Smart Energy Utilities based on Real-Time GIS Web Services and Internet of Things

Transcription

1 Available online at ScienceDirect Procedia Computer Science 110 (2017) 8 15 The 14th International Conference on Mobile Systems and Pervasive Computing (MobiSPC 2017) Smart Energy Utilities based on Real-Time GIS Web Services and Internet of Things Reza Nourjou a,, Mehdi Hashemipour b a GIS Researcher, Washington D.C., USA b Systems Engineer, Washington D.C., USA Abstract Problem: The power & electric outages cause business disruptions that result in economic impacts to cities. A web map service (WMS), as a RESTful API with URL, that makes real-time information of power outage and economic loss estimation accessible and available across the web would be valuable and important for decision makers in smart cities and smart energy utility. Objective: This paper presents a web-based system that is capable of creating real-time WMS. Our system enables geo-enabled applications (such as web mapping, mobile GIS app, operations dashboard) to access, query, and display a web map of real-time outage status and economic loss information associated with outages in the county scale. Method: Our method is based on the Internet of Things (IoT), web services, GIS web services, web mapping, and web GIS (Geographic Information Systems). To implement our system, we used esri ArcGIS for Server, SQL Server Spatial, Amazon Web Services, and C#.NET. The system architecture consists of autonomous computer program, spatial database, GIS Server, cloud computing platform, a WMS, web feature service (WFS). Our insight is in real-time by continuously 1) connecting to web sensors and collecting outage data, 2) connecting to web services for calculating economics loss associated with outages, 4) manipulating (modify, insert, delete) our spatial database over the Internet, 4) publishing a real-time WMS. Result: The system implementation resulted in a back-end server that includes /serves a real-time WMS. A data record includes: name of county, name of state that contains this county, total number of costumers located within the county, total number of costumers without electricity who are located into the county, and total economic losses (modeled as total jobs loss, total GDP loss, and total cost in $) caused by outages for the county. Findings: Real-time WMS created by our system can be used by applications to display and query a web map of electric outage status and their macroeconomic loss in real-time. They provide a decision support tool for disaster emergency management in the energy sector by aiding in prioritization of restoration operations and recovery phase in electric utilities in order to reduce the economic impact of outages. Our system provides an important platform for smart cities, real-time GIS, Internet on Things, map-based crowdsourcing applications, real-time intelligent systems, and for real-time monitoring and analytical applications. c 2016 The Authors. Published by Elsevier B.V The Authors. Published by Elsevier B.V. Peer-review under responsibility of the Conference Program Chairs. Keywords: cloud; Internet of Things; sensor network; power outage; economic loss; electric utilities; spatial web service; smart cities; real-time GIS; web GIS Corresponding author. address: nourjour@gmail.com The Authors. Published by Elsevier B.V. Peer-review under responsibility of the Conference Program Chairs /j.procs

2 Reza Nourjou et al. / Procedia Computer Science 110 (2017) Introduction Problem: Electric & power outage in power & electric utilities is a short-term or long-term loss of electric power in an area. In a macro-geographic scale, an outage record is a set of information: name of county for which the outage is reported, name of a state that contains this county, number of costumers who are served by and located within this county, number of customers (business and commercial units) without electricity, and time of last update in which these data are observed. These data are open and free and are accessible in power & electric companies web sites. A web sensor reports a number of outage records in real-time. Indeed, a web sensor makes outage data publicly available through web services API (URL) with web data formats such as JSON across the web. For example, we are able to access and collect outage data sensed by the Dominion power company through an API (please see footnote 1 ). These outage data are updated every 15 minutes to show the real-time status of power outages (please see footnote 2 ). It is assumed to dedicate a web sensor to a power company. Therefore, a sensor network indicates a set of web sensors. Outages cause economic loss (E.L. in short) due to business disruptions. Outages can impact electricity consumers primarily through property loss and business disruption. The Electric Power Research Institute estimated that the annual cost of outages across all U.S. business sectors was $104 billion to $164 billion in With the estimating the average number of power outages in a year over five minutes, a reasonable estimate for the annual economic cost yields a possible $20 billion to $31 billion as a lower range 1. The 2003 blackout in the northern and eastern U.S. and Canada which caused a $6 billion loss in economic revenue is one of many indicators that the current electrical grid is outdated 2. Economic models are used to assess E.L. associated with outages in counties. Losses are usually presented by the total GDP loss in USD$, total jobs loss, and total cost in USD$ in each county. Our focus is on the county level since the web sensors reports outages in counties. These models should be executed in real-time as soon as a county experiences an outage situation relative to the anticipated duration of the outage. Outages become a crisis when an unpredictable event such as a hurricane or flood causes a widespread outages. Consequently, outages generate vast negative economic outcomes over an extended period of time. Specially, when an affected county contains many business and commercial units. Motivation: Imagine a catastrophic hurricane Matthew hit the Florida State on October 2016 and caused widespread outages in 67 counties. An efficient action plan for restoration operations results in reducing the economic impacts of outages. Thus, the disaster emergency management department are faced with some main questions: 1) what is current (real-time) status of outages and also spatial distribution of outages in 67 counties? 2) what is current (real-time) status of E.L. associated with outages in each county? 3) which county should be considered as the first priority in restoration operations? 4) how much loss can be reduced if a certain county is restored? Importance of Problem: It is important to collect outage data from the sensor network in real-time, analyze the gathered data and estimate E.L., and share information (collected outages and calculated E.L.) in real-time through real-time and dynamic maps. This type of real-time map creates situational awareness. This process is valuable for decision makers, emergency managers, industry partners, and operational decision support, as well as pre- and post-facto planning to have an accurate real-time perception (understanding) of outage crisis. Thus it enables them in a real-time mechanism to monitor outage status and their economic impacts. Moreover, it assists them a real-time mechanism to prioritize the restoration/recovery operations of critical hubs and services in order to reduce economic loss due to outages. Research Question: Considering the fact that there is a sensor network, how can outage data be collected in realtime, E.L. be estimated in real-time, and this information as a real-time and dynamic map be shared by users through real-time web services API? Objective: We aim to build a web-based system for smart cities. Our system is capable of 1) connecting to web sensors and collecting outage data from them, 2) connecting to web services that provide economic loss estimation (E.L.E. in short), 3) manipulating our spatial database across the web, 4) providing real-time WMS for applications 1 generation_data/2017_01_24_20_30_30/report_region.json?_= report-panel-county-muni

3 10 Reza Nourjou et al. / Procedia Computer Science 110 (2017) 8 15 (web mapping, mobile, desktop) to access, query, and display real-time maps of outages and E.L. associated to outages. Innovation and Contribution: Our innovation is a computer system that connects outage sensor network and E.L.E. services and shares real-time outages data and real-time E.L.E through real-time GIS gis web services (URL). Significance and Importance: We highlight the significance and importance of our work as follows: It provides a valuable platform to publish and serve a real-time WMS that contains real-time outage data and E.L.E. calculation. It provides a valuable platform to build a WFS that allows geo-enabled applications to manipulate an enterprise spatial database on the web through URL (web services). It is valuable for smart cities and smart energy utilities during crisis response operations to power outages. It proposes a flexible and scalable solution to be used in various problem domains, sensors, real-time analysis and computation, real-time monitoring, real-time GIS, location-based services, crowdsourced maps, and webbased services. It uses the cloud technology to embed the spatial database,gis web services, and computer programs. It enables real-time analysis and provides a scalable framework to execute different data analysis. It provides an enterprise spatial database with real-time information. It provides an editable spatial database to be manipulated by applications across the web. 2. Background to GIS Web Services Web services provide a standard means of communication (interoperating) between different software applications, running on a variety of platforms and/or frameworks, by using open standards such as HTML, XML, WSDL, and SOAP. So a web service is a mechanism that provides some kind of data and functionality over the web. A service can be consumed by a web application to access its data or functionality. (Source: World Wide Web Consortium) GIS web services share (provides access to) GIS resources over the internet through standard internet protocols. A GIS WS is NOT a web mapping application. This paper focuses on two types of GIS WS: Web Feature Service (WFS) and Web Map Service (WMS). A WMS is a standard protocol for serving (over the Internet) georeferenced map images which a map server generates using data from a GIS database. The WMS interface returns an image which cannot be edited or spatially analyzed. On the other hand, the WFS standard defines interfaces and operations for data access and manipulation on a set of geographic features across the Internet in the form of URLs. Data manipulation operations include the ability to 1) get or query features based on spatial and non-spatial constraints, 2) create a new feature instance, 3) get a description of the properties of features, 4) delete a feature instance, 5) update a feature instance, 6) lock a feature instance. (Source: Open Geospatial Consortium). For an example, GIS gis web services are implemented in esri ArcGIS for Server (Source: 3. Review Related Works A number of good works have contributed to the field of smart cities and smart energy utilities by using Internet of Things (IoT), cloud, web mapping, Internet-based services, and sensor network. To highlight our contribution and innovation in this paper, Table 1 shows a list of some related works and their features. The innovation of our work is to estimate outage E.L.E. and share those data in real-time. In this work 8, a macroeconomic loss estimation model is implemented as a component in system architecture, and this model is executed in real-time whenever new outage data are collected by the system. While in our approach, economic loss calculations are done by relevant web services over the Internet. The smart grid can be considered a modern electric power grid infrastructure for enhanced efficiency and reliability through automated control, high power converters, modern communications infrastructure, sensing and metering

4 Reza Nourjou et al. / Procedia Computer Science 110 (2017) Table 1. Comparison between our solution and related works.(description: Y: Yes N: No WS4CM: Web Service of Computing Models ) Work Cloud Real-Time Outage Real-Time E.L.E. Real-Time WMS Sensor Network WS4CM Our system Y Y Y Y Y Y 8 Y Y Y Y Y N 9,10,11 Y Y N Y Y N 3,2 Y Y N Y Y N technologies, and modern energy management techniques based on the optimization of demand, energy and network availability, and so on 3,5. Smart grid makes restoration faster, easier for utilities 2. The disadvantage in these solutions is the outages E.L. that is not estimated. They focused on monitoring outages in real-time while the E.L.E. is important for action planning in restoration operations. The US energy & power companies usually publish current map of outages in their service areas, cities and counties. Moreover, those data are accessible though RESTful API. Please see these samples 9,10,11. The main drawback of these maps is the E.L.E.. They do not estimate the E.L.E. due to outages, therefore those types of information are not available. 4. Objective Our objective is to create and publish a WMS (as URL) that can be used for applications to openly access real-time outage data and real-time E.L.E. information and to visualize them on dynamic, real-time maps. To achieve this goal, we developed a real-time GIS, which is considered a back-end program, on web and cloud. The system architecture contains WMS, WFS, GIS server, a spatial database which is editable on the web, and an autonomous computer program. Our insight is through an automated and continuous mechanism to 1) connect to web sensors and collect/gather outage data, 2) analyze data, 3) connect to web services that provide E.L.E. models and E.L.E. calculation, 4) manipulate the spatial database on the web through the WFS, 5) make information accessible through the WMS. Please see Figure 1. Fig. 1. The system proposed by this paper for smart energy utilities.

5 12 Reza Nourjou et al. / Procedia Computer Science 110 (2017) Method We used esri ArcGIS for Server, Amazon Web Services (AWS), Microsoft SQL Server Spatial, and C#.NET to develop our system. More details about our initial achievements are described in paper 8. The system components shown in Figure 1 are explained in the following subsections Web Sensor A web sensor is a web service or a web page that reports real-time outage statuses across the Internet. A web sensor provides RESTful APIs that enable us (the Autonomous Computer Program component) to collect and receive outage data with the JSON format. Those data are updated every 15 minutes to show the real-time and live status of a power outage. Indeed, each power & electric company shares a list of the service areas, such as counties, where customers are experiencing power outages. For example, the Dominion power company s website, shown in footnote 2, and a publicly provided web API, shown in footnote 1, presents 118 outage records, each associated with a county located in the Virginia State. According to the Wikipedia website, the United States has 48 electric companies. Therefore, a network of 48 web sensors can monitor real-time outage statuses in the US Web Services for Computing Models E.L.E. models enable us (the Autonomous Computer Program component) to calculate and assess regional economic loss impact due to outages. Whenever outage data are collected, executing a loss model at the same time results in real-time E.L.E. There are two main approaches to estimating E.L.. The first approach is to select an appropriate loss model and implement it in our system by programming. Our previous work applied this strategy 8. The second one is to recognize and use available web servers that provide web services for E.L.E.. As Figure 1 shows, an API enables the autonomous computer program to send outage data to a web service and receive economic loss data that are calculated by that web server Autonomous Computer Program The autonomous computer program, which is considerable an intelligent system, is the essential component in our system that provides four main capabilities as follow: connecting to web sensors to receive and collect electric outage data from them though APIs, connecting to and communicating with computing web services for to get E.L.E. from them though APIs, manipulating (modifying, inserting, updating, accessing, querying) the spatial database component though the WFS component, be an autonomy to execute some reactive and proactive actions, data analysis, and problem-solving algorithms. This autonomous computer program executes these functions in real-time, repeatedly, continuously, and automatically in order to provide a real-time situational awareness. Actually, this component executes these functions in real-time whenever new outage data are observed. As Figure 1 shows, this component is embedded in a cloud computing platform. To support its functionalities, an internal database stores some essential information such as addresses of web services, web sensors, etc. To implement this component, we deployed C#.NET, LINQ and JSON in.net, Network Programming in the.net Framework, jquery, GeoJSON, ArcGIS REST API, ArcGIS Runtime SDK for.net, ASP.NET Web API. This component is run on AWS.

6 Reza Nourjou et al. / Procedia Computer Science 110 (2017) Spatial Database Data is the core of any information system. Roles that the Spatial Database component plays in our system include: 1) store and manage data (geographic information and non spatial data) that are required by system components for data-related operations and mapping, 2) provide and contain data that are required/used by the WMS and WFS components, 3 ) provide a centralized database and share data, as a geodata service, with system components, and 4) make data available and accessible for applications and for the autonomous software via spatial web services. This spatial database has five important characteristics: be real-time by containing real-time data, be a spatial database by containing geographic and geospatial information, be editable (manipulatable) and replicated across the Internet, be a central database, be embedded (installed, stored, and executed) in the cloud. To implement the spatial database, we first designed a data model as a UML class diagram. Then, we created a spatial database using Microsoft SQL Server, esri Enterprise Geodatabase, and ArcGIS for Server on AWS 8. Finally, counties data as a GIS map were inserted into the spatial database using esri ArcGIS Catalog GIS Server The GIS Server allows us to: 1) create and share GIS resources such as WMS and WFS, 2) manage and server the Spatial Database component on the cloud, 3) respond to and process web requests sent by the autonomous computer program and devices (geo-enabled applications), 4) server and manage WMS and WFS components, 5) and communicate with applications and autonomous program through WFS and WMS. We used esri ArcGIS for Server on Amazon Web Services in order to create and run a GIS Server on the cloud. We used the ArcGIS Server Cloud Builder on Amazon Web Services tool Web Map Service (WMS) This WMS component shares a real-time map that contains two types of information: 1) real-time power outage statuses and 2) real-time E.L.S.. This software component is a web service (a RESTful API with URL) that can be used by any geo-enabled application to access and get this map across the Internet. Actually, it makes this map available and accessible on the Internet through a URL. The WMS is managed by the GIS Server, and it enables applications to send requests to the GIS Server. The most important feature of this component is to be real-time. To server a real-time map service, this WMS is connected via the GIS server to the spatial database, which contains real-time data. This WMS needs to be frequently updated to present real-time and updated data. We used ArcGIS for Server and esri ArcMap to create and publish this type of spatial web service in our work Web Feature Service (WFS) This WFS enables the autonomous computer program to access the Spatial Database component and manipulate data over the web via URL. This web service behaves like a WMS. We used ArcGIS Server and esri ArcMap to create this component in our work Cloud Cloud computing is a model for enabling ubiquitous, on-demand access to a shared pool of configurable computing resources e.g. computer networks, servers, storage, applications and services. We used the cloud for four purposes:

7 14 Reza Nourjou et al. / Procedia Computer Science 110 (2017) 8 15 Fig. 2. Consume our real-time WMS by a web mapping application. 1) embedding (installing, storing, and executing) the autonomous computer program, 2) embedding the Spatial Database component, 3) embedding the GIS Server component, 4) and embedding the WMS and WFS components and making them available and accessible across the globe. We used Amazon Web Services (AWS) for the cloud platform and utilized tools to install these components Geo-enabled Application Our WMS with format URL (web API) can be used easily by any geo-enabled application such as web mapping, a mobile GIS app, and an operational dashboard. Any device running a such app can get /receive real-time web map of outages and E.L.E. from our system and display it on a map-based user-interface. 6. Implementation and Results To validate and evaluate our system presented in Figure 1, we first implemented and developed its components. The system implementation results in a back-end serve, which includes WMS. This WMS has a defined URL to share real-time outages and real-time E.L.E.. Then, we built web mapping applications using the ArcGIS Online, Web AppBuilder tool, and Operations Dashboard for ArcGIS. These geo-enabled applications used the WMS component. Please see Figure 2. E.L.E. information is modeled in cost (in USD$) of extended business disruptions (i.e. total loss GDP, total loss output, total loss jobs) due to an extended power outage. As the result of the electric outage change in real-time, the economic loss estimation varies over time Findings Our system provides a flexible platform to implement real-time GIS. Some applications are as follows: 1) Internet of Things, 2) crowdsourcing, 3) smart meters and smart grid in utilities (gas, water, electricity) management, 4) coordinate disaster emergency operations, run the agent-based simulation, and plan actions of first respond teams 4,6,7 in real-time, 5) monitor and track dynamic assets (vehicles, cargo, trains, cars, aircraft, vessels or stationary sensors) and field crews on a map, 6) develop real-time dashboards for supply chain managers.

8 Reza Nourjou et al. / Procedia Computer Science 110 (2017) Conclusion Our system creates and shares real-time WMS that can be used by any geo-enabled application to display a web map of electric outage statuses and their macroeconomic loss in real-time. These spatial web services provide a decision support tool for disaster emergency management and crisis response in the energy sector in real-time through aiding in prioritization of operations restoration in electric utilities in order to reduce the economic impact of outages. It can enhance efficiency of recovery operations in energy utilities, improve real-time decision making and real-time planning, enable real-time monitoring, support operational decisions, solve critical problems, and create new and innovative experiences in smart electric utilities and smart cities. As a result, our communities will be more resilient, as evidenced by lives saved, property protected, and businesses staying open in the wake of a catastrophic event. One drawback to be considered in our approach is our use of eari ArcGIS. Our system uses ArcGIS for Server which is a costly software. Moreover, we need to develop a scalable and flexible GIS server. Another one drawback is to use simple E.L.E. models while our autonomous computer program can provide a scalable framework to execute different data analysis for various requirements. The future works would use open source tools such as GeoServer for sharing geospatial, PostGIS for spatial database, and MapServer for web mapping. References 1. Campbell, R.J., 2012, August. Weather-related power outages and electric system resiliency. Washington, DC: Congressional Research Service, Library of Congress. 2. Cavoukian, A., Polonetsky, J., & Wolf, C. (2010). Smartprivacy for the smart grid: embedding privacy into the design of electricity conservation. Identity in the Information Society, 3(2), Gungor, V. C., Sahin, D., Kocak, T., Ergut, S., Buccella, C., Cecati, C., and Hancke, G. P. (2011). Smart grid technologies: Communication technologies and standards. IEEE transactions on Industrial informatics, 7(4), Hashemipour, M., Stuban, S. M., & Dever, J. R. (2017). A community-based disaster coordination framework for effective disaster preparedness and response. Australian Journal of Emergency Management, The, 32(2), Hourdouillie, R., and D. Pollock (2016). How multi-services gateways enable IoT for utilities. Journal of Energy and Power Engineering 10 (2016) Nourjou, R. and J. Gelernter (2015). Distributed Autonomous GIS to Form Teams for Public Safety. In Proceedings of the 4th ACM SIGSPA- TIAL International Workshop on Mobile Geographic Information Systems, pp , Seattle, Washington, November 2015, ACM. 7. Nourjou, R., H. Tatano, H. Aghamohammadi (2015). Search Algorithm for Optimal Execution of Incident Commander Guidance in Macro Action Planning. International Journal of Intelligent Systems Technologies and Applications, 14(3/4), Nourjou, R. and J. Thomas (2016). System Architecture of Cloud-based Web GIS for Real-Time Macroeconomic Loss Estimation. In Proceedings of the 5th ACM SIGSPATIAL International Workshop on Mobile Geographic Information Systems, San Francisco Bay Area, California, November 2016, ACM. 9. Map of outages reported by the Dominion company. Online. Available: Accessed 10 February Map of outage information reported by the ComEd company. Online. Available: Accessed 10 February Map of current outages reported by the Southern California Edison company. Online. Available: OutageMap_new.html. Accessed 10 February 2017.