging Technologies, 24, 9-18 (2012) Lei Tang, Piyushimita (Vonu) Thakuriah, Ridership effects of real-time bus information system: A case study in the City of Chicago, Transportation Research Part C: Emerging Technologies, 22, (2012) Jerald Jariyasunant, Daniel B. Work, Raja Sengupta, Branko Kerkez, Steven Glaser, Alexandre Bayen, Mobile Transit Trip Planning with Real- Time Data, Transportation Research Board 89th Annual Meeting, Washington, D.C., Jan (2010) Nigel Hardy, Provision of bus real time information to all bus stop in London, 19th ITS World Congress (2012) 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 161

169 SESSION: TRANSIT QUALITY PERFORMANCE

170 SESSION: TRANSIT QUALITY PERFORMANCE Sustainable Urban Mobility Indicators for Medium-Sized Cities. The Case of Serres, Greece Dimitrios Karagiannakidis, Alexandros Sdoukopoulos, Nikolaos Gavanas*, Magda Pitsiava-Latinopoulou Transport Engineering Laboratory, Faculty of Engineering, Department of Civil Engineering, Aristotle University of Thessaloniki, Thessaloniki, Greece *Corresponding author: Tel: Abstract The operation of the urban transport system generates impacts which affect directly the vast majority of the population, who live in urban agglomerations of various sizes, while at the same time impose pressures on environmental sustainability at the global scale. In this context, the promotion of sustainable urban mobility is of high priority in the European policy framework through a series of strategies and recommendations, such as the Sustainable Urban Mobility Plans. The effective assessment of these strategies implementation is based on the use of a system of appropriate indicators which reflect the components of sustainable mobility (i.e. economic, social and environmental issues). This system of indicators should be measurable, conceptually relevant to the examined objective and adjusted to the city scale. In addition this system should be fed with valid and compatible transport related data sets, in order to provide an efficient evaluation tool of the sustainable mobility conditions across the European cities (benchmarking). The latter constitutes the main drawback especially for mediumsized cities. Towards this direction, the current research aims to develop a methodological approach for assessing and evaluating the sustainable mobility conditions in medium-sized cities, taking into account the relevant available indicators from the international literature review and adjusting them properly. The proposed methodology is implemented in a case study for the city of Serres, a typical mediumsized city of Greece. The results of this application apart from highlighting the priorities for the improvement of the specific city s sustainable mobility conditions, they also outline the framework for the establishment of a benchmarking mechanism. Keywords: Sustainable urban mobility indicators, Sustainable city index, Mediumsized cities, Serres. 1. INTRODUCTION The European Commission s 1990 Green Paper on the Urban Environment (COM (90) 218 final) set forth the principles for upgrading the urban environment, with reference to the role of the transport system and the need to restrain the role of the private car. At the meantime, the first Transport White Paper, i.e. The future development of the common transport policy (COM (92) 492 final), mentioned the significance of a Europe-wide transport observatory which will contribute to the collection and exchange of valid and compatible transport related data. In addition, the 1995 Green Paper on The Citizens Network (COM (95) 601 final) highlighted the goal of an integrated transport system by enhancing public and combined transportation. In this framework, the European Commission s Communication on 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 162

171 SESSION: TRANSIT QUALITY PERFORMANCE Developing the Citizens Network (COM (98) 431 final) proposed methods in order to increase the sustainability of transport systems and promoted benchmarking in the transport sector. The value of such benchmarking mechanisms refers to the ability of comparative evaluation between different entities, such as Member States, regions or cities according to the level of reference, but also to the potential of collecting compatible data at the European level. At the turn of the century, the second Transport White Paper, i.e. European transport policy for the year 2010 (COM (2001) 370 final), established more specific targets for sustainable transportation in terms of promoting social inclusion and economic growth, upgrading the quality and safety of services and achieving a more balanced and environmentally friendly transport system. Based on the aforementioned principles, the 2007 Green Paper Towards a new culture for urban mobility (COM (2007) 551 final) aims at the formulation of a common European approach on free-flowing, greener and smarter cities through an effective, safe and accessible multimodal transport system. In this context, the Action Plan on Urban Mobility (COM (2009) 490 final) included six main priority axes in order to support sustainable urban mobility, i.e. i. Promoting integrated policies, ii. Focusing on citizens, iii. Achieving cleaner urban transport, iv. Strengthening funding, v. Sharing experience and knowledge and vi. Optimizing urban mobility. Regarding the priorities for sharing experience and knowledge, specific reference is made to the upgrade of data and statistics, the establishment of an urban mobility observatory and the promotion of international dialogue and information exchange concerning urban mobility. The third Transport White Paper, i.e. Roadmap to a Single European Transport Area - Towards a competitive and resource efficient transport system (COM (2011) 144 final), points out that cities provide favourable conditions for the implementation of multimodal and alternative mobility solutions due to the lower requirements for vehicle range and the higher density of population. Furthermore, the high concentration of population in urban areas enhances the significance of developing sustainable solutions for urban mobility. Approximately 70% of the European citizens live in cities with population of or more producing more than 80% of Europe s Gross Domestic Product (GDP) (COM (2013) 913 final). At the administrative level, 12.6% of the European population lives in municipalities of 50, ,000 inhabitants, i.e. the sizeclass of population which corresponds to the case study presented in this paper (ESPON 2006). Despite the potential and necessity for sustainable mobility in the cities, urban transport is responsible for the 25% of CO2 emissions from transport and almost 70% of road accidents (COM (2011) 144 final). Nonetheless, the European experience provides a wide inventory of interventions that support sustainable urban mobility. However, a large share of them is implemented under the initiative of local authorities or in the context of collaborative initiatives which include a group of European cities. Moreover, these initiatives are often concentrated on the problems of larger cities. Recently, the Sustainable Urban Mobility Plan (SUMP) framework provides the guidance and methodology at the level of strategic, integrated transport planning (COM (2013) 913 final). A problem which still remains refers to the development of valid and compatible transport related data sets which can be processed in order to provide an efficient evaluation of the sustainable mobility conditions across the European cities. The problem is more intense in the case 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 163

172 SESSION: TRANSIT QUALITY PERFORMANCE of medium-sized cities, i.e. cities with less than 250,000 inhabitants, where there is no sufficient data in order to promote effective strategies for sustainable development. A common approach for the acquisition of the required data is the development and application of appropriate indicator systems that enable the planners and decision makers to identify trends, compare different time periods and urban areas, evaluate policies and formulate strategic objectives (Litman 2006). Furthermore, the international experience provides methodologies, such as the Sustainable City Index, in order to compose these indicators for the assessment of an integrated measurement of sustainable mobility which characterises each city and allows comparisons with other cities of similar or different features (Choon et al. 2011). In this context, the scope of the current paper comprises the presentation of a methodological approach for assessing and evaluating the sustainable mobility conditions in medium-sized cities, taking into account the international literature review and adjusting it for implementation in a medium-sized city of Greece. The paper is organised as follows: After the above presented overview of European policy concerning the significance of sustainable urban mobility and the necessity for compatible indicators for accurate and comparable transport related data, a description of the selected study area, i.e. the city of Serres, is made. The next part of the paper refers to the presentation of the methodological approach of the research while in the final section the conclusive remarks are presented. 2. DESCRIPTION OF THE STUDY AREA The study area of the research is the city of Serres, which is a typical medium-sized city of Greece, the second biggest city in the region of Central Macedonia and the capital of the regional unit of Serres. The administrative area of Serres extends over an area of 252 km 2 including the suburbs, while according to the results of the recent (2011) census, its population reaches approximately 76,800 inhabitants. ( /page/portal/esye/page-cencus2011tables, accessed March 22, 2014) The study area s road network consists of arterials, collector and local roads and its structure is a combination of a radial and a grid pattern. According to the latest General Transportation Study (completed in 2002), the city centre is the origin or/and destination of the vast majority of daily trips while the main trip purpose is commuting to work. The concentration of population and activity in the city centre in combination with the dominant role of the private car in daily mobility and the inefficient control of illegal parking impose significant pressures on the city s transport system and result to traffic congestion mainly during morning and evening peak-hours. Regarding the city s spatial structure, the central and peri-central areas comprise mixed land use areas where residential, commercial and institutional uses co-exist with specific poles of cultural and recreational activity. The Technological Educational Institute (TEI) of Serres, which is located at the south edge of the city, is the main pole of educational activity, attracting and generating a total of approximately 10,000 daily trips. The study area is presented in the figure below (Figure 1). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 164

173 SESSION: TRANSIT QUALITY PERFORMANCE 3. METHODOLOGICAL APPROACH Figure 1. Study area The methodology of the research is presented in Figure 2 and consists of the following steps: a) Selection of the appropriate set of indicators and development of the sustainable mobility indicator system taking into account the findings of the literature review and the features of the study area, b) Assessment of indicators using the available data sources and carrying out on-field measurements, c) Normalisation of indicators combining the guidelines of related European projects, i.e. the Urban Transport Benchmarking Initiative and the Ecomobility Shift, d) Assessment of the integrated Sustainable City Index (SCI) by composing the normalised indicators, e) Evaluation of the sustainable urban mobility features of the study area. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 165

174 SESSION: TRANSIT QUALITY PERFORMANCE 3.1. System of indicators Figure 2. Followed methodology The process of developing the appropriate system of indicators is based on the international literature review with the purpose of investigating the quality and quantity of available sources. The results of the review are listed in Table 1. From this Table, it can be observed that there is no specification for the number of indicators that compose an indicator system. The number of indicators depends on the requirements of the case study while the final selection should be balanced between ease and completeness (Munier 2011; Litman 2007). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 166

175 SESSION: TRANSIT QUALITY PERFORMANCE Table 1. Review of main sustainable urban mobility indicator sets No Authors Year References Number of indicators 1 Kupiszewska, D Modelling for sustainable cities: the transport sector 32 2 OECD 1999 Indicators for the integration of environmental 27 concerns into transport policies 3 Newman, P., Kenworthy, J Sustainability and cities: overcoming automobile 22 4 Department of the Environment, Transport and the Regions 5 European Environmental Agency 6 Gilbert, R., Irwin, N., Hollingworth, B., Blais, P., Lu, H., Brescacin, N. 7 Jones, P., Jucas, K., Whittles, M. 8 Minken, H., Jonsson, D., Shepherd, S., Järvi, T., May, A., Page, M., Pearman, A., Pfaffenbichler, P., Timms, P., Vold, A. 9 Nicolas, J., Pochet P., Poimboeuf, H. 10 Gilbert, R., Irwin, N., Hollingworth, B. 11 Department for Environment, Food and Rural Affairs dependence 2000 Local quality of life counts Indicators of transport and environment integration 38 TERM 2002 Sustainable Transport Indicator Project, CST The Civilising Cities initiative PROSPECTS project s methodological guidebook Towards sustainable mobility indicators application to the Lyons conurbation Sustainable transportation performance indicators 14 (STPI) 2005 Securing the future Department for Transport 2005 How to monitor indicators in Local Transport Plans and Annual Progress Reports Update 13 Jeon, C., Amekudzi, A Addressing sustainability in transportation systems: definitions, indicators and metrics 14 Zegras, Ch Sustainable transport indicators and assessment methodologies 15 Savelson, A., Colman, R Sustainable transportation in Halifax regional municipality, GPI (Genuine Progress Index) for Atlantic Canada 16 Moles, R., Foley, W., Morrissey, J Practical appraisal of sustainable development, methodologies for sustainability measurement at settlement level 17 Litman, T Sustainable transportation indicators Appleton B., Davies M SMART transportation ranking report (27 Canadian 12 cities) 19 Litman, T Sustainable transportation indicator data quality and 35 availability 20 Litman, T Well measured developing indicators for 41 comprehensive and sustainable transport planning 21 Doody, D.G., Kearney, P., Barry, J., Moles, R., O Regan, B Evaluation of the Q-method as a method of public participation in the selection of sustainable development indicators 5 22 Castillo, H., and Pitfield, D. E. 23 Tanguay, A., Lefebvre, J.F., Lanoie, P. 24 Mascarenhas, A., Coelho, P., Subtil, E., Ramos, T.B. 25 Choon, S. W., Siwar, C., Pereira, J. J., Jemain, A. A., Hashim, H. S., & Hadi, A. S. 26 Shen, L. Y., Jorge Ochoa, J., Shah, M. N., & Zhang, X ELASTIC - A methodological framework for identifying and selecting sustainable transport indicators 2010 Measuring the sustainability of cities: an analysis of the use of local indicators (23 study) The role of common local indicators in regional 5 sustainability assessment 2011 A sustainable city index for Malaysia The application of urban sustainability indicators - A comparison between various practices nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 167

176 SESSION: TRANSIT QUALITY PERFORMANCE 27 Zheng, J., Garrick, N. W., Atkinson-Palombo, C., McCahill, C., & Marshall, W Guidelines on developing performance metrics for evaluating transportation sustainability Source: (Haghshenas and Vaziri 2012; Castillo and Pitfield 2010 and own elaboration) The objective of the literature review is the analysis and evaluation of the indicators implemented in previous research and the selection among them of the indicators which are more suitable to be inserted to the indicator system. According to Litman (2009), an option may seem to be appropriate and desirable if evaluated by one set of indicators but unsustainable if evaluated by others and thus, a concrete set of selection criteria was set by the specific research (Atkinson et al. 1997; Barrera-Roldán and Saldıvar-Valdés 2002; Lee and Huang 2007; Castillo and Pitfield 2010; Litman 2007; Haghshenas and Vaziri 2012; Pitsiava 2012; Zheng et al. 2013): Relevance to sustainability: Indicators must describe performance in the fields of social, economic and environmental sustainability. Policy relevance: Indicators should be able to illustrate the impact of transportrelated policies so as to indicate progress towards sustainability. Data availability: The assessment of an indicator requires the availability of specific data or methods to obtain the data. Continuity: The ability to assess an indicator at regular time intervals and evaluate trends through time is an essential asset which enhances its durability. Reliability: The validity of an indicator depends on the quality of data, which should derive from official sources and scientific methodologies. Simplification: The structure of an indicator should be simple in order to be easily understood and used. Compatibility: Indicators should be compatible and enable comparisons between urban areas. Affordability: The cost for assessing an indicator is a main criterion that ensures its potential use. Accessibility and transparency: The features of the indicator, such as scope, method of calculation, results, metadata etc., should be accessible to stakeholders. Sensitivity: Indicators must be capable of capturing and illustrating changes concerning sustainable mobility. Suitability: The features which are measured by an indicator as well as its spatial and time reference should be suitable to the specific case study. According to the first criterion, the indicator system must cover the three main fields of sustainability, i.e. economic competitiveness, social welfare and environmental safeguard. In this context, it is common for an indicator system to comprise the following categories (Choon et al. 2011; Litman 2007; Zheng et al 2013; Litman 2006; Haghshenas and Vaziri 2012): a) Environmental, b) Social and c) Economic. A change in mobility conditions can produce a conflict concerning the impact on the aforementioned categories. For instance, an increase in private car use may imply an increase in economic growth while generating a negative impact on the environment and some aspects of social welfare nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 168

177 SESSION: TRANSIT QUALITY PERFORMANCE Taking into account the above criteria and categorization, a set of indicators was selected for the city of Serres deriving from various sources. However, the main source was the Programme Urban Transport Benchmarking Initiative (European Commission. Directorate General for Energy and Transport 2006), which developed and assessed sets of indicators for cities of different population size-classes. More specifically, eleven (11) medium-sized cities participated in the Programme. In addition, a number of selected indicators derives from the Programme: EcoMobility Shift ( shift-manual?download=32:appendix-1-indicator-descriptions, accessed October 16, 2013). The indicator system which was developed and implemented by the specific research for the city of Serres is presented in Table 2. Table 2. Overview of the indicator system I.D. Indicator Description Unit Policy goal Environmental EN-1 Road density per surface (land take) EN-2 PT infrastructure density EN-3 Share of traffic calming roads EN-4 EN-5 Pedestrian infrastructure density Cycling infrastructure density Road network length per km 2 of surface area Public transport (PT) network length per km 2 of surface area Roads with traffic calming measures as a proportion of total road network Sidewalks length per km 2 of surface area Cycle-route network as a proportion of total road network km/ km 2 Effective traffic management km/ km 2 Improved PT LOS - Regeneration of urban area % Effective traffic management - Increased road safety km/ km 2 Improved pedestrian LOS - Regeneration of urban area % Improved bicycle LOS - Regeneration of urban area EN-6 Age of PT fleet Average age of PT vehicles Years Improved PT LOS EN-7 Engine technology Share of PT vehicles with EURO % Reduction of fuel of PT fleet 4 engine consumption and pollutant EN-8 Social SC-1 SC-2 SC-3 SC-4 SC-5 SC-6 SC-7 Fuel efficiency of PT fleet Road density per population (availability) PT density per population PT network coverage PT size in relation to population Accessibility for all (PT vehicles) Accessibility for all (PT infrastructure) Cycle parking availability Number of kilometres travelled per litre of fuel used Road network length per 1,000 inhabitants PT network length per 1,000 inhabitants Share of people living within a distance of 350 m from a PT stop Average number of inhabitants per PT vehicle Share of the PT vehicles that are wheelchair accessible Share of the PT stops that are wheelchair accessible Number of bicycle parking spaces per 1,000 inhabitants SC-8 Road safety Number of accidents per 10,000 inhabitants SC-9 Road safety and vulnerable users Number of accidents involving vulnerable users per 10,000 inhabitants km/ l km/ 1,000 inh. emissions Reduction of fuel consumption and pollutant emissions Effective traffic management km/ 1,000 Improved PT LOS - inh. Regeneration of urban area % Improved access to services and goods inh./ Improved PT LOS - vehicle Increased mobility % Improved access to services and goods % Improved access to services and goods number of Encouragement of parking alternative modes of spaces/ transport 1,000 inh. number of Increased road safety accidents/ inh. number of Increased road safety accidents/ 10,000 inh. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 169

178 SESSION: TRANSIT QUALITY PERFORMANCE Economic EC-1 GDP per capita Gross Domestic Product (GDP) in Euros divided by the number of inhabitants EC-2 Car ownership Number of owned private vehicles per 1,000 inhabitants EC-3 Share of PT PT as an average share of the modal distribution of traffic EC-4 Rate of PT use Average annual number of PT trips per inhabitant EC-5 EC-6 EC-7 PT peak-hour speed Car speed to PT speed PT frequency during peak-hour Average speed of PT vehicles during peak-hour Ratio of the average speed of private vehicles to the average speed of PT vehicles Most frequent peak-hour service intervals (in minutes) Improved access to services and goods number of vehicles/ 1,000 inh. Improved access to services and goods % Increased multimodality trips/ person/ year Encouragement of alternative modes of transport km/ h Improved PT LOS - Increased mobility - Improved PT LOS - Increased mobility min. Improved PT LOS - Increased mobility In order to obtain a complete description of each indicator, a technical report was developed and filled in. The format of the technical report is based on the corresponding indicator sheets of the Egnatia Odos Observatory ( accessed 29 March, 2014). An example of a technical report for indicator: EN-3. Share of traffic calming roads is presented in Table 3. I.D. Name Description Methodological approach Table 3. Example of an indicator s technical report Technical report of indicator EN-3 Share of traffic calming roads Roads with traffic calming measures as a share of the total length of the road network a. Analysis of the basemap of the study area b. Measurement of the total length of the road network c. Measurement of the total length of roads with traffic calming measures d. Division of the total length of roads with traffic calming measures by the total length of the road network Unit % Calculation frequency Short-term (2 years) Medium-term (5 years) Long-term (10 years) Policy goal Effective traffic management - Increased road safety Promotion of Active Transport Spatial reference Regional unit, Municipality Data sources General directorate for technical services 3.2. Assessment of indicators Data sources for the calculation of indicators depend on the form and methodological approach of the indicator. Different methods of data collection are used for various purposes, such as (Limoges 2000): a) Experiments and desk-top processes, b) Conduction of questionnaire surveys, c) Observation, monitoring and measurement of traffic features, d) Use of Intelligent Transport Systems (ITS), e) Acquisition of data from information management systems, f) Combination of methods. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 170

179 SESSION: TRANSIT QUALITY PERFORMANCE Due to the absence of reliable data sources, a combination of methods was implemented for the collection of data for the city of Serres, i.e.: On-field observations and measurements of traffic volume and speed. Web-based data collection from official statistical databases, such as the Eurostat and the Hellenic Statistical Authority (ELSTAT). Scheduled meetings and interviews with the officials of the city s administrative departments, such as the municipality of Serres for the collection of maps and regulatory decisions, the Public Transport Authority of Serres (KTEAL) for information on urban bus services and the city s Traffic Police for the collection of accident reports. It is worth mentioning that the local authorities of Serres were able to provide only few organised data concerning the monitoring of the city s transport system. In order to cope with this deficiency, specific assumptions, such as homogenous yearly growth factors, were applied in order to bring the available primary data to the same base year and calculate the corresponding indicators. The results from the assessment of the indicators for the city of Serres are presented in Table 4. Table 4. Results from the assessment of indicators (time reference: 2013) I.D. Value I.D. Value I.D. Value Environmental Social Economic EN km/ km 2 SC km/ 1,000 inh. EC-1 11,200 EN km/ km 2 SC km/ 1,000 inh. EC vehicles/ 1,000 inh. EN % SC % EC-3 2.5% EN km/ km 2 SC-4 2,133 inh./ vehicle EC trips/ person/ year EN % SC % EC-5 12 km/ h EN years SC-6 0.0% EC EN-7 28% SC parking spaces/ 1,000 inh. EC-7 15 min. EN km/ l SC-8 13 accidents/ 10,000 inh. SC Normalisation of indicators 0 accidents/ 10,000 inh. The assessment of an integrated expression of the sustainable mobility conditions, i.e. the Sustainable City Index, requires the composition of the indicators that describe environmental, social and economic sustainability. In order to compose these indices using a common measuring system, the results of the indicators were ranked in a scale from 1 to 5. The range of values per score for each indicator is based on the findings of the Programmes: Urban Transport Benchmarking Initiative for medium-sized cities and EcoMobility Shift. Thus, the ranking system of the research is presented in Figure 3. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 171

180 SESSION: TRANSIT QUALITY PERFORMANCE Figure 3. Indicator ranking system Based on the ranking system, the scores of each indicator per field for the city of Serres are presented in Figure 4. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 172

181 SESSION: TRANSIT QUALITY PERFORMANCE Figure 4. Ranking scores of indicators Regarding the diagrams presented above it can be seen that only eleven (11) indicators out of twenty four get the lowest score (i.e. one) while only three (3) indicators are ranked to the highest level. The average score of the environmental category is 2.00 while the scores of the social and economic categories are 2.33 and 2.00 respectively. The weakest category is related to the sustainability of the public transport system with seven indicators ranked to the lowest score 3.4. Sustainable city index The normalisation of the indicators enables the assessment of the SCI. The initial step towards this purpose is the definition of weights which are attributed to each indicator category so as to express the gravity that environmental, social and economic sustainability holds in relation to the overall goal of sustainable urban mobility. In the international literature various methods are applied for defining these weights that can be divided into two main approaches, i.e. the use of equal weights and the use of an evaluation method, such as multicriteria analysis, in order to obtain weights (Singh et al. 2009; Castillo and Pitfield 2010). The main disadvantage of the first approach refers to its simplicity while the main disadvantage of the second approach refers to its subjectivity (Tanguay et al. 2010). In the specific research the method of equal weighting is chosen in order to avoid subjectivity deriving from different weights in each indicator. Thus, the SCI is calculated by the following equation: (1) Where: ien: score of each indicator of the Category: Environmental indicators nen: total number of indicators in the Category: Environmental indicators isc: score of each indicator of the Category: Social indicators nen: total number of indicators in the Category: Environmental indicators iec: score of each indicator of the Category: Economic indicators nec: total number of indicators in the Category: Economic indicators ntot: total number of indicators in all categories According to the value of the SCI, four different levels of sustainable mobility conditions can be identified (Van Dijk and Mingshun 2005): a) Sustainable 4, b) 4> Moderate >3, c) 3 Weak >2 and d) 2 Unsustainable. The SCI for Serres is 2.125/5, 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 173

182 SESSION: TRANSIT QUALITY PERFORMANCE suggesting weak sustainable mobility conditions for the city s transport system (Figure 5). Figure 5. Size and level of the Sustainable City Index 4. CONCLUSIVE REMARKS Within the framework of this paper, the sustainable mobility conditions of a Greek medium-sized city were evaluated using a representative selection of sustainable urban mobility indicators. The selection of indicators was based on the review of international literature as well as the knowledge of the specific local characteristics obtained through on-site observations of the study area. The assessment of indicators highlighted the lack of reliability and consistency in the available data and the necessity of conducting own measurements to bridge essential gaps. Thus, the present research confirms the already ascertainment made in previous research i.e. the need for developing an urban mobility observatory under an established framework by the European Union. In addition what is worthwhile to mention for the applied methodological approach in the specific research is the calculation of the composite Sustainable City Index (SCI) through the normalization of the considered indicators, which gives the possibility to evaluate the general mobility condition in terms of sustainability of the examined city. Concerning then the application in the city of Serres the result of the SCI indicates its weak level of sustainable mobility conditions and highlights the necessity of taking measures towards sustainability. In specific, looking through average scores of the indicators those concerning the sustainability of the public transport system are ranked to the lowest score, a fact that shows that the improvement of the quality of public transport services consists a matter of great importance for the city of Serres. Moreover, the synthetic analysis of all indicator categories suggests that focus should be also given on the available infrastructure for active transport, i.e. traffic calming and cycling networks. An opportunity for improvement towards the aforementioned aspects is the ongoing project for the urban regeneration of the city centre of Serres, which includes the implementation of traffic calming and pedestrian-friendly interventions. In this context, the proposed indicator system in the present research should be re-assessed for two periods after the completion of this project according to the guidelines of the corresponding technical reports, in order to investigate the improvement of the city s Sustainable City Index. Finally it should be pointed out that the specific research outlines also the framework for the establishment of a benchmarking mechanism under the assumption of a widescale implementation of coordinated Sustainable Urban Mobility Plans in Greek urban areas Thus this mechanism, with the support of local authorities and stakeholders, will enable the monitoring of the comparative progress in the conditions of sustainable urban mobility among cities of different size. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 174

183 SESSION: TRANSIT QUALITY PERFORMANCE References Atkinson, G., Dubourg, R., Hamilton, K., Munasinghe, M., Pearce, D., & Young, C. (1997). Measuring sustainable development: macroeconomics and the environment. Edward Elgar Publishing Ltd. Barrera-Roldán, A., & Saldıvar-Valdés, A. (2002). Proposal and application of a Sustainable Development Index. Ecological Indicators, 2(3), Castillo, H., & Pitfield, D. E. (2010). ELASTIC A methodological framework for identifying and selecting sustainable transport indicators. Transportation Research Part D: Transport and Environment, 15(4), Choon, S. W., Siwar, C., Pereira, J. J., Jemain, A. A., Hashim, H. S., & Hadi, A. S. (2011). A sustainable city index for Malaysia. International Journal of Sustainable Development & World Ecology, 18(1), ESPON. (2006). The Role of Small and Medium-Sized Towns (SMESTO). ESPON Final Report European Commission. (1990). Green Paper on the Urban Transport. COM(90) 218 final, Brussels, 27 June 1990 European Commission. (1992). White Paper - The future development of the common transport policy. COM(92) 494 final, Brussels, 2 December 1992 European Commission. (1995). Green Paper - The Citizens Network. COM(95) 601 final, Brussels, 29 November 1995 European Commission. (1998). Developing the Citizens Network. COM(1998) 431 final, Brussels, 10 July 1998 European Commission. (2001). White Paper - European transport policy for 2010: time to decide. COM(2001) 370 final, Brussels, 12 September 2001 European Commission. (2007). Green Paper - Towards a new culture for urban mobility. COM(2007) 551 final, Brussels, 25 September 2007 European Commission. (2009). Action Plan on Urban Mobility. COM(2009) 490 final, Brussels, 30 September 2009 European Commission. (2011). White Paper - Roadmap to a Single European Transport Area Towards a competitive and resource efficient transport system. COM(2011) 144 final, Brussels, 28 March 2011 European Commission. (2013). Together towards competitive and resource efficient urban mobility. COM(2013) 913 final, Brussels, 17 December 2013 European Commission. Directorate General for Energy and Transport. (2006). The Urban Transport Benchmarking Initiative. Year three final report. UTB3-A0-FINAL-REPORT.pdf, accessed March 31, 2014 Haghshenas, H., & Vaziri, M. (2012). Urban sustainable transportation indicators for global comparison. Ecological Indicators, 15(1), Lee, Y. J., & Huang, C. M. (2007). Sustainability index for Taipei. Environmental Impact Assessment Review, 27(6), Limoges, E., Purvis, C. L., Turner, S., Wigan, M., & Wolf, J. E. A. N. (2000). Future of urban transportation data. Transportation in the New Millennium. Litman, T., & Burwell, D. (2006). Issues in sustainable transportation. International Journal of Global Environmental Issues, 6(4), Litman, T. (2007). Developing indicators for comprehensive and sustainable transport planning. Transportation Research Record: Journal of the Transportation Research Board, 2017(1), nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 175

184 SESSION: TRANSIT QUALITY PERFORMANCE Litman, T. A. (2009). Sustainable transportation indicators: a recommended research program for developing sustainable transportation indicators and data. In Transportation Research Board 88th Annual Meeting (No ). Munier, N. (2011). Methodology to select a set of urban sustainability indicators to measure the state of the city, and performance assessment. Ecological Indicators, 11(5), Pitsiva-Latinopoulou, M. (2012). Indicators for sustainable mobility in urban areas. Newsletter, Hellenic Institute of Transportation Engineers, No. 182: Singh, R. K., Murty, H. R., Gupta, S. K., & Dikshit, A. K. (2009). An overview of sustainability assessment methodologies. Ecological indicators, 9(2), Tanguay, G. A., Rajaonson, J., Lefebvre, J. F., & Lanoie, P. (2010). Measuring the sustainability of cities: An analysis of the use of local indicators. Ecological Indicators, 10(2), Van Dijk, M. P., & Mingshun, Z. (2005). Sustainability indices as a tool for urban managers, evidence from four medium-sized Chinese cities. Environmental Impact Assessment Review, 25(6), Zheng, J., Garrick, N. W., Atkinson-Palombo, C., McCahill, C., & Marshall, W. (2013). Guidelines on developing performance metrics for evaluating transportation sustainability. Research in Transportation Business & Management, 7, accessed October 16, accessed March 30, accessed March 22, nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 176

185 SESSION: TRANSIT QUALITY PERFORMANCE Spatio-Temporal Parameters Affecting Sustainable Urban Mobility in Greek Medium Sized Cities: The Case of Volos A. Trampa Department of Planning and Regional Development, University of Thessaly, Volos, Greece tel , fax: Abstract This paper brings to attention spatio-temporal parameters affecting citizens daily travel behavior as well as citizens perception on travel time required for their daily mobility needs. Our aim is to shed light on the complexity and difficulties towards the integration of alternative mobility policies in Greek medium sized cities which, like most southern European cities, face great difficulties and barriers towards the integration of a sustainable urban mobility strategy. We focus mostly on spatialtemporal components affecting sustainable urban mobility and preferred transport modes in different residential areas (i.e. city center, inner or outer suburban, peri-urban areas). For this purpose, qualitative research was conducted in the Greek medium sized city of Volos, in order to estimate existing alternative mobility conditions in different residential areas and record citizens perception about time required for traveling by using different modes of transport on the same journey. The results of the research show a significant variation between different urban zones (i.e. city center, inner or outer suburban zones) regarding travel mode options as well as the low competitiveness of the existing public transport system (bus). The major issue is the perceived loss of time in almost all residential zones by those using the existing public transport system (bus network) compared with the use of other transport modes (private car, motorcycle, bicycle) on the same journey leading to the feeling that they are unable to meet their daily mobility needs. As a conclusion, we support the proposition that the selection and promotion of sustainable mobility policies should be informed by spatio-temporal characteristics and specific local conditions of urban space in order to achieve viable and acceptable solutions for better mobility in the city. 1. INTRODUCTION In the last decades motor vehicle use has grown rapidly, due to favorable technical and economic development with the cities unable to deal with traffic problems. Nowadays there has been a public awareness of environmental impacts and the loss of urban quality due to increasing use of private mode of transport (automobile). Urban sprawl is widely recognized as a political reality addressed by developers and governments. (V. Kaufmann, C. Jemelin, 2005). The debate on urban sustainable mobility seems to be primarily concerned with capturing and addressing the demand for regional travel needs associated with the formation of sprawling urban regions and edge cities, thus concentrating on urban form and land use issues (Ewing and Cervero, 2001). Even though modal share varies across space and time, private car is the predominant mode of transportation; even in cities in southern regions where walking is still a prevalent mode of travel at least for short 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 177

186 SESSION: TRANSIT QUALITY PERFORMANCE distances, even if at ever decreasing rates, as opposed to northern ones. The need to address car dependence through better infrastructure options is clear, but the urban design implications are not. Urban design is focusing on making urban centers accessible and more efficient in their functioning simply by facilitating car access, creating lots of problem in urban environment. Sustainable Urban Mobility (SUM) is key to European policy, setting the common goals of economic growth, prosperity and social equity together with the protection of the environment and the improvement of the quality of urban life. This strategy implies the management of traffic flows, the organization of public transport systems as well as the promotion of soft mobility (walking, cycling) especially in the city center. Combined urban transports, (cooperation between different transport possibilities private car /public transport/ walking / bicycle), is also of great importance, in order to achieve better mobility and the improvement of living conditions in European cities (E.C. Green Paper, 2007). Seeking urban sustainability, a major challenge has been set: this is the organization of alternative transportation in order to reduce automobile dependence and create more viable urban centers. (EC, 2007). The coordination between urban design and transport planning is also of great importance, as they form a unique system, complementary and competitive at the same time and therefore should never be seen separately. (Wiel, 2010). However, changing the modal share at the city level still remains an unmet objective. Several barriers to the implementation of urban sustainability are of great concern regarding the political acceptability of radical policies and a feeling of resentment from the general public. As has been proven, it is much easier to introduce policies which directly improve the quality of the urban environment than policies which are perceived as negative, but would also have a significant -but indirect- impact on the quality of city centers. (e.g. limitation of car access to the town centre). (Banister, 1998) Nowadays, urban mobility is associated with two main characteristics. Firstly, on a daily basis, work related trips are not the dominant ones, mainly because of changes in the locational relationship between employment and housing. Instead, what we observe is a continuous increase in trips related to other purposes (e.g. personal services to other, the school run and extracurricular activities, etc) to recreation, shopping which in total represent the majority of daily trips. Secondly, the geography of urban mobility has changed due to technological changes that through increased speed and improved transportation network have compacted time and allowed the dispersal of usually low density development in space, thus multiplying mobility and the use of private automobile at the regional level. On the other hand, acceptance of alternative mobility strategy requires changes. Various strategies have been proposed to arrive at a more sustainable transport system with a clear distinction between technological and behavioral changes. Technological solutions are aimed at reducing the negative impact per car and per km (by using hybrid vehicles or Intelligent Technologies (IT) for minimizing trips etc). Behavioral changes are aimed to influence car users decisions by shifting to alternative mode of transport 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 178

187 SESSION: TRANSIT QUALITY PERFORMANCE (public transport/walking/cycling) or by changing destination choices, combining trips, traveling less, etc). Extensive research on sustainable transport has shown that people usually prefer technological solutions (the use of energy efficiency cars) to behavioral changes. This is because they are directly linked to individual perceptions of mobility as well as quality of life issues reflecting the degree by which important human values and needs are met. But the mitigating environmental effects of new technologies tend to be overshadowed by the continuing growth of car use. (Steg and Gifford, 2005). Thus, the underlying rationale is that acceptance of sustainable transport strategies requires the achievement of balance between improvement in collective quality of life, such as environmental quality and fulfillment of individual interests such as comfort, safety and health. 2. SPATIO-TEMPORAL FACTORS AFFECTING URBAN MOBILITY Distance between place of residence and place of everyday activities (work, education of children, shopping, entertainment, other) / trip destination, is a basic parameter affecting travel mode choice, (mechanical or not), leading to the distinction between distance dependent or distance independent travel mode choice. Land use factors such as density 1, centrality or centeredness 2, land use mix and connectivity determine the amount and type of daily travel activity in the city. Centrality allows different activities to be grouped together at the centre of an area, thus reducing overall travel needs. The degree of proximity will also affect travel mode choice and travel duration. Central areas are usually characterized by greater density, mixed land uses and therefore, higher accessibility 3 and connectivity. Higher urban density results in a shorter journey length and maximizes the potential for soft mobility (cycling and walking). Once you get above a certain density, it is probable that you get less travel by car and more people using public transport or soft mobility (walking/biking). In dense urban areas, it is easier to organize and provide effective public transport services because they offer population density to support public transit. Many researches ( ) indicate that residents of central areas typically drive 20-40% less and they use soft mobility choices (walking and cycling) more often, than residents of other urban zones. Generally, at distances shorter than 2 km (acceptable length for walking) there is a substantial potential for walking, while most cycle journeys are usually shorter than 8 km. Therefore, most daily mobility needs could be carried out by these green modes (i.e. walking and cycling) (Banister 1998). Central areas should be viable, sustainable in the long-term, providing economic, social and environmental amenities. Viable centres cannot be car dependent because the domination of car (traffic, parking) defeats walk-ability and other vital functions that centres need. They should be complemented by other quality factors such as vitality, security, accessibility and a better urban environment. Thus, greater attention should be placed on creating a citizen friendly urban environment to improve safety and to facilitate soft mobility. Therefore the policy, of invisible infrastructure'(sully, 1 higher concentration of population 2 represents the concentration of jobs and related activities 3 people s ability to reach desired services and activities 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 179

188 SESSION: TRANSIT QUALITY PERFORMANCE 2005), proposes the reconstruction of road axes to ensure the sidewalk continuity, road redistribution in favor of alternative modes of transport, including traffic calming measures, parking restraint, speed reduction etc. According to research, although various land use factors have modest individual impacts, typically affecting just a low percentage of travelers, they are cumulative and synergistic and together they have a large impact on travel demand. (Litman, 2012). In inner or outer residential zones, usually characterized by lower urban densities and greater diffusion of residence, the organisation of public transport services gets more challenging and usually it does not present a cost effective solution. Thus, automobile use is the predominant mode of transportation and in many cases it represents the faster and cheaper mode of transport per mile or the only solution facilitating daily travel needs. Travel time needed for daily commuting is a crucial factor determining travel mode choice as the desire for rapid movement prevails, especially for work journeys, regarding the improvement of both effectiveness and competitiveness of human activities. But the value of time is an important component because there is a significant variation between different commuters (age, profession, socio-economic category) traveling for different purposes (work, leisure, shopping, other etc) leading to the distinction between work related or other purpose trips (recreational, educational, shopping related etc.) According to conventional transport analysis, travel time represents a real cost and, for this reason, it should be minimized. In contrast, Sustainable Urban Mobility approach indicates that traffic should slow down, especially in central urban areas, in order to create a safer and more citizen friendly urban environment. Therefore, the key policy objective is that of a reasonable travel time rather than the sole minimization of travel time. (Banister, 2008). The main issue is that of the use of public space simultaneously by different categories of users (automobilists and non-automobilists) creating conflicts and road accidents or collisions which reduce the level of the service of the existing transport network. The point is the fair share or more efficient use of public space between different groups of users as well as between various road functions (rapid link, access or parking). Further and more detailed analysis of a road network (road hierarchy / category of link / function) should be made in order to determine the basic role of every road section (ie. link between cities or access to the city center or neighborhood service) and set the priorities: road main usage, public transport priority, may exclude some road services (parking), or introduce a pricing policy. Spatial analysis can evaluate the quality of accessibility for specific groups and locations and overall accessibility can be evaluated with regard to time and money budgets. (Litman, 2012) 3. RESEARCH METHODOLOGY The assessment of a city s sustainable urban mobility efficiency is typically based on objective measuring (existing infrastructures) and analysis. However, citizens travel behaviour is a subjective and constantly changing process depending on several objective factors (like existing road networks, private vehicle ownership, public 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 180

189 SESSION: TRANSIT QUALITY PERFORMANCE transport network etc) as well as other subjective parameters such as the feeling of comfort or convenience, safety, personal mood, physical abilities, health, weather conditions etc. In this research, we try a qualitative approach of parameters affecting citizens travel behavior. We used as a case study, the medium sized city of Volos where we have carried out a qualitative research 4, with personal interviews of citizens living in different residential areas. For this reason, the city was divided, initially, into 3 zones (A, B, C) 5 (scheme in annex) depending on the distance and its connectivity 6 to the city centre. The aim of this research is to highlight the link between spatial parameters of different residential zones and travel mode preferences regarding different travel purposes (work, shopping, education children, leisure etc) as well as other determining factors affecting travel mode options. We focus mostly on residents perception on alternative mobility existing possibilities (public transport, cycling, walking) in different urban zones as well as travel time perceived for daily commuting, by using different means of transport (public transport / private car / or soft cycling, walking) on the same trip. In order to facilitate interviews and collect the adequate information, a short, semistructured questionnaire has been designed. This questionnaire consists of three sections: the first one contains questions relative to socio-demographic information (age, sex, income, vehicles ownership) including the place of residence and the place of the main daily trip (work, education, shopping etc). The second section focuses on the preferred modes of transport used for different travel purposes, including questions concerning the relevant travel time required and the factors influencing the residents travel mode choices. The third section concerns the problems encountered in terms of existent infrastructures, especially regarding alternative possibilities (public and private modes of transport) as well as necessary conditions for changing travel habits and shifting towards a greater use of alternative modes of transport. The sample used consisted of 75 individual reviews (25 interviews per zone) randomly assigned. It is not considered to be representative of the city s population, but only indicative of the prevailing tendencies. Statistical analysis, SPSS method was performed in order to record the degree of dependency between the use (or not) of a specific mode of transport and the residential zones, using the well-known Khi-square test (x2). This analysis has been implemented for each of the main travel purposes. (work, shopping, service of others, leisure, entertainment. According to Table 1, it is possible to observe some significant differentiation 7 between residential zones A, B, C., in travel mode options, according to the four main travel purposes considered. 4 based on a semi-structured questionnaire and excluding socio-economic components, due to insufficient data 5 zone A(Central Business District (CBD) and immediately adjacent areas, up to 2-3km), zone B(inner, suburb area, up to 4-5km), zone C(outer suburb, peri-urban, more than 5km ) 6 existing road and bus network is taken into consideration 7 The level of significance as regards the differentiation between the residential zones is given by the p- value with: p < 0, 01: very significant, 0, 01 < p < 0, 05: relatively significant and p >0, 05 not at all significant. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 181

190 SESSION: TRANSIT QUALITY PERFORMANCE 4. RESULTS Data analysis points to the predominance of private car use in almost all residential zones A,B,C, (Table 1), as well as very restricted use of existing public transport service (buses) within the city. In addition, we observe exclusive private car usage for special purposes such us: the service of others (driving children to school or after school activities, elderly activities etc), for week-end leisure activities, for entertainment and nights out, (car use for safety reasons due to lack of late bus lines etc) Furthermore the interpretation of Table 1 shows that there is a significant differentiation among different urban zones, in travel mode preference which is directly related to distance from the city center (zone A, B, C) as well as travel purpose. More specifically, residents in central areas (zone A) show a greater preference to alternative modes of transport for almost all travel purposes (work, shopping, entertainment), while they use a private car exclusively for special services (the school run). Sub-urban residents (Zone B) show a relatively greater preference to public transport (bus), for almost all travel purposes (work, shopping, and entertainment) but they also use a private car exclusively for the school run. Intensive private car use for all travel purposes is prevalent among residents in peripheral urban areas (peri-urban). This is an expected result justified by the relevant lack of mobility alternatives: less attractive public transportation 8 (inexistent or insufficient bus service, very few buses/day) and greater distances excluding soft mobility options (walking and cycling). 8 exception along main bus line 1,3 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 182

191 SESSION: TRANSIT QUALITY PERFORMANCE Table 1. Trip distribution per zone/ per purpose / per mode TRAVEL PURPOSE: WORK Mode of transport Zone A Zone B Zone C p-value* differentiation a Private car 48%(*) 48% 88% 0,004 Very significant b Motor-cycle 16% 24% 28% 0,587 not at all significant c Bicycle 24% 20% 8% 0,298 Relatively significant d Public Transport 16% 28% 4% 0,069 Relatively significant e Pedestrians 32% 12% 0% 0,005 Very significant TRAVEL PURPOSE: SHOPPING Mode of transport Zone A Zone B Zone C p-value differentiation a Private car 20% 28% 80% 0,000 Very significant b Motor-cycle 16% 20% 20% 0,916 not at all significant c Bicycle 32% 8% 8% 0,028 Relatively significant d Public Transport 4% 28% 16% 0,069 Relatively significant e Pedestrians 68% 48% 20% 0,003 Very significant TRAVEL PURPOSE: SERVICE OF OTHERS CHILDREN S EDUCATION Mode of transport Zone A Zone B Zone C p-value differentiation a Private car 80% 80% 80% - not at all significant b Motor-cycle 8% 12% 4% - not at all significant c Bicycle (no sufficient data) d Public Transport (no sufficient data) e Pedestrians 4% 8% - - not at all significant TRAVEL PURPOSE: LEISURE, ENTERTAINMENT Mode of transport Zone A Zone B Zone C p-value differentiation a Private car 88% 84% 88% 0,891 not at all significant b Motor-cycle 20% 20% 8% 0,409 Relatively significant c Bicycle 16% 4% 4% 0,196 Relatively significant d Public Transport 4% 8% 8% 0,807 not at all significant e Pedestrians 48% 28% 4% 0,002 Very significant (*) The values presented in the Table 1 correspond to the percentage of individuals in each zone which declare that they use the relative mode of transport. In other terms, if only 48% of individuals are using the car in zones A and B, this percentage reached 88% in the zone C. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 183

192 SESSION: TRANSIT QUALITY PERFORMANCE Comparing estimated average travel time required for daily commuting (for main travel purpose), with the use of public transport or private cars (for the same trip/itinerary), travel time appears consistently more when using public transport (bus) compared to car use for the same trip and this difference is statistically significant in all residential zones. Table 2. Comparison of perceived waste of time by using bus or private vehicle ZONES Paired Differences Std. Error Mean 95% Confidence Interval of the Difference Lower Upper t df Sig. (2- tailed) Mean (min) Std. Deviation A -11,9 7,539 1,729-15,581-8,314-6, ,000 B -16,8 6,480 1,449-19,783-13,717-11, ,000 C -17,5 8,259 2,065-21,910-13,109-8, ,000 More specifically, the average loss of time for travelers using public transport (bus) compared to car use (on the same trip), amounts to approximately 12min/ journey within the central zone (A) and around 17min/ journey in other residential zones (B, C), while in theory it can be considered double (24 or 34 min respectively) per round trip. This feeling of wasting time is considered to be very significant and discourages citizens from opting for public transportation, especially when it comes to successive, multipurpose traveling with many stopovers, or for nights out etc. Certainly, the definition of a long commute varies across people and communities for example, in medium sized cities or in metropolitan areas. According to ACS (American Community Survey Reports) the 60 min travel time threshold is also roughly twice that of metro areas with the longest average travel times, which exceed 30 min. In 2011 workers in the N.Y. City metro area and Washington DC metro area had the 2 longest average travel time among metro areas, at 34,9 min and 34,5 min respectively. ( (see figure and table in annex). To summarise results regarding citizens perception of alternative mobility possibilities in the city, we recorded the feeling of depreciation towards public transportation and people prefer avoiding losses. Bus service is considered as a 'low level public service in the city because of low frequency, long delays and an incomplete network but it is also viewed as an expensive travel mode because of a lack of ticket policy (correspondence between different bus lines on the same ticket or ticket reduction for families, etc). Soft mobility alternatives (walking or cycling) is seen in a positive light by the majority of interviewees especially because of the city s center urban morphology (sea front, flat ground) which facilitates and attracts pedestrians or cyclists, especially under good weather conditions (generally mild temperature). But, as it has been pointed out by the citizens themselves, there are many obstacles and barriers (such as narrow 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 184

193 SESSION: TRANSIT QUALITY PERFORMANCE sidewalks, incomplete, fragmentary or unsafe infrastructures for pedestrians and cyclists), leading to a feeling of insecurity which discourages potential users. Particular emphasis was placed on the poor quality of public spaces and sidewalks, which are often occupied by hindrances (goods, rubbish bins) or illegally parked vehicles. Unlawful parking seems to be an acceptable practice by authorities, who declare unable to control total urban area. Finally, poor alternative mobility opportunities in the city, complexity of daily commuting, (successive different destinations, services to other persons etc) as well as time required for daily travels determine travel mode choice. Even in today s harsh economic situation, which creates considerable stress and uncertainty, many people continue to use private cars for everyday mobility needs because they lack alternatives. But, as respondents have declared, they have mostly abandoned leisure travel on week-ends. Concerning the necessary conditions for changing mobility practices, citizens declare that they are positive and willing to make changes in their daily mobility habits, only in the event of well-organized alternative solutions, ameliorate and safer network for pedestrians and cyclists, time and cost effective public transport (better frequency, network, ticket etc). 5. CONCLUSIONS Greek medium sized cities, like most southern European cities, face great difficulties regarding the integration and implementation of a sustainable urban mobility strategy. Nowadays, we face a serious economic recession with a significant reduction of family income together with an explosion of fuel cost. Thus, there is a widespread desire to adopt policies and find solutions ensuring that people do not have to rely on a private car and have the burden of fuel and from the other hand not to be deprived of the right to travel as a basic element of personal progress, productivity. For this reason, transport planning should be flexible and responsive to continuous growing travel demands by adopting multi-modal transport planning which should be informed by specific local characteristics especially those related to the use of public space as this determines mobility practices and needs. Changing modal share in favour of alternative mobility could never be achieved unless we manage to offer time & cost effective alternative solutions. Spatial analysis as well as time distance criteria are crucial to determine priorities in urban transport planning. The backbone and top priority to achieve sustainable urban mobility in all urban zones is the organization of a time and cost effective public transport system which should be able to guarantee seamless connections between the city center and inner or outer urban zones. Central zones allow several possibilities for private car-independency and the replacement of car use is possible because there is a large proportion of activities that can be easily reached by alternative mobility. Greater attention should be placed on creating a citizen friendly urban environment, by improving safety and facilitating soft mobility (pedestrians and cyclists) together with dissuasive measures for car use(speed reduction, parking charges etc). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 185

194 SESSION: TRANSIT QUALITY PERFORMANCE In inner sub-urban or outer peripheral zones characterized by relative distancedependent mobility, facilitating multi-modal mobility and increasing the effectiveness of public transport services could have a significant impact on modal shift at the city level. Several measures to improve the effectiveness of public transport such as giving buses priority, the introduction of exclusive bus-lanes combined with parking facilities on public transport junctions are of great importance. References Banister, D. (1998) Barriers to the implementation of urban sustainability International Journal of Environment and Pollution, Vol 10, No 1, pp Banister, D. (2008) the sustainable mobility paradigm, Transport Policy, no 15, pp European Commission (2007) Green Paper Towards a new culture for urban mobility, (online) Kaufmann V. et Jemelin C., Articulation entre urbanisme et transport : quelles marges de manœuvre?, Revue international des science sociales 2003/2, No 176, p Litman, T., and Steel R. (2012): Land use impacts on Transport: How land use factors affect travel behaviour, VTPI, (online) Litman, T., (2012): Evaluating Accessibility for Transportation Planning, VTPI, (online) Litman, T., (2013): The future isn t what it used to be VTPI, (online) Sally A. (2005) Invisible Infrastructure Steg L., and R. Gifford (2005) Sustainable transportation and quality of life, Journal of Transport Geography, 13: U.S. CENSUS BUREAU, B. Mc Kenzie, 2013 Out of state and long commutes: Wiel M., (2010 ) Etalement urbain et mobilité, La documentation française Statistique.developpement-durable.gouv.fr, 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 186

195 SESSION: TRANSIT QUALITY PERFORMANCE Annex City of Volos. Division in 3 zones 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 187

196 SESSION: TRANSIT QUALITY PERFORMANCE Figure 1 shows that the national average travel time fluctuated little between 2000 and 2011, with the national average travel time of 25, 5 min in 2011 Table 1 lists the distribution of commuting times across several intervals in the N.Y. City metro area, in As we observe travel time interval of 58% of workers is below 24 min. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 188

197 SESSION: TRANSIT QUALITY PERFORMANCE Environmental Impact Assessment of Urban Public Transport Systems: a Scenario Study for the Northwest Part of City of Volos Lambrini Papafoti*, Konstantinos Papoutsis, Eftihia Nathanail Transportation Engineering Laboratory, Department of Civil Engineering, University of Thessaly, Volos, Greece *Corresponding author: lambrinp@gmail.com, Tel: Abstract Urban areas represent a significant challenge regarding both environmental impacts and traffic congestion. Residents and visitors of cities require high level of service in their travelling and local communities strive for respect to environment, social equity and economic balance. Thus, sustainability in urban transport becomes an overarching concern for a city s policy-makers. In this paper we suggest an economically feasible scenario for restructuring the urban public transport network of the northwestern area of city of Volos, while also evaluating the environmental impacts of several alternative green bus fleets in order to achieve an economically effective and environmentally efficient transport service, both for the operator and local community. This is achieved by the most suitable models and software available in the literature. The most efficient bus fleet is then subject to cost-benefit analysis in order to ensure the economic viability of such investment. Sensitivity analysis is conducted in order to estimate the fluctuation range of profit in case of volatile economic environment. In this regard, a sound and sustainable operational model is suggested for the public transport service of the northwestern area of city of Volos. Keywords: urban public transport, environmental impact assessment, Cost-Benefit analysis, Volos. 1. INTRODUCTION The impact of transportation on the environment is multidimensional; it consumes energy, generates noise, pollutes the air, land and water, and consumes materials and land. Current vehicle technologies are mainly petroleum based but those resources are not replenished at higher rates than they are consumed, therefore have a strong negative impact (Mitropoulos, 2011). Since the early 1990s, a series of strict emission standards which are also known as Euro emission standards were progressively introduced across the European Union. These standards define the permitted limits for exhaust emissions of new vehicles sold in the European Union member states. These Euro standards have encouraged vehicle operators to reduce the emissions of their older vehicles either by repowering or by fitting emission reduction after treatment equipment. The introduction of cleaner vehicles comes at a higher cost and technological risk (Wall et al., 2008). The legislation focuses on the emissions of carbon monoxide (CO), hydrocarbons (HC), Nitrogen oxides (NOX), Particulates (PM10) as well as on smoke. It is of note that carbon dioxide (CO2) emissions are not included in the Euro standards. Although CO2 poses no health risks, it has been linked to global warming (Wall et al., 2008). As far as heavy- 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 189

198 SESSION: TRANSIT QUALITY PERFORMANCE duty vehicles are concerned, the current Euro standard is Euro VI. It should be noted that the Euro standard EEV (Enhanced Environmentally-friendly Vehicle) is a nonbinding standard ( Table 1. Euro emission standards for heavy duty engines (g/kwh) Euro Euro Euro I Euro II Euro III Euro IV Euro V EEV VI (1993) (1996) (2000) (2006) (2009) (1999) (2014) Carbon monoxide (CO) 4,50 4,00 1,00 2,10 1,50 1,50 1,50 Hydrocarbons (HC) 1,10 1,10 0,25 0,66 0,46 0,46 0,13 Nitrous oxides (NOX) 8,00 7,00 2,00 5,00 3,50 2,00 0,40 Particulate matters (PM10) 0,36 0,15 0,02 0,10 0,02 0,02 0,01 Smoke (m -1 ) - - 0,15 0,80 0,50 0,50 - The pollutants which are emitted by vehicles can have a harmful effect on both people s health and the environment as a whole. Tailpipe emissions are pollutants released directly from vehicle exhaust pipes. Motor vehicles produce various harmful air emissions, some of them are responsible for the greenhouse effect while others contribute to various health problems. A list of the most common pollutants emitted by vehicles and their effects are discussed here (Wall et al., 2008; EPA, 2009; EEA, 2012). Carbon monoxide (CO) is a colourless, odourless and tasteless but highly toxic gas. Carbon dioxide (CO2) comes from the burning of carbon fuels and is a major contributor to global warming. Methane (CH4) is also a greenhouse gas that is over 25 times more effective in trapping heat in the atmosphere than carbon dioxide (CO2) over a 100-year period. Nitrogen oxides (NOX) consist of various compounds, some of them are toxic but all of them contribute to ozone. Nitrous oxide (N2O) is a greenhouse gas and has 298 times more impact per unit weight than carbon dioxide over a 100-year period. Volatile organic compounds (VOC). Particulate matters (PM2,5, PM10) are the residues of incomplete burning of hydrocarbon fuels. Hydrocarbons (HC) are produced with the burning of hydrocarbon fuels. Non-Methane Volatile Organic Compounds (NMVOC). In the white paper European transport policy for 2010: time to decide it is mentioned that the increased traffic and urban congestion have resulted in air and noise pollution. Frequent short journeys made with the engine cold increase fuel consumption exponentially, and emissions may be three or four times higher while traffic speed is three or four times slower. In the green paper of 2007 it is proposed that the introduction of clean and energy efficient vehicles could be supported by green public procurement. A possible approach could be based on the internalization of external costs by using life-time costs for energy consumption, CO2 emissions, and pollutant emissions linked 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 190

199 SESSION: TRANSIT QUALITY PERFORMANCE to the operation of the vehicles to be procured as award criteria, in addition to the vehicle price (European Commission, 2007). Based on all the above mentioned, our main objective is by suggesting a balanced, economically viable scenario of re-designing two bus routes that serve the Northwest areas of the city of Volos and introducing a new one and evaluating four scenarios of alternative green bus fleets, to suggest a feasible, efficient and environmentally friendly operational model. The routes that serve the Northwest areas have been redesigned so that more dwellers of the area use the public means of transportation. Firstly, different fleets of buses are proposed and they are evaluated based on the amounts of pollutants they emit. The produced emissions are estimated using the calculating program COPERT4. A cost benefit analysis is conducted for the fleet of buses and the data which are needed for the cost-benefit analysis are obtained by the literature. We estimate the capital, the operating, the maintenance costs as well as the labor costs in order to investigate the economic viability of the investment. Finally, a sensitivity analysis is used in order to assess the impact of potential changes in ridership and fuel price on total profit. 2. SCENARIOS DEVELOPMENT 2.1. The city of Volos and basic transport characteristics The scenarios that are tested refer to the public transport system of city of Volos, Greece which is a coastal city located almost in the middle of the country. The conurbation of Volos covers an area of 385,614 sq. km² and it is administratively located in the southeast of the Prefecture of Thessaly. The city of Volos has a centroidal character as it is very close to the salient transport axis of Greece, being almost equidistant to Athens, the capital of Greece, and Thessaloniki, the second biggest city. Main spatial and transport particularities are the port of Volos, the vicinity to the main road axis (P.A.Th.E.) and the international airport of Nea Aghialos (Municipality of Volos, 2012). The city is identified as mono-centric where the most and important operations are deployed within the historic center. The land uses are basically mixed either in the urban core or the suburban areas. The industrial area of the city is spotted in the west, close to road and rail infrastructure. The majority of economic and commercial operations are taken place within the historic center. Any state agencies are also located within the historic center. Any small scale commercial operations are pinpointed in the center of the city, mostly along Ermou str., where other private and commercial services are provided (banks, offices, infirmaries, etc.). The coastal front of the city and Ag. Nikolaos are considered as leisure areas whereas Portaria, Makrinitsa and Iolkos are known as the main touristic places around the city. Finally, transport infrastructure inside the city includes the railway station and urban-interurban bus station both located at Sekeri str. close to each other. The population of the city wider area is ( Census). The growth in the population of the city of Volos was slightly positive across the last decade. The population in the municipal district of Volos has been increased by 1,23 % between 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 191

200 SESSION: TRANSIT QUALITY PERFORMANCE whereas the increase in the wider conurbation of the city reaches 1,07% (A) ( Public transport service The passenger transport service in the city of Volos is undertaken by ASTIKO KTEL VOLOU SA, a company under private law employing 54 vehicles. The transport service is supported through 15 public transport lines covering a wide area of destinations in the conurbation of Volos. The lines feed the northwest part of the city are No 2 (Central Station Ampelokipi) and No 9 (Central Station Chiliadou). The restructuring anticipated new design for the lines and new fleet scheduling and routing. This study investigated several scenarios of transport service in the northwest area of the city using lines No 2 and No 9 and the introduction of a local bus line in the north of the city, which will operate in combination with existing bus lines. It is believed that the northwest area of the city of Volos is not efficiently served and there are areas which indicated overlapping in transport service and other areas are not covered at all. The criteria that fostered the planning and the restructuring of the transport service were (Chakroborty and Dwivedi, 2002): the satisfaction of total transport demand or the highest part of it, tackling the needs for transit and saving time (and money) for transport. Τhere is also an array of criteria that fall into the design framework of new public transport service, and that are (Karlaftis and Lymperis, 2009; Frantzeskakis et al., 1997; Transportation Research Board, 1995; Giannopoulos, 1994): Lines density Population density Bus stops location Spatial deviation from the primary line Avoidance of overlapping Descent passenger flows Length of bus route Satisfactory road infrastructure and geometry 2.2. Basis scenario The study examined several scenarios for the lines No 2 and No 9. Regarding the particularity of this case-study, there was an additional concern which comprised the economic viability of the new service. In this respect, the final outcome should have implied operational profits. After various sensitivity checks between the scenarios and selected sensitivity criteria, such as volatility in passenger flows or fluctuation in fuel price, a final, sound scenario was produced as a basis for the analysis in this paper (Nathanail, 2014). The basis scenario constitutes three bus lines: 1. The bus line No 2 follows the route: Pedion Areos Lampraki Iasonos Kartali Analipseos Pagason Anapafseos Merkouri Ag. Nektariou Fitokou 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 192

201 SESSION: TRANSIT QUALITY PERFORMANCE Panepistimio and for the return Panepistimio Fitokou Ag. Nektariou Merkouri Anapafseos Pagason Analipseos Venizelou Dimitriados Lampraki Pedion Areos. 2. The bus line No 9 follows the route: Pedion Areos Lampraki Kartali Dimou Paraskevopoulou Fitokou Nea Ionia Swimming Centre (Natatorium) and for the return Ag. Dionisiou Paraskevopoulou Dimou Venizelou Dimitriados Lampraki Pedion Areos. 3. The local bus line follows the route: Makariou Panagouli Iatridi Ag. Paraskevis K. Varnali Naxou Ag. Kuriakis K. Varnali Ag. Paraskevis Iatridi. Table 2. Routes length. Length (in kms.) Outward Return Line 2 6,10 6,27 9 5,90 5,82 Local 3,66 (Source: Nathanail, 2014) The service frequency for the line No 2 is every 12 minutes for the typical weekdays, for the line No 9 the frequency varies over the day depending on peak hours, and for the local bus line the frequency is intertwined with line No 4 as these two lines are complementary. For this scenario, the expected annual passenger demand of the bus lines No 2, 9 and local for outward and return runs counts cumulatively passengers. Table 3. Number of runs for the bus lines No 2, No 9 and local one under the basis scenario. No of runs Day Outward Return Bus line 2 Typical weekday Saturday Sunday Typical weekday Saturday Sunday Local Typical weekday 43 Saturday 40 Sunday 36 (Source: Nathanail, 2014) 2.3. Analysis of the different scenarios of bus fleets Five different scenarios of fleets of buses are proposed for the operation of the redesigned lines. The emissions produced by the potential implementation of each scenario are calculated. The five scenarios examined here are: 1. Scenario of operation using the current fleet of buses: Operation of the redesigned lines 2 and 9 using a fleet of buses that comprises of 54 diesel-fueled midi buses (16 conventional buses, before the imposition of Euro emission standards, 26 Euro I buses, 11 Euro II buses, 1 Euro I bus) and operation of the new local line using a Euro III diesel-fueled mini bus. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 193

202 SESSION: TRANSIT QUALITY PERFORMANCE 2. Alternative scenario of operation (a): Operation of the redesigned lines 2 and 9 using Euro VI diesel-fueled midi buses, and similarly, operation of the new local line using a Euro VI diesel-fueled mini bus. 3. Alternative scenario of operation (b): Operation of the redesigned lines 2 and 9 using Euro VI biodiesel-fueled midi buses and operation of the new local line using a Euro VI diesel-fueled mini bus. 4. Alternative scenario of operation (c): Operation of the redesigned lines 2 and 9 using Euro EEV compressed natural gas-fueled (CNG) midi buses and operation of the new local line using a Euro VI diesel-fueled mini bus. 5. Alternative scenario of operation (d): Operation of the redesigned lines 2 and 9 using hybrid Euro VI diesel-fueled midi buses and operation of the new local line using a hybrid Euro VI diesel-fueled mini bus. It should be noted that all the scenarios include diesel-fueled mini buses solely because we are not able to estimate emissions produced by alternative fueled vehicles using the calculating program COPERT4. However, this was not a hurdle in the assessment process of the different fleets of buses, since the use of hybrid buses (either midi or mini) seems to lead to the lowest emission factors (emissions per kilometer). 3. METHODOLOGICAL APPROACH FOR THE EVALUATION OF THE ACQUISITION AND OPERATION OF ALTERNATIVE BUS FLEETS 3.1. Environmental impact assessment As it has already been mentioned the air pollutants and the greenhouse gas emissions by each fleet of buses are calculated with COPERT4. All the calculations are made for a one-year-period. The data that need to be imported so as to estimate the pollutants produced by the operation of the lines using the different fleets of buses are: 1. Fuel specifications 2. Euro emission standard of each bus 3. Type of fuel used by each bus 4. Number of buses 5. Annual kilometers driven (km) 6. Average speed (km/h) 7. Average occupancy (average passengers per maximum capacity) 8. The environment of the road (urban, rural, highway) The average speed of the midi buses that are used for the redesigned lines 2 and 9 is assumed to be 15 km/h while the average speed of the mini bus is assumed to be 18 km/h (Μορφουλάκη και Κoτούλα, 2011). In the case where we calculate emissions produced by the operation of the lines using the current fleet of buses, the number of buses is already known, and therefore, we assume that each bus contributes equally to the total annual kilometers driven. However, for the examination of the alternative scenarios we need to define the number of buses that are needed for the operation of the three lines. The number of buses is calculated based on the frequency and the average speed of each line. Thus, it is concluded that five and four buses need to be purchased and used for the operation of lines 2 and 9, respectively, while one bus is needed for the new local line. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 194

203 SESSION: TRANSIT QUALITY PERFORMANCE A previous broad survey took place during summer 2013 and included stated and revealed questionnaires addressed to passengers and evaluation of quality indicators regarding public transport (Papafoti, 2013). Among the most interesting findings of this survey was the fact that in case of the introduction of environmentally friendly buses the 14% of the respondents would intend to increase the use of public transport modes. Another study aiming at restructuring the urban passenger transport service provided by ASTIKO KTEL VOLOU SA was elaborated in autumn 2013 (Nathanail, 2014). Except for the air pollutants and the three greenhouse gases, CO2, CH4, N2O, we also estimate the total greenhouse gas emissions based on the global warming potential of each greenhouse gas relative to CO2, which is 1, 25 and 298, respectively Economic analysis The cost-benefit analysis investigates the socio-economic balance between costs and benefits that are generated by the purchase and operation of a new, hybrid bus fleet: one small vehicle for the operation of the local bus line and nine midi vehicles for the operation of bus lines No 2 and No 9. In order to perform a cost benefit analysis it is vital to capture all the sources of expenses and all the revenue streams in a given period of time. The cost of acquisition of the new fleet (purchase cost) is a lump sum cost that is embedded in this process. For the sake of analysis, the service life span of the fleet was determined in 12 years (Bartin, 2013; Lajunen, 2014). However, according to the national legislation No 2963/2001, bus vehicles could be scrapped after 23 years since manufacturing. Nevertheless, the analysis is strictly limited to 12 years. Beyond this threshold it is assumed that the scrap value, namely the resale value, is zero. The assumptions that were adopted for the estimation of the total costs (one-off costs and operational expenses) incurred in the introduction of new fleet are: Table 4. Assumptions adopted towards the estimation of operational costs for the new fleet Characteristics of a Fuel consumption (lt./100 km.) 23,59 middle-sized hybrid Total capacity of a midi vehicle (seats) 100 vehicle Acquisition cost ( ) Tire market price ( ) 350 Average speed (km/hr) 15 Characteristics of a Fuel consumption (lt./100 km.) 15,62 small-sized hybrid Total capacity of a mini vehicle (seats) 20 vehicle Acquisition cost ( ) Tire market price ( ) 350 Average speed (km/hr) 18 Fuel Fuel price ( /lt) 1,37 Maintenance Average rate (up to km) of the annual maintenance 8,04% cost to the acquisition cost of a conventional bus Tire wear Annual number of tires that need replacement (per ,0793 km) Wage costs Hourly wage of the driver ( ) 12 (Sources: Μορφουλάκη και Κωτούλα, 2011 ; Transportation Research Board 2009; Chandler and Walkowicz, 2006; Lajunen, 2014) Cost factors 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 195

204 SESSION: TRANSIT QUALITY PERFORMANCE The cost categories for a hybrid vehicle are described and estimated below: Acquisition cost: Hybrid vehicles operating with diesel are 30-70% more expensive than the conventional ones. According to several estimations, it could be assumed that the upfront acquisition cost for a hybrid bus is 50% more expensive compared to a conventional Euro VI bus (Lajunen, 2014). The cost of a conventional mid-sized bus is and the cost of a conventional mini bus is Therefore, the cost of a hybrid midi-bus is estimated to C acq = and for a hybrid mini bus to C acq = (1) Fuel costs: It is estimated that hybrid mid-sized and mini buses consume 27% and 23 % less fuel compared with the similar conventional ones (Transportation Research Board, 2009). Total fuel costs include an additional safety net of 10% to incorporate any delays, discrepancies, etc. (Nathanail, 2014). Then, the annual fuel cost is estimated as follows: C fuel = 110%*fuel consumption per 100 kilometers * number of annual vehicle kilometers *diesel fuel price/100 (2) Tires wear costs: There is no difference between conventional and hybrid vehicles. As such: C tw = 0,0793* number of annual vehicle kilometers *cost of a tire/1000 (3) Maintenance costs: The maintenance cost of a hybrid vehicle is 4% lower than in conventional vehicle (Chandler and Walkowicz, 2006). Consequently: C main = acquisition cost of a conventional diesel vehicle*0,0804 (4) Wage costs: Estimated exactly with the same way as the conventional vehicles: C wage = hourly wage* number of annual vehicle kilometers/average speed (5) As far as revenues are concerned, the only source of revenues is the bus tickets that are purchased. The estimated revenues that stem from the expected passenger flow are calculated as follows: due to the different categories of tickets (civil-student, in-bus or pre-purchased), an analysis was made on the ticket data which were provided by the transport provider. A weighted average ticket price was produced taking into account for a specific ticket category: the amount of tickets purchased the price for each category Finally, the average price that was calculated is 1,037 per passenger. Apart from tickets, there are monthly passenger cards providing access to public transport modes that are purchased by passengers creating additional revenue. The total revenues that are generated from monthly passenger cards are factored within the rest revenue sources producing an updated value that reflects the average price per ticket. It was estimated that the revenues from monthly passenger cards represent the 2,53% of 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 196

205 SESSION: TRANSIT QUALITY PERFORMANCE total revenues. In this regard, the total revenues are estimated with the following formula: R total = (number of passengers*average ticket price) + 2,53% * (number of passengers * average ticket price) (6). 4. RESULTS 4.1. Environmental impact 1. Scenario of operation using the current fleet of buses: The operation of lines 2 and 9 is conducted using 54 diesel buses. The annual kilometers driven by each bus for the lines 2 and 9 are 5.383,99 and 4.785,23 respectively. The corresponding occupancies are 10,8% and 9,4%. Lastly, in the case of the new local line the annual kilometers are ,12 and the occupancy is 19,9%. Table 5. Calculated emissions (kg) per year for the scenario of operation using the current fleet Pollutants Line 2 Line 9 New local line CO 1.363, ,70 124,39 CO , , ,37 CH4 46,87 41,65 0,00 NOX 4.363, ,95 507,02 NO 3.881, ,43 436,04 NO2 482,16 427,52 70,98 N2O 5,01 4,45 0,00 PM2,5 204,36 181,31 11,84 PM10 212,87 188,79 13,57 PM exhaust 196,70 174,58 10,28 VOC 503,10 447,51 27,58 NMVOC 456,23 405,86 27,58 2. Alternative scenario of operation (a): It is assumed that the occupancies for each line are increased by 14%, similarly to all the alternative scenarios. The annual kilometers per vehicle are ,13 (line 2) and ,64 (line 9), as in all the following scenarios. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 197

206 SESSION: TRANSIT QUALITY PERFORMANCE Table 6. Calculated emissions (kg) per year for the alternative scenario of operation (a) Pollutants Line 2 Line 9 New local line CO 530,15 470,66 68,31 CO , , ,93 CH4 1,53 1,36 0,00 NOX 420,59 376,21 50,32 NO 378,53 338,59 45,29 NO2 42,06 37,62 5,03 N2O 12,07 10,72 0,00 PM2,5 9,52 8,37 1,82 PM10 18,12 15,93 3,59 PM exhaust 1,77 1,57 0,23 VOC 12,61 11,20 1,56 NMVOC 11,08 9,84 1,56 3. Alternative scenario of operation (b): COPERT4 considers operation of the buses at high biodiesel blending ratios (up to 30%). Emissions of all pollutants except NOX appear to decrease when biodiesel is used. The fact that NOX emissions increase with increasing biodiesel concentration could be a detriment in areas that are out of attainment for ozone (Demirbas, 2007). Table 7. Calculated emissions (kg) per year for the alternative scenario of operation (b) Pollutants Line 2 Line 9 New local line CO 482,44 428,30 68,31 CO , , ,93 CH4 1,53 1,36 0,00 NOX 435,31 389,38 50,32 NO 391,78 350,44 45,29 NO2 43,53 38,94 5,03 N2O 12,07 10,72 0,00 PM2,5 9,25 8,13 1,82 PM10 17,86 15,69 3,59 PM exhaust 1,51 1,33 0,23 VOC 10,72 9,52 1,56 NMVOC 9,19 8,16 1,56 4. Alternative scenario of operation (c): CNG and diesel bus emissions of regulated pollutants have become increasingly similar over the last decade. As far as the greenhouse gases are concerned, methane emissions are the greatest greenhouse gas concern in the case of CNG vehicles (TRB, 2011). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 198

207 SESSION: TRANSIT QUALITY PERFORMANCE Table 8. Calculated emissions (kg) per year for the alternative scenario of operation (c) Pollutants Line 2 Line 9 New local line CO 365,37 324,74 68,31 CO , , ,93 CH4 284,92 253,24 0,00 NOX 1.406, ,19 50,32 NO 1.350, ,18 45,29 NO2 56,27 50,01 5,03 N2O 0,00 0,00 0,00 PM2,5 11,54 10,18 1,82 PM10 20,05 17,66 3,59 PM exhaust 3,88 3,45 0,23 VOC 355,91 316,33 1,56 NMVOC 70,99 63,09 1,56 5. Alternative scenario of operation (d): COPERT4 does not enable the estimation of emissions produced by hybrid buses. Hybrid electric vehicles have two forms of drive power: electrical power, and another power source from a combustion engine. The engine and the electric motor compose the mix dynamic system to actuate the vehicles travel together, causing the noxious gas in vehicles' fuel consumption and the waste gas discharges to reduce massively (Houyu and Guirong, 2011). The reductions in the exhaust emissions of hybrid buses are associated with reduced use of the combustion engine (TRB, 2011). Therefore, they are associated with the reduced fuel consumption. In the TRCP (2009) two fuel economy functions for diesel and hybrid diesel buses are proposed as functions of speed. Using these equations it is concluded that hybrid midi and mini buses consume 27% and 23% less diesel, respectively, comparing to conventional buses. We make the assumption that in this scenario the annual driven kilometers decrease as much as the fuel consumption does. Consequently, the annual kilometers driven by each bus for the lines 2 and 9 are ,40 and ,47 respectively. The corresponding kilometers which are driven by one mini hybrid bus for the operation of the new line are ,00. Table 9. Calculated emissions (kg) per year for the alternative scenario of operation (d) Pollutants Line 2 Line 9 New local line CO 387,01 343,58 52,60 CO , , ,72 CH4 1,11 0,99 0,00 NOX 307,03 274,63 38,75 NO 276,32 247,17 34,87 NO2 30,70 27,46 3,88 N2O 8,81 7,83 0,00 PM2,5 6,95 6,11 1,40 PM10 13,23 11,63 2,76 PM exhaust 1,29 1,14 0,18 VOC 9,21 8,17 1,20 NMVOC 8,09 7,18 1,20 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 199

208 SESSION: TRANSIT QUALITY PERFORMANCE 4.2. Economic analysis In order to perform economic appraisal of such investment, it is crucial to calculate the expected revenue streams and the costs that are incurred. The costs that are allocated to this basis scenario are depicted in the table below. Table 10. Total and partial annual costs for the operation of bus lines No 2, 9 and local Bus line Costs/year Fuel Tire wear Maintenance Wages 2-Outward , , , ,3 2-Return , , , ,2 9-Outward , , , ,6 9-Return , , , ,5 Local line , , , ,1 Total , , , ,6 (source: Papafoti, 2013) The total annual costs reach and the total demand counts passengers. Taking into account the fact that 14% of the respondents in a recent survey stated that they would surely use public transport instead of other transport modes or increase their use of public transport, the forecasted demand that might use public transport operated by hybrid fleet reaches Ds=1,14 * = passengers annually. According to (6), the total annual revenues generated by the passenger flow of become Rtotal = (Papafoti, 2013). The economic indicators that were investigated within the context of the appraisal of the investment of purchasing and operating new hybrid fleet of buses are: Net Present Value (NPV) Cost-Benefit ratio (B/C) Internal Rate of Return (IRR) Net Present Value: within this framework, the present value of all future revenues and costs that are incurred during the service lifespan of hybrid fleet is estimated. The Net Present Value is estimated as follows: Whereas B0 and C0 are the benefits and costs arising by the time of adoption of a new option and they are one-off. Bi stands for the benefits in the year i and Ci stands for the cost in the year i. Another data that should be incorporated is the expected increase rate of the population which according to (A) is 0,11% per year. In order to estimate the NPV we assume that the discount rate is 5% and for a period of 12 years the NPV is ,7. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 200

209 SESSION: TRANSIT QUALITY PERFORMANCE Cost-Benefit ratio: This method regards the calculation of the ratio discounted value of incremental benefits to discounted value of incremental costs. One way to calculate the benefit-cost ration is given below: B/C = B stands for discounted value of benefits, E for the discounted value of expenses, C for the present value of purchase cost and RV for the Residual Value which we assume as zero (after the twelfth year of operation) (Profyllidis, 2008). The B/C ratio for 12 years of service lifespan is equal to 1,14. Internal Rate of Return: the internal rate of return is the value of the discount rate for which the net present value of benefits equals to the net present value of costs. If IRR>I then the investment is considered as viable, else it is not suggested to be initiated. In case IRR~i then a sensitivity analysis is important (Profyllidis, 2008). For 12 year service lifespan, it is IRR=7,34% Sensitivity analysis Due to the dynamic economic environment, before investing onto a project it is wise to perform a sensitivity analysis in order to be on the safe side no matter what shortcomings might appear. As monetary and economic policies change and adjust to social and global economic trends leading in changes in the consumer behavior and demand, several fundamental dimensions should be examined to bear witness of the range of expected economic benefits or losses. Two independent core scenarios that determine the economic outcomes are the discount rate and passenger flows. Eight different scenarios have been investigated: Discount rate 2% Discount rate 4% Discount rate 8% Discount rate 10% Increase of passenger flows to 10% Increase of passenger flows to 20% Decrease of passenger flows to 10% Decrease of passenger flows to 20% The following table presents the sensitivity analysis of purchasing and operating a fleet of hybrid buses with expected service lifespan of 12 years. The analysis investigates the economic outcomes on NPV and B/C ratio in the aforementioned optimistic and pessimistic scenarios. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 201

210 SESSION: TRANSIT QUALITY PERFORMANCE Table 11. Sensitivity analysis of economic indicators for the discount rate and passenger flows for a service lifespan of 12 years Basic scenario Discount rate Passenger flows 2% 4% 8% 10% +10% +20% -10% -20% NPV Β/C 1,14 1,36 1,20 0,97 0,87 1,47 1,79 0,81 0,48 As it was previously calculated the investment is viable for a discount rate less than 7,34%. Higher discount rates lead to negative net present value and then the investment is not economically viable. On the other hand, an increase in passenger flows per 10% or 20% results in boosting NPV and B/C ratio and beneficial investment. Negative trends of demand should drive economic flows to negative zones shrinking the NPV and B/C ratio to prohibitive levels. 5. DISCUSSION Urban areas represent a significant challenge regarding both environmental impacts and traffic congestion. As there are great volumes of people travelling within city contexts in every day basis, special attention should be paid on the environmental burden generated by these needs. To this end, attempts are being made in global level to introduce a more sustainable culture of living and moving. One of these legislative interventions made by European Commission the last two decades, is the vehicleoriented policy of introducing Euro standards, namely restrictions that regard pollutants emitted by vehicles. Furthermore, European Union has issued a set of policy documents (COMs) that communicate its strategic directions, suggesting guidelines to mitigate environmental burden. In this paper, a scenario was introduced that included several modifications compared to the existing route scheduling in the bus lines No 2, 9 and the introduction of a local bus line in the city of Volos. This scenario was strictly shaped to address emerging economic operational deficits and served as the basis to investigate several different scenarios that anticipate the use of cleaner vehicles for urban public transport service. The environmental impact assessment between several types of fleet buses came up with some interesting findings: in terms of GHG emissions, the most environmentally friendly type of bus vehicle was hybrid bus (scenario D). The existing fleet is the second most burdensome close to the first one defined under scenario C. This scenario constitutes CNG and diesel fuel mix enabling sharp reduction of CO and N2O pollutants. However, due to the use of CNG, methane emissions (CH4) are much greater than of any other technology. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 202

211 SESSION: TRANSIT QUALITY PERFORMANCE Table 12. Annual emissions (in kg) produced by different bus fleets Pollutants Existing Environmentally friendly vehicles fleet Scenario Α Scenario Β Scenario C Scenario D CO 2.697, ,12 979,05 758,42 783,19 CO , , , , ,91 CH4 88,52 2,88 2,88 538,16 2,10 NOX 8.739,45 847,11 875, ,12 620,41 NO 7.758,79 762,40 787, ,82 558,37 NO2 980,66 84,71 87,50 111,30 62,04 N2O 9,46 22,79 22,79 0,00 16,64 PM2,5 397,51 19,71 19,21 23,54 14,46 PM10 415,24 37,64 37,14 41,30 27,62 PM exhaust 381,56 3,57 3,07 7,56 2,61 VOC 978,19 25,37 21,80 673,80 18,58 NMVOC 889,67 22,49 18,91 135,64 16,48 GHG , , , , ,04 The greenhouse gas emissions are calculated through the following formula: GHG=CO2+25*CH4 +298*N2O The potential of CH4 for the global warming is 25 times stronger than CO2 whereas the N2O potential is 298 times stronger. That is how this formula was build. It is proved that hybrid buses consume less fuel than the conventional ones. Moreover, they have slightly less maintenance cost leading in lower operational cost. Besides this, the introduction of greener vehicles would be more attractive to road network users resulting in ridership increase and higher revenues. However, in order to specify costs and benefits, a long-term cost-benefit analysis and sensitivity checks are useful. Based on analysis outcomes, using normal discount rates and a moderate value of expected service lifespan, the investment seems viable and relatively safe under light discount rate variations and smooth passenger flow fluctuations. The investigation of totally electric fleet of buses could be the next step in introducing a sustainable approach in urban public transport. Assessing the whole life-cycle impact of the introduction of such technologies could be also a challenge, allowing room for wider impact monitoring. Life-Cycle Analysis could also be implemented onto the existing fleet in order to assess and compare longer term impacts of alternative and conventional engine technologies and scrutinize broader implications in people health. References Bartin, B., Life cycle cost analysis of Hybrid Bus Deployment on Transit Lines. Transportation Research Board s 92nd Annual Meeting, Washington, D.C. Chandler, K. and Walkowicz, N., King County Metro Transit Hybrid Articulated Buses: Final Evaluation Results. National Renewable Energy Laboratory. Chakroborty, P. and Dwivedi, T., Optimal Route Network Design for Transit Systems Using Genetic Algorithms. Engineering Optimization, 34(1), Demirbas, A., Importance of biodiesel as transportation fuel. Energy Policy 35, nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 203

212 SESSION: TRANSIT QUALITY PERFORMANCE EPA U.S. Environmental Protection Agency, Air and Radiation. < European Commission, Green paper: Towards a new culture for urban mobility. European Environmental Agency, The contribution of transport to air quality. TERM 2012: transport indicators tracking progress towards environmental targets in Europe. Frantzeskakis J., Pitsiava-Latinopoulou M. and Tsamboulas D. "Traffic Management". 2nd edition, Papasotiriou, Athens 1997 Giannopoulos G. (1994). Urban public transport: Buses. Paratiritis, Thessaloniki.. Houyu, L. and Guirong, Z., Hybrid Electric Vehicle Drive Control, rd International Conference on Environmental Science and Information Application Technology (ESIAT 2011), Procedia Environmental Sciences 10 (2011), Karlaftis M. and Lymperis K. (2009). Urban transport systems. Symmetria editions, Athens. Lajunen, A., Energy consumption and cost-benefit analysis of hybrid and electric city buses. Transportation Research Part C (38),1-15. Mitropoulos, L., Sustainability framework for urban transportation modes and exploratory applications. Dissertation. Municipality of Volos, Operational Programme of Municipality of Bolos Phase A : Strategic Planning, Volos. Nathanail, Upgrading of transit service quality in the northwest area of city of Volos, Volos. Papafoti Lambrini (2013). Green urban transportation in Volos. Diploma thesis. Profyllidis, B. (2008). Transport Economics. Papasotiriou, Athens. Transportation Research Board, Bus route evaluation standards. TCRP Report 10. Transportation Research Board, Assessment of Hybrid Electric Transit Bus Technology. TCRP Report 132. Transportation Research Board, Guidebook for Evaluating Fuel Choices for Post Transit Bus Procurements. TCRP Report 146. Wall, G., Felstead, T., Richards, A., Mcdonald, M., Cleaner vehicle buses in Winchester. Transport Policy 15, Μορφουλάκη, Μ., Κoτούλα, Ν., Παροχή υπηρεσίας σύνταξης μελέτης για την οργάνωση των Δημοσίων Συγκοινωνιών του Δήμου Κω: Π2. ΠΡΟΤΑΣΗ ΑΝΑΔΙΑΡΘΡΩΣΗΣ ΚΑΙ ΛΕΙΤΟΥΡΓΙΑΣ. ΕΚΕΤΑ/ΙΜΕΤ. Θεσσαλονίκη. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 204

213 SESSION: TRANSIT QUALITY PERFORMANCE A Decision Tree Application in Transit Quality of Service in the City of Volos Maria Tsami 1,2, Eftihia Nathanail 1,2 1 University of Thessaly, Department of Civil Engineering, Transportation Engineering LaboratoryPedion Areos, 38334, Volos, Greece 2 Centre for Research and Technology Hellas (CERTH) / Hellenic Institute of Transport (HIT), 6 th km Charilaou - Thermi Rd., P.O. Box: 361, P.S.:57001, Thessaloniki, Greece *Corresponding author: martsami@civ.uth.gr, Tel.: , Fax: Abstract Transit Quality of Service seems to affect significant travelers choices towards a transit trip. Men and women travelers perceive differently the quality of service, and women travelers are usually linked with highest quality expectations than men. The present paper aims to identify the different correlations among transit quality of service parameters perceived importance between male and female travelers and to examine the role of transit information in combination with other quality parameters, taking into account sample s socioeconomic characteristics. Towards this aim, two decision trees have been developed and analyzed, giving significant input for the improvement of crucial quality parameters according to users perceptions. Results showed that in a gender analysis the most important parameter to classify the sample is the costumer services indicator while when analyzing the information importance the most crucial parameter is the service reliability. Keywords: Transit Quality of Service; WEKA; decision trees; J48 algorithm; Transit information importance 1. INTRODUCTION Quality of Services (QoS) of transit systems has been examined to a great extent by many researchers (TRB,1999; Glerum et.al.,2011; Tsami and Nathanail, 2012; Tsami and Nathanail, 2014a; Tsami and Nathanail, 2014b, Nathanail,2008; Tyrinopoulos and Antoniou, 2008; Eboli and Mazzulla 2007; Eboli and Mazzulla 2008; Eboli and Mazzulla 2011; Dell Olio 2010; de Oña et.al.,2012). Tyrinopoulos and Antoniou (2008) examined five transit systems in Greece based on importance and satisfaction criteria [14]. Tsami and Nathanail (2012), examined the perceived and importance level users recognize in transit quality indicators, the gender differentiation among quality importance parameters and the overall quality assessment formulation based on the evaluation of sub-quality indicators (2014a). Many researchers furthermore, tried to identify the difference between actual and perceived by users transit QoS (Tsami and Nathanail, 2014b; Eboli and Mazzulla, 2008; Eboli and Mazzulla, 2011). 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 205

214 SESSION: TRANSIT QUALITY PERFORMANCE QoS parameters are also considered to affect travel choices and users strategies for transit trips (Spiess and Florian, 1989). Tsami and Nathanail stated that transit quality parameters influence the optimal strategies users develop before transit choices (Tsami and Nathanail, 2014a) and Glerum et.al.(2011) examined the mode choice selections of users based on users perceptions on quality parameters. In the present study, two decision trees have been developed to classify the quality parameters following a categorization of the sample based on gender and transit information. The aim of the present paper therefore is considered to be twofold: to examine the gender differences on QoS assessment to assess the transit information by examining its correlation with other transit quality indicators and users socioeconomic characteristics. 2. DATA COLLECTION AND SAMPLE CHARACTERISTICS 2.1. Design and realization of the survey The data used in this analysis, were collected within the framework of a household survey conducted in the city of Volos, Greece, during a two-month period, January- February In total 876 questionnaires were collected and analyzed. Respondents were asked to evaluate the importance of eleven (11) QoS indicators represented in Table 1. All 11 quality indicators were evaluated by transit users, regarding the importance users recognize on a 5 point likert scale (1 represented the lowest importance rating and 5 the highest importance). Except of the quality indicators, a list of socioeconomic indicators were examined as well, including age (as a numeric variable), gender (1 men, 2 women), education (1Did not finish the primary school, 2 Primary school education, 3 High school, 4 Secondary education, 5 College/University student, 6 Higher education degree holder), occupation (1 Private sector employee, 2 Public sector employee, 3 Free lancher,4 Farmer, 5 Student, 6 Pensioner, 7 Unemployed, 8 Other), number of cars in the household (0 Zero,1 One, 2 Two, 3 More than two) and monthly family allowance (0 Do not answer, 1 <=1172 euro, euro, 3 >=2344 euro). Table 1. Transit Quality indicators and symbols used in the analysis Quality indicator Symbols Route characteristics RC Service characteristics SC Service reliability Re Comfort Cf Cleanliness Cl Cost Cs Information Inf Safety and security Sf Personnel Pl Customer services CS Environmental protection EP 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 206

215 SESSION: TRANSIT QUALITY PERFORMANCE 2.2. Socioeconomic characteristics of the sample As indicated in Table 2, 49% of the sample consisted of male respondents and 51% of female. The majority of them (37%) had a higher education degree, followed by a percentage of 30% with a secondary education level. Most of respondents were pensioners (23%) and many of them (19%) private sector employees. A percentage of 12% were public sector employees and another percentage of 12% were students. Almost half of the sample respondents (47%) had a low monthly family allowance (47%), 40% stated that their monthly family allowance is within euro and only 13% exceed this amount of money in a monthly rate. Table 2. Sample Socioeconomic Characteristics Indicators Attributes N % Gender Men Women Education Did not finish the primary school 32 4 Primary school education High school 34 4 Secondary education College/University student Higher education degree holder Occupation Private sector employee Public sector employee Free lancer Farmer 6 1 Pupil 40 5 Student Household Pensioner Unemployed 59 7 Other 0 0 Monthly Family Allowance <=1172 euro euro >=2344 euro METHODOLOGICAL APPROACH 3.1. Decision Trees Decision Trees are among the popular approaches to represent classifiers with extended applicability in many disciplines (statistics, data mining, machine learning etc.) (Quinlan, 1993; Breiman et.al, 1984). A decision tree represents a graph that can support decision making process by modeling decisions and possible impacts including resource costs and utility. Decision trees are considered to be a useful tool to identify the most appropriate strategy to meet a goal. The tree has a root node (main classifier), interior nodes (representing the input variables) leaves/end nodes (representing the class labels) and branches (representing conjunctions of features that lead to those class labels). Each leaf (end node) represents a value of the target variable given the values of the input variables represented by the path from the root node to the leaf. For example 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 207

216 SESSION: TRANSIT QUALITY PERFORMANCE in the transit operation case, if the target is to make female travelers to use more the transit system, a decision tree representing the importance role women recognize in transit quality parameters (input variables) could provide input to make the appropriate decision on selecting the parameters that need to be improved in order to attract female users in the service. Two decision trees have been developed, as discussed in the present section QoS importance by gender The first decision tree developed in terms of the present study, aimed at identifying the role of gender in the formulation of the correlations among the different quality indicators. Figure 1. J48 decision tree for a gender categorization of Quality of Service indicators importance. Figure 1, represents the graph of the J48 tree impelemented in Waikato Environment for Knowledge Analysis (WEKA). This tree represents the quality paths for a gender categorization of the importance corellations of the tested quality indicators. In Table 3, all tree paths are presented, each one linking to a gender categorization of the sample. 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 208

217 SESSION: TRANSIT QUALITY PERFORMANCE Table 3.Tree paths linking to Gender NO Tree path Gender 1 Costumer services<=3, Costumer Services <=2, Service Characteristics <=1, Costumer Men Services<=1, Comfort <=3 2 Costumer services<=3, Costumer Services <=2, Service Characteristics <=1, Costumer Women Services<=1, Comfort > 3 3 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1 Women 4 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Men <=1, Cost<=3 5 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Women <=1, Cost>3 6 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Men >1, Cost<=3, Cleanliness<=3, Information<=3,Cost<=2<Reliability<=2,Comfort<=3,Route Characteristics<=2 7 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Women >1, Cost<=3, Cleanliness<=3, Information<=3,Cost<=2<Reliability<=2,Comfort<=3,Route Characteristics> 2 8 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Men >1, Cost<=3, Cleanliness<=3, Information<=3,Cost<=2<Reliability<=2, Comfort > 3 9 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Men >1, Cost<=3, Cleanliness<=3, Information<=3,Cost<=2<, Reliability>2 10 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Women >1, Cost<=3, Cleanliness<=3, Information<=3,Cost>2 11 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Women >1, Cost<=3, Cleanliness<=3, Information>3 12 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Men >1, Cost<=3, Cleanliness>3, Route Characteristics<=1 13 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Women >1, Cost<=3, Cleanliness>3, Route Characteristics>1 14 Costumer services<=3, Costumer Services <=2, Service Characteristics > 1, Comfort Men >1, Cost>3 15 Costumer Services>3 Women 16 Costumer Services<=3,Costumer Services>3, Cost<=2,Information<=3, Service Women Characteristics<=2, Reliability<=1 17 Costumer Services<=3,Costumer Services>3, Cost<=2,Information<=3, Service Women Characteristics<=2, Reliability>1, Comfort<=1 18 Costumer Services<=3,Costumer Services>3, Cost<=2,Information<=3, Service Men Characteristics<=2, Reliability>1, Comfort>1 19 Costumer Services<=3,Costumer Services>3, Cost<=2,Information<=3, Service Women Characteristics>2, Service Characteristics<=3 20 Costumer Services<=3,Costumer Services>3, Cost<=2,Information<=3, Service Men Characteristics>2, Service Characteristics>3, Service Characteristics <=4 21 Costumer Services<=3,Costumer Services>3, Cost<=2,Information<=3, Service Women Characteristics>2, Service Characteristics>3, Service Characteristics >4 22 Costumer Services<=3,Costumer Services>2, Cost<=2,Information>3, Safety<=3 Men 23 Costumer Services<=3,Costumer Services>2, Cost<=2,Information>3, Safety>3 Women 24 Costumer Services<=3,Costumer Services>2, Cost>2,Environmental Protection<=1 Women 25 Costumer Services<=3,Costumer Services>2, Cost>2,Environmental Protection>1 Men 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 209

218 SESSION: TRANSIT QUALITY PERFORMANCE What we can notice from the tree paths is that there are 11 out of 25 different paths of quality indicators correlations that could conclude to men travellers. This means that for a decision maker who requires to attract men travellers based on a quality importance assessment feedback there are 11 alternatives. Still, all paths do not have similar weights. Thus, one way to select the crusial decision paths is to select the paths with the more correctly classifyed instances (indicated by the first number in the parenthesis in the end nodes of the tree). In this approach for example, path NO:25 is the most crucial path for men and path NO:15 the most crusial for women travellers. The pruned tree developed (pruning value: 0.05) had 25 leaves and the time taken to build model was 0.05 seconds. The time taken to test model on training data was 0.01 seconds. The total number of instances was 876 and the test option used was to use the training set. Tables 4 and 5 represent the output and detailed accuracy of the decision trees by class respectively. Table 4. WEKA J48 output Correctly Classified Instances 571 ( %) Incorrectly Classified Instances 305 ( %) Kappa statistic Mean absolute error Root mean squared error Relative absolute error % Root relative squared error % Coverage of cases (0.95 level) 100 % Mean rel. region size (0.95 level) % Table 5. WEKA J48 detailed accuracy by class Class TP 1 Rate FP 2 Rate Precision 3 Recall 4 F- Measure 5 MCC 6 ROC 7 Area PRC 8 Area Weighted Avg True Positive Rate (correctly identified) 2 False Positive Rate (incorrectly identified) 3 Precision is the probability that a (randomly selected) retrieved instance is relevant 4 Recall is the probability that a (randomly selected) relevant instance is retrieved in a search 5 F-Measure is the harmonic mean of precision and recall (F-Measure = 2* (Precision*Recall)/(Precision+Recall)) 6 Matthews Correlation Coefficient 7 The area under the ROC curve as the Wilcoxon-Mann-Whitney statistic. 8 The area under the precision-recall curve 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 210

219 SESSION: TRANSIT QUALITY PERFORMANCE Importance of information The second decision tree is represented graphically in Figure 2 and in greater detail by representing all the paths that are linked with a certain information importance scores in Table 6. This decision tree helps us to identify the importance of information relating information with other quality of service parameters and the socioeconomic characteristics of the sample. In such an analysis the route node that defines the most important parameter is the reliability indicator. Figure 2. J48 Tree for the information importance evaluation 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 211

220 SESSION: TRANSIT QUALITY PERFORMANCE Table 6. Tree paths linking to Information NO Tree path Information Importance Rating 1 Reliability <=4, Personnel <=4, Safety and Security<=1, Cleanliness<=2, 1 Reliability<=3,Cost<=2 2 Reliability <=4, Personnel <=4, Safety and Security <=1, Cleanliness<=2, 2 Reliability<=3,Cost>2 3 Reliability <=4, Personnel <=4, Safety and Security <=1, Cleanliness<=2, Reliability>3 3 4 Reliability <=4, Personnel <=4, Safety and Security<=1, Cleanliness>2, Environmental 2 Protection<=2 5 Reliability <=4, Personnel <=4, Safety and Security <=1, Cleanliness>2, Environmental 3 Protection>2 6 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 2 Protection<=3,Personeel<=2,Customer Services<=2 7 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 3 Protection<=3,Personnel<=2,Customer Services>2 8 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 2 Protection<=3,Personnel>2,Customer Services<=1 9 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 3 Protection<=3,Personnel>2,Customer Services>1 10 Reliability <=4, Personnel <=4, Safety and Security>1, Environmental 3 Protection>3,Cost<=2,Cleanliness<=2 11 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 2 Protection>3,Cost<=2,Cleanliness>2, Cleanliness<=3 12 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 1 Protection>3,Cost<=2,Cleanliness>2, Cleanliness>3, Reliability<=2 13 Reliability <=4, Personnel <=4, Safety and Security>1, Environmental 3 Protection>3,Cost<=2,Cleanliness>2, Cleanliness>3, Reliability>2 14 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 3 Protection>3,Cost>2,Environmental Protection<=4,Monthly Family Allowance=0 15 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 3 Protection>3,Cost>2,Environmental Protection<=4,Monthly Family Allowance=1, Route Characteristics<=3 16 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 3 Protection>3,Cost>2,Environmental Protection<=4,Monthly Family Allowance=1, Route Characteristics>3,Cleanliness<=3 17 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 4 Protection>3,Cost>2,Environmental Protection<=4,Monthly Family Allowance=1, Route Characteristics>3,Cleanliness>3 18 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 2 Protection>3,Cost>2,Environmental Protection<=4,Monthly Family Allowance=2, Reliability<=2 19 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 3 Protection>3,Cost>2,Environmental Protection<=4,Monthly Family Allowance=2, Reliability>2,Gender=1 20 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 4 Protection>3,Cost>2,Environmental Protection<=4,Monthly Family Allowance=2, Reliability>2,Gender=2 21 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental 2 Protection>3,Cost>2,Environmental Protection<=4,Monthly Family Allowance=3 22 Reliability <=4, Personnel <=4, Safety and Security>1, Environmental Protection>3,Cost>2,Envirnonmental Protection>4,Personnel<=3 4 2 nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 212

221 SESSION: TRANSIT QUALITY PERFORMANCE 23 Reliability <=4, Personnel <=4, Safety and Security >1, Environmental Protection 3 >3,Cost>2,Environmental Protection>4,Personnel>3 24 Reliability <=4, Personnel>4, Gender=1,Customer Services<= Reliability <=4, Personnel >4, Gender=1,Customer Services>4, Route 4 Characteristics<=3,Environmental Protection<=3 26 Reliability <=4, Personnel >4, Gender=1,Customer Services>4, Route 2 Characteristics<=3,Environmental Protection>3 27 Reliability <=4, Personnel >4, Gender=1,Customer Services>4, Route Characteristics> Reliability <=4, Personnel >4, Gender=2,Environmenta Protection<= Reliability <=4, Personnel >4, Gender=2,Environmental Protection> Reliability >4, Service Characteristics<=2,Comfort<= Reliability >4, Service Characteristics <=2,Comfort> Reliability >4, Service Characteristics>2 4 As we can notice, there are three alternative paths to achieve an importance rating for information equal to 1 (paths:1,12,31), ten paths for a rating equal to 2 (paths:2,4,6,8,11,18,21,24,26,30 ), twelve paths for a rating equal to 3 (paths:3,5,7,9,10,13,14,15,16,19,23,29 ), (paths:17,20,22,25,28,32) and only one path for the highest importance rating score, equal to 5 (path:27). The pruned tree developed (pruning value: 0.05) (Figure 2) has 32 leaves and the time taken to build model was 0.14 seconds. The time taken to test model on training data was 0.01 seconds. The total number of instances was 876 and the test option used was to use the training set. In Tables 7 and 8 are represented the model output and the detailed accuracy by class respectively. Table 7. WEKA J48 output Correctly Classified Instances 492 ( %) Incorrectly Classified Instances 384 ( %) Kappa statistic Mean absolute error Root mean squared error Relative absolute error % Root relative squared error % Coverage of cases (0.95 level) % Mean rel. region size (0.95 level) % Table 8. WEKA J48 detailed accuracy by class Class TP Rate FP Rate Precision Recall F- Measure MCC ROC Area PRC Area Weighted Avg nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 213

222 SESSION: TRANSIT QUALITY PERFORMANCE 4. CONCLUSIONS The scope of this paper was to identify the gender differences in the quality importance recognition and the role of transit information compared to travelers socioeconomic characteristics and importance evaluation of other transit quality indicators. In total 876 respondents from the city of Volos, Greece responded to the survey conducted towards the aim of the paper. Research shows that costumer services is the most important indicator for a gender classification, while reliability is the main classifier in the information importance classification. Two decision trees have been developed to underline the decision paths one should follow to address the study objectives. There are alternative paths linking to the same scores and results. The selection of the decision path to follow usually is made by the path with the highest number of correctly classified instances. For the gender analysis women could be represented by the sample that gives an importance score of >3 in the indicator Costumer Services, while men could be represented by the class of the sample that evaluates Costumer Services importance with a medium score (=3), Cost >2 and Environmental Protection >1. Similarly, for the highest importance score of information evaluation, decision makers should follow the path where travelers evaluate the importance of quality indicators as follows: Reliability <=4, Personnel >4, Gender=1, Customer Services>4, Route Characteristics>3. Next step of this research is to examine the performance evaluation of the travelers and develop the relevant decision trees for the performance evaluation. By analyzing the performance-importance scores in a comparative way, useful input will be provided for the decision makers who wish to attract transit travelers and achieve higher evaluation scores by them. References Breiman, L., Friedman, J H., Olshen, R.A., & Stone, C.J. (1984). Classification and regression trees. Monterey, CA: Wadsworth. Dell Olio, L., Ibeas, A. and Cecín, P. (2010) Modelling user perception of bus transit quality. Transport Policy, 17(6), Eboli, L. and Mazzulla, G., Service quality attributes affecting customer satisfaction for bus transit. Journal of Public Transportation 10 (3), Eboli, L. and Mazzulla, G. (2008) Willingness to pay of public transport users for improvement in service quality. European Transport, 38, Eboli, L. and Mazzulla, G. (2011) A methodology for evaluating transit service quality based on subjective and objective measures from the passenger s point of view. Transport Policy, 18, Glerum A., Atasoy B., Monticone A., Bierlaire M. (2011) Adjectives qualifying individuals' perceptions impacting on transport mode preferences. International Choice Modelling Conference Juan de Oña, Rocío de Oña Francisco Calvo (2012). A classification tree approach to identify key factors of transit service quality, Expert Systems with Applications, Volume 39, Issue 12, 15 September 2012, Pages Nathanail E., Measuring the quality of service for passengers on the Hellenic railways, Transp. Res. Part A Policy Pract., vol. 42, no. 1, pp , Jan nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 214

223 SESSION: TRANSIT QUALITY PERFORMANCE Quinlan (1993). C4.5: Programs for machine learning. San Francisco: Morgan Kaufmann. Spiess, H. and Florian, M. (1989) Optimal strategies: a new assignment model for transit networks. Transportation Research, 23B(2), pp Transportation Research Board, A Handbook for Measuring Customer Satisfaction and Service Quality. TCRP Report 47. National Academy Press, Washington, DC. Tsami Maria and Nathanail Eftihia, Assessing the quality of service in public transport. 5th International Conference on Traffic and Transport Psychology, August 29-31, 2012, Groningen, Netherlands. Tsami Maria, Eftihia Nathanail, 2014a. Examining travelers Optimal strategies in transit trip choice, applying a classification tree approach on transit quality of Service Indicators. OPT-I, International Conference on Engineering and Applied Sciences Optimization, 4-6 June 2014, Kos Island, Greece. Tsami Maria and Nathanail Eftihia, 2014b. Opening ground to female transit movements. Women's vs operator's perspective in transit quality of service. 5th International Conference on Women's Issues in Transportation - Bridging the Gap, Paris - La Défense, France, 14-16, Apr Tyrinopoulos, Y. and Antoniou, C. (2008) Public transit user satisfaction: Variability and policy implications, Transport Policy 15 (4), pp nd Conference on Sustainable Urban Mobility, 5-6 May 2014, Volos, Greece 215

224