A GIS Methodology to Map Routes from a Truck Permit Database Utilizing a Linear Reference System and Network Analysis

Transcription

1 Dayan et al A GIS Methodology to Map Routes from a Truck Permit Database Utilizing a Linear Reference System and Network Analysis Sinaya Dayan M.S. Geospatial Analyst Rahall Transportation Institute 00 Third Ave., Huntington, West Virginia 0 Telephone: sdayan@njrati.org Andrew P. Nichols, Ph.D., P.E. Associate Professor Weisberg Division of Engineering Marshall University One John Marshall Drive, Huntington, West Virginia Telephone: andrew.nichols@marshall.edu Chih-Sheng Chou, Ph.D. Research Associate Rahall Transportation Institute 00 Third Ave., Huntington, West Virginia 0 Telephone: jchou@njrati.org Sanghong Yoo, M.S., M.S.E GIS Data Coordinator Office of Enterprise Web Portal and GIS (EWPG) Department of Information Technology (IT) Washington Metropolitan Area Transit Authority (WMATA) 00 Fifth St NW, Washington, DC 000 Telephone: syoo@wmata.com Tuan Nguyen, M.S. Geospatial Analyst Rahall Transportation Institute 00 Third Ave., Huntington, West Virginia 0 Telephone: tnguyen@njrati.org Eric Pennington M.A. Research Assistant Rahall Transportation Institute 00 Third Ave., Huntington, West Virginia 0 Telephone: epennington@njrati.org August, 0 (Revision ) November, 0 Word count:, + Figures + Table =, Submitted for presentation at the nd Annual Meeting of the Transportation Research Board and publication in the Transportation Research Record Corresponding author

2 Dayan et al. 0 ABSTRACT Trucking companies can obtain permits from state agencies allowing them to haul oversize and overweight (OSOW) loads. Most permit systems assign a specific route to be followed that accounts for various network travel restrictions. The routes assigned in these permits can provide a wealth of information for different types of transportation analyses. Unfortunately, the permit data is not always generated by a Geographic Information System (GIS) or archived in a manner that facilitates analysis in a GIS platform. This paper presents a methodology that was developed to process archived permit records from the West Virginia Department of Transportation (WVDOT) so that they could be imported into GIS and plotted using the existing WVDOT Linear Referencing System (LRS). Some agencies do not have a GIS-based OSOW permitting system, and those who use a GIS permitting systems are still faced with the challenge of integrating archive data with existing GIS/LRS platforms. The methodology presented in this paper should be widely applicable for those facing such challenges. The automated procedure was able to assign an LRS code and map.% of the permits that contained route data for the month of July 0.

3 Dayan et al INTRODUCTION Shipping by commercial truck is an incredibly popular means of transporting goods. However this form of transportation does a great deal of damage to the nation s highways and bridges. Often, large pieces of equipment and other cargo are heavy enough that they can t be divided into smaller loads that are within the legal hauling limits for size and weight. In this situation, companies can apply for special permits issued by either the state transportation agency or state law enforcement that will allow them to exceed the legal limits. Most transportation agencies utilize automated permitting systems for proper evaluation and subsequent permitting of oversize and/or overweight vehicles. Permitting systems streamline the permitting process while allowing for optimal route selection necessary to insure that the infrastructure is protected by accounting for certain travel restrictions and travel time windows (e.g. avoid bridges with weight or height restrictions, avoid congested time periods). Permitting systems vary in automation, feature number, and feature functionality. An important feature not always available through those systems is the GIS interface, which is not only capable of presenting selected routes on a map but, more importantly, can utilize spatial routing algorithms for optimal route selection and allow for the integration of permit information with enterprise data like the locations of the road network, bridges, and other transportation assets. West Virginia (WV) utilized a permitting system that does not archive data in a format that allows easy integration with a GIS system for analysis. GIS analysis of the permit data helps to validate and calibrate automated systems, optimize route selection, assess the impact of commercial trucks on infrastructure along permitted routes, and extract and analyze travel patterns. The methodology developed in this study converts archived OSOW permit data to be compatible with GIS, particularly the West Virginia Department of Transportation (WVDOT) Linear Referencing System (LRS) of route designations and mileposts. The original concept for the data analysis discussed in this paper was presented by Chou et al. () as part of a methodology to estimate the number of overweight trucks on specific roadways that do not have the proper permits. The methodology presented in that paper required the GIS integration of archived OSOW permit data and subsequently matching the permits with weigh-in-motion (WIM) stations that the vehicle should have crossed. This paper will expand on the permit mapping portion of that methodology. Several researchers have previously examined OSOW truck aspects using GIS. Previous GIS work in this field has dealt with designing and developing routing procedures for the permits (,,), network optimization (,), and automation of route selection and permit processing (). Each of these papers uses GIS to either optimize or automate decision making to initially assign a route to the permit. These papers however do not explore actually mapping the routes in order to determine the spatial occurrence and travel patterns of permitted vehicles for planning and analysis purposes. The only study that could be found that dealt with mapping archived OSOW permits in order to explore this information for planning purposes was conducted by Li et al. (). Li developed an approach to convert the Texas DOT archived OSOW permits into a GIS format. They retrieved the data from a centralized database, processed it to be GIS compatible, and directly mapped the truck routes in a GIS system. The objectives of their research were to identify strategic infrastructure improvements to accommodate the extreme loads and to develop optimal routes for certain load groups between the most common origins and destinations using the Texas DOT highway network. A significant amount of their effort was in processing and correcting the travel route data in the permit record in order to be mapped with GIS.

4 Dayan et al The research presented here is different from Li s research primarily due to the format of the data stored for the permit route, which required a different set of procedures to format and map the data. The permit data for WV used route numbers and milepost information to designate the permitted routes, which was conducive to processing and mapping the routes using the LRS. The Texas permit data contained non-standardized route information (e.g., some route numbers and some route names) and did not contain milepost information, so utilizing an LRS was likely not a viable option in their study. The approach using LRS reduces the data processing time and allows for a more accurate determination of locations where the permit route transitions between roads (used to construct the overall route). Most state transportation agencies, including the WVDOT, use a LRS to identify and define the location of their assets on their roadway network, so this methodology should be widely applicable. Additionally, this research also includes an automated method to identify the direction of travel for each permit on a specific roadway segment, which can be useful on certain types of data analysis. This paper presents a methodology to map OSOW permits to a LRS using GIS. Although it is implemented with permit data from WV, it can be applied to data from other states whose route data is not archived in a format readily compatible with an LRS. Thus, an overview of the WV permit system is presented in the next section. It is followed by a five-step GIS mapping methodology with LRS functionality. Finally, some sample applications of the mapped permit data are presented, including the specific application referenced earlier where the permits are assigned to a weigh-in-motion system in order to quantify illegal overweight truck activity. Conclusions and future study directions are then derived. OVERSIZE/OVERWEIGHT PERMIT SYSTEM In WV, the process of acquiring a permit for an oversize or overweight truck is conducted electronically through the WVDOT website (). The West Virginia Department of Motor Vehicles issues five different permit types; Blanket, Mobile Home Blanket, Seagoing, Single Trip Mobile Home, and OS/OW/Superload. The only permit type that is assigned a specific route is the OS/OW/Superload permits because their weight limits approach the limits of some bridges across the state and require that a certain route be followed. Permit applicants submit sufficient information for the WVDOT to evaluate the vehicles, including the vehicle model, dimensions, axle spacing, axle and gross weights, desired travel dates, and origin and destination (within WV). With user defined origin and destination information, the online permit system will generate a recommended route that accounts for various factors, such as bridge weight limits, underpass height limits, and other general travel restrictions. This system appears to have some linear referencing components as it incorporates milepost information and consistent route name information in the Routes field. This system, however, does not seem to be fully integrated with the WVDOT LRS because the route name does not follow a route numbering structure common for an LRS. After the user enters this information, WVDOT personnel will review and approve, modify, or deny the permit. Of particular interest in this research are the fields in the permit database Permit ID (as a unique permit identifier), Origin, Destination, and Routes. Table shows the data contained in these four fields for a sample permit. As shown in the table, the Routes field contains the permitted route as a single string of text, which is a set of segment-by-segment instructions with the keywords START, TO, END, and RETURN. This sample route happens to be a roundtrip permit that starts on a state route (WV-), continues on the interstate (I-), and ends on a US route (US-) where it intersects with a county route (C). The

5 Dayan et al information after the RETURN keyword indicates the route to take from the destination back to the origin. The format of this data is not compatible for GIS mapping purposes without additional processing, as will be discussed in the next section. Thus, this study discusses a methodology that will convert this field of data into a format that is LRS compatible for mapping purposes. Permit data from WV for the month of July in 0 were utilized in this analysis. There were a total of,0 permit records during this time period. There were, records that contained no route information and were excluded from further analysis. These records corresponded to the permit types previously discussed that are not assigned to specific routes. Table Sample Permit Database Record Permit ID Origin Destination Routes 000 NITRO ST ALBANS START WV- W MP PUTNAM. TO WV- MP I- W TO I- MP US- S END CAND RETURN US- C TO US- MP I- E TO I- MP WV- E END WV- MP PUTNAM. LINEAR REFERENCE SYSTEMS (LRS) Linear referencing is a method of specifying a location as a distance or offset measurement (e.g., milepost) along a linear feature (e.g., a roadway), from a known reference point (e.g., milepost 0.0) (0). An LRS is one type of location referencing system with the primary benefit of establishing intuitive reference points which are easily identified in the field. For example, transportation agencies use routes and mileposts to define the locations of assets (e.g., bridges, signs, structures) and events (e.g. road conditions, traffic counts, incidents) (). The WVDOT has developed a Road Inventory Log (RIL) that uses LRS as a reference system. WVDOT RIL is a transportation network database defined and maintained in a tabular form and records transportation assets or activities on or along the route. A key feature of an LRS is a systematic way to define route segments, (i.e., Route ID). Each agency defines its own Route ID structure as well as the number designations for each component of the Route ID. The current WVDOT structure is shown in Figure. The county code is a two digit number referring to one of the counties. The road classification is a single digit from 0 to, where is an Interstate, is a US route, is a State route, etc. The route number is a four digit number corresponding to the assigned route number. The sub-route number is the assigned route designation for those roadways that have a sub-route designation. Generally, primary roadways (road classification,, and ) only have a route number and nonprimary roadways (typically road classification and above) will have both a route number (corresponding to its adjacent primary route) and a sub-route number. The supplemental code is a code that provides additional information about certain roadway characteristics (e.g., toll roads, entrance/exit ramps, spurs). The LRS codes for each road segment in the permit route will be automatically assigned within Step of the methodology. XX X XXXX XX XX County Code Road Classification Route Number Sub-route Number Supplemental Code Figure WVDOT LRS Route ID Structure ()

6 Dayan et al. 0 GIS ROUTE MAPPING METHODOLOGY To facilitate the analysis and visualization of the permit data, the route information from each permit must be converted to an LRS route for mapping in GIS. Since the Routes field in the permit database contains route and milepost information, it is feasible to assign the trips to the WV LRS after the data is converted to be compatible. The methodology used in this study to convert the route information and plot it consists of a five-step process, summarized in Figure. Step seeks to decompose the single string of text into specific transition points along the permit route, which would occur when the vehicle must turn onto a new route. Step creates a unique Route ID for every transition point, which corresponds to the Route IDs used in the WVDOT LRS. In Step, all transition points in each permit are plotted on a map. These plotted points are then connected and the segments merged to form a continuous permit route in Step. Finally, Step assigns the cardinal direction of travel on each tangent segment in the permit route. Step is not needed for plotting purposes, but provide the directional travel information at particular locations that is useful in some types of analysis and verification of routing results. 0 Figure Methodology Flowchart Step. ROUTES Field Decomposition This step decomposes the continuous text string from the Routes field into segments corresponding with a transition point in the permit trip. A Visual Basic program was written to search for key words in the route string, including START, RETURN, TO, and END, which are

7 Dayan et al then used to partition the string. This exercise is continued until the end of the string to get a complete list of transition points associated with a permit. The transition points of each permit are also identified with the order in which the transition points occurred, which is important in Step when the continuous route is constructed. The program generates a new field called order and assigns the sequence value for each transition point. This process served as the initial stage for LRS assignment and GIS transition point plotting and permit mapping. Step. LRS Assignment For each transition point in the route, the text content is further evaluated to assign an -digit Route ID that corresponds to the WVDOT LRS format. The text in each transition point field was processed automatically using a Visual Basic script that utilized lookup tables and logic statements to determine each individual component of the Route ID. Each component had a separate lookup table that contained all possible values from the WVDOT LRS. The individual components were then concatenated to form the overall Route ID, which is shown in Table. General Road Name I- US WV Table LRS Route ID Generation for Sample Permit ID 000 County 0 (Putnam) 0 (Putnam) 0 (Putnam) Road Classification (Interstate) (US Route) (State Route) Route Number 00 () 00 () 00 () Sub Route 00 (none) 00 (none) 00 (none) Supplemental Code 00 (n/a) 00 (n/a) 00 (n/a) Concatenated -digit Route ID The results of Steps and are a new data table similar to the one in Table, which lists the order, Route ID and milepost for the decomposed text of Permit ID 000, all necessary components for LRS/GIS plotting. Table LRS Compatible Transition Points for Permit ID 000 Transition Point Order Route ID Milepost WV- W MP PUTNAM WV- MP I- W I- MP US- S US- MP PUTNAM C US- N MP PUTNAM C US- MP I- E I- MP WV- E WV- MP PUTNAM Table summarizes the records that were processed in Steps and. The scripts automatically processed, of the, records that contained route information. There were, records that could not be automatically processed and required manual inspection. Problems with these records included missing key words (i.e. START, RETURN, TO, END), missing transition point information or milepost in route description, or inconsistent descriptive structure (e.g., missing indicator for an intersection). Of these, records, 0 were

8 Dayan et al successfully coded manually and had insufficient information to generate a Route ID or milepost and could not be processed or included for further analysis. The resulting permit database of,0 unique records was utilized for Steps and. Table Summary of Record Processing Statistics after Steps and Description Frequency Total Records for July 0 with Route Information, Successfully Processed with Automated Procedure in Step and Step, (.%) Required Manual Processing, (.%) Manual Route ID Generation Successful 0 Insufficient Route or Milepost Information and Excluded Step. Plotting Transition Points Using LRS The plotting of transition points from the permits was possible using the make route event layer in the LRS tools in ArcGIS. This tool uses a reference network, in this case the WVDOT LRS, to locate events or points along the network, using the Route ID field and the milepost information for each transition point as a reference. The WVDOT network already includes the Route ID field for proper matching as well as the calculated measurements, referred to as M values, along each of the routes. This allows the identification of specific routes and the location of transition points along such routes, providing alternative location determination to records or features without the use of longitude or latitude information. This layer was needed for the route generation and automated mapping process carried out in Step. The plotted points for the sample permit are shown in Figure. The GIS process generated a point feature class for all successfully located transition points as well as an error field flagging transition points that could not be automatically located on the LRS due to a problem with the Route ID or with the milepost. Two types of errors were generated in the LRS plotting process, Route Not Found and Route Measure Not Found. The Route Not Found error generally indicates a segment where the -digit Route ID generated in Step does not correspond to an actual Route ID in the WVDOT LRS. This is most likely caused by a clerical error in the reference WVDOT network. The Route Measure Not Found error indicates that the milepost in the transition point is outside the milepost limits in the WVDOT LRS. This generally occurs near the end of a route (at the maximum milepost or county line) or if the milepost information assigned by the permit system is invalid. Example for this error type is presented in Figure. In this example, a transition point within a permit was assigned the milepost., however, the route only goes up to milepost. which yields an LRS plotting error. All errors were manually inspected and resolved in an iterative process until all permit records were properly matched with the route network.

9 Dayan et al. Figure Permit 000 Transition Point Plot in GIS in Step Route ID: Milepost:. Route ID: Milepost:. Route Measure Not Found 0 Route ID: Milepost: 0.00 Figure Example for LRS Error Route Measure Not Found. Step. Connecting Transition Points to Create Continuous Route After the transition points are plotted along the road network, a continuous permit route is created by connecting these points along the mapped roadway, as opposed to a straight line connecting the points. A batch process was developed within ESRI ArcGIS Model Builder and utilized Network Analyst and Tele Atlas Premium StreetMap North America to automatically

10 Dayan et al. 0 0 construct the continuous routes (, ). The batch process is illustrated in Figure. Model elements include an iterator to process all permits in the database and their transition points. The model integrated Make Route Layer, Add Locations, and Solve tools from the Network Analyst extension in ArcGIS. The first tool creates a route analysis layer, namely Output Route, for determining the optimized routes between a set of transition points. The output is carried over to the next tool, Add Locations, which adds the transition points (Route Stops in Figure ) to the network created by the first tool. All transition points are sorted by the order value assigned in Step for a hierarchical assignment. Both the outputs for the Add Locations and Make Route Layer tools are analyzed to solve the network analysis layer. The Solve tool determines the optimal route by identifying the barriers and constraints within the network transition points, and accounting for their hierarchical order by which their connectivity is determined. The final output, Network Analyst Layer Solved captures the actual traveled route for each permit within the database, accounting for all transition points. Figure shows Permit 0000 plotted in GIS after connecting the transition points in Step. Notice that the plotted route follows the mapped roadways rather than the straight line distance, which would have missed the ramps connecting I- to US. Figure GIS Batch Routing Model for Step

11 Dayan et al. Permit Figure Permit 000 GIS Plot in Step Step. Assigning direction of travel to route segments It was desirable to assign the direction of travel to certain segments of the permit route in order to facilitate future directional analysis. In order to derive the directional information of a route at any specific location, the permit was segmented in order to identify tangent sections. Once the straight segments were identified, the Linear Directional Mean (LDM) was calculated using the corresponding ESRI ArcGIS Spatial Statistics tool. The LDM computes the azimuth for a line, referenced from north (0 degrees) in a clockwise direction. A Python script was used to convert the azimuth to one of the four primary directions North, South, East, and West. Figure illustrates the azimuths assigned to each travel direction of the segment of I- in Permit 000. Permit 000 Figure Assigned Azimuth for I- Segment from Permit 000

12 Dayan et al. 0 SAMPLE APPLICATIONS Matching Permits to WIM Stations This methodology was applied in previous research to help estimate the percentage of overweight trucks on certain roads that do not have proper permits. This was accomplished by integrating the permit data discussed here with weight data measured in the field at specific locations. In WV, truck weight data is collected at weigh-in-motion (WIM) stations. The physical location of the WIM stations was known, so by plotting their location in GIS and identifying the permits that should have crossed the corresponding tangent roadway section, the permit data and the WIM data could be directly compared. Figure shows the frequency of unique permits that crossed each WIM station during July 0. It is easy to see that the WIM sites located along interstates and other primary routes experienced more permitted overweight activity. By comparing the quantity of permits crossing the WIM with the actual overweight truck counts crossing the WIM, compliance rates were estimated. This information is useful for overweight enforcement purposes. Figure Frequency of Permits Crossing WIM Sites (July 0)

13 Dayan et al. 0 Statewide Roadway Permit Frequency To protect the highway infrastructure and prioritize maintenance activities, a map illustrating the routes that overweight trucks are taking is useful. Figure illustrates the mapped results from the,0 permits that were processed for July 0. The roadway segments with the highest frequency of permitted loads can easily be identified, which tend to be the interstate system. I- between Morgantown, WV and the Pennsylvania border showed the highest number of truck permit loads in the range of, to, permits during the study period. Heavy truck loaded roadway segments warrant frequent inspections to protect the safety of the infrastructure. Additionally, this information might assist the authorities in selecting segments for permit compliance enforcement. 0 Figure Statewide Frequency of Permits on Roadway Network (July 0) Origin-Destination Analysis An origin-destination (OD) matrix can be created to identify all OD pairs, which can be used to identify frequently used travel paths, which can be useful in planning new facilities or upgrading existing ones. As an example, all permits with an origin in the vicinity of Nitro, WV were plotted, both as Euclidean Distance (Figure 0) and the permitted routes (Figure ). The data in Figure 0 could have been plotted without the data processing described in this paper since it

14 Dayan et al. does not utilize the route information. Figure can only be derived after processing the route data, using the procedure discussed here. Figure 0 Euclidean Distribution of Permit OD Pairs Leaving Nitro, WV (July 0) 0 0 Figure Actual Route Distribution of Permit OD Pairs Leaving Nitro, WV (July 0) CONCLUSION This paper presents a methodology to convert OS/OW permit data into a format compatible with a LRS for GIS mapping purposes. Overall, the methodology resulted in successful mapping of.% of the permits that contained route information during July 0 in WV. This methodology should be widely applicable, as most state transportation agencies manage their assets with a LRS and have network established. Results from the mapped permit routes can be integrated with other databases to derive valuable knowledge for traffic analysis and planning purposes. For example, the integration between permit routes and WIM locations can facilitates the derivation of illegal overweight truck activity. This can improve enforcement mechanisms, but can also contribute to roadway planning, construction, and maintenance, creating a safer infrastructure network.

15 Dayan et al. 0 With this analysis, the actual route that is occupied during travel can be determined, rather than using a less exact Euclidian distance. The mapping procedure can also provide information on OSOW permit travel frequency, which could be used in economic development analysis, land use planning, transportation planning, and other analyses. ACKONWLEDGEMENTS This work was supported by the West Virginia Department of Transportation, Division of Highways and the Nick J. Rahall II Appalachian Transportation Institute at Marshall University. The contents of this paper reflect the views of the authors, who are responsible for the facts and the accuracy of the data presented herein, and do not necessarily reflect the official views or policies of the sponsoring organizations. These contents do not constitute a standard, specification, or regulation. REFERENCES Chou, C.-S., A.P. Nichols, S. Yoo, and M. Cetin, "Methodology to Estimate Percent of Overweight Trucks Without Proper Permits," Proceeding of Transportation Research Board nd Annual Meeting, no Osegueda, R., A. Garcia-Diaz, S. Ashur, O. Melchor, S.-H. Chang, C. Carrasco, and A. Kuyumcu, "GIS-based network routing procedures for overweight and oversized vehicles," Journal of Transportation Engineering, no. pp.-,. Nord, M., and G. Hovey. "Load Rating and Permit Vehicle Routing." In Eighth Transportation Research Board Conference on Bridge Management, no. L-, IBMC-0.. Datla, S.K., R.S. Moorthy, and K.K. Rao, "A GIS for Routing of Oversized and Hazardous Material Carrying Vehicles," Proceeding of Map Asia Conference, pp Adams, T.M., S. Malaikrisanachalee, C. Blazquez, S. Lueck, and A. Vonderohe, "Enterprisewide data integration and analysis for oversize/overweight permitting," Journal of computing in civil engineering, no. pp.-, 00. Ray, J.J. "A web-based spatial decision support system optimizes routes for oversize/overweight vehicles in Delaware," Decision Support Systems, no. pp.-, 00. Adams, T.M., S. Malaikrisanachalee, C. Blazquez, and A. Vonderohe, "GIS-Based Automated Oversize/Overweight Permit Processing," Computing in Civil and Building Engineering, pp.0-. ASCE, 000. Li, Y., J.T. Le, D.R. Middleton, and C.A. Quiroga, "Mapping Oversized and Overweight Truck Routes with Procedure Based on Geographic Information Systems," Transportation Research Record: Journal of the Transportation Research Board, no. pp.-, 0. Hauling Permits, Website, West Virginia Department of Transportation, accessed June, 0, 0 Federal Highway Administration, Federal Highway Administration Linear Referencing Practitioners Guidebook, GIS/Trans Ltd,. Curtin, K.M. Linear Referencing, The Encyclopedia of Geographic Information Science, pp. -. K. Kemp ed. Sage Publications, 00.

16 Dayan et al. Geospatial Transportation Information, Website, West Virginia Department of Transportation, accessed June, 0, ESRI 0. ArcGIS Desktop: Release 0.., Redlands, CA: Environmental Systems Research Institute, 0. ESRI ArcPad 0 StreetMap Premium Tele ATLAS North America, Redlands, CA: Environmental Systems Research Institute. 00.