Home

Final Report - Wisconsin Transportation Center

1. suonoes YIOMION uno eysou y im rer ni TE A nm mi MI vil A 17 BEFORE Roadway centerlines necessary AFTER Roadway centerlines are digitized to allow turns on and off of ramps are from orthophotos allowing travel on the omitted from dataset Two small segments highway and all possible turns on and off of the freeway are omitted the highway Figure 16 Missing Roadway Segments A EA BEFORE Travel from off ramps directly AFTER Travel from off ramps directly back to on ramps is not permitted back to on ramps is permitted and all other necessary topology is preserved Figure 17 Roadway Segments Require Modification In other places the manner in which roadways were represented could not be integrated into an accurate roadway network The most common problem of this nature was encountered when a single roadway centerline was used to represent two adjacent unconnected lanes of traffic flowing in opposite directions This often occurred on ramps 18 see Figure 18 There were also number of undershoots and overshoots as with the Kenosha County data BEFORE Two lanes of on ramp off AFTER The single linear feature is re ramp structure separated by a physical digitized as two features using barrier are improperly represented as orthophotos
2. e k EF rm in Bold Red Figure 11 Patrol Sections Columbia County Network Shapefile Columbia Road Pavement Centerlines Shapefile Description Attributes Keywords Theme WiscPlow Wisconsin Winter Maintenance Concept Vehicle Place Wisconsin Columbia County Description Abstract The roadway spatial database was created by the UW Madison Research Team in conjunction with the Land Information Department in Columbia county It depicts the roadway centerlines in Columbia County and contains attributes used by the WiscPlow software among them attributes necessary to create a network shapefile using Netshape The spatial database was first obtained from the Columbia Land Information Department and some of attributes for roadway centerlines were modified or created for use of WizcPlow Purpose The roadway spatial database for Wiscplow is utilized to represent locations of the Winter Maintenance Concept Vehicle on pavement centerlines within Columbia County Information contained in this database was designed and modified to show performance measures of winter operation Status of the data Time period for which the data is relevant Publication Information Data storage and access information Details about this document Contents last updated 20060807 at time 12042700 Standards used to create this document Standard name FGDC Content Standards for Digital Geospatial Metadata Standard vers
3. gt lt Average Temperature F Patrol Section fH Salt GSand a Pavement Temperature Figure 35 Example Material Analytical Decision Tool Analytical decision tools can be printed or stored as Microsoft Excel spreadsheets 6 1 9 REPORT Equipment Tab The REPORT Equipment Tab appears in Figure 36 Users can request up to ten reports and Ge WiterP live iteration and Pas Events Eniry L amp bor TART Process REFET Cost Operations Cox for each selected attachment unit for each event and patrol Section for all attachment units foreach event and patrol section Cast for all attachment unas for each event All roads Percentage of operating time season to date f r each attachment type Interstate Federal and State roads only All roads Total operation time season To date lorcach Federal and State roada only Compound Measures Chowne duration factor for selected stanm s Cycle time for each patrol section by selected storms ata rgo Samen abtachnmient typi erstave and Stale roads on Analytical Decision Tools r gurult hours for nels al arn egugpnmuent type peril hours for selected type ard selected event Produce Results Figure
4. sAewe ubisse pue opooaq 549 55 4 3dooadd shee oBessoui 539 554 5 54 50 3 WISCPLOW FOLDER STRUCTURE PURPOSE This chart shows the general structure of WiscPlow Version 3 beginning with the main WiscPlow folder and moving down into the contents of the folders and files within WiscPlow current database Folder This folder contains the two databases below They are crucial to WiscPlow current mdb database Tables contain information sorted from specified datafiles in datafile databases Folder These tables are used when calculating performance measures info mdb database Tables contain information such as county projection parameters and labor rates Datafiles Folder Contains raw unprocessed datafiles e g 01040389 003 It is not necessary to keep datafiles here but is provided as an option to the user datafile databases Folder Contains Access databases specific to each processed datafile e g 10210418 807 mdb These databases are built and datafile shapefiles Folder Contains shapefiles of points from GPS data for each datafile that has been processed These shapefiles are added to the map display in WiscPlow Each shapefile consists of 6 7 files with a variety of extensions shp dbf shx sbx sbn prj 51 pathways Folder Contains the two text files shown below current ini file This file tells WiscPlow where certa
5. Sun mon Tue wea Thu Sat 5 gt bo h b ao qu 2 ha ha hs az hs 2 ba 25 pr 2 b b Figure 25 Operator Information Screen 23 Pushing the Enter New Operating Day button on the Labor Data Entry Tab causes the screen in Figure 26 to appear Here the start and end times for an operating day and the effective start and end dates for that operating day are picked from scrolled lists and calendars respectively This information is necessary to assign regular and overtime wages to work times reported by the vehicles Enter New Operating Day Operating Day Times Start Time 9 00004M End Time _ 4 00 00PM Effective Dates Start Date End Date Jan 2005 Jan 2005 Dec 2005 p Dec 2005 Tue rue Wed 26 27 js FE 6 RT Figure 26 Enter New Operating Day Screen Pushing the View Labor Rates Information button on the Labor Data Entry Tab causes a county dependent labor rates entry screen to appear The screen for Columbia County is shown in Figure 27 Data in the table can be modified by deletion of existing and insertion of new full records Effective start and end dates for labor rates are chosen from calendars and hourly w
6. Code 80 10 Dry Salt Drine E 50 40 salt Magnesium 70 5 30 salt m IMS0 RO Enter New Product Delete Selected Agenutis Figure 23 Material Data Entry Tab 22 6 1 4 Labor Data Entry Tab The Labor Data Entry Tab appears in Figure 24 Data on operators and operating days are displayed along with a View Labor Rates Information button Data in the tables can be modified by deletion of existing and entry of new full records Gol WisePlow Empan Entry and Eniey baiiia Data Friry make changes in the tables you musi enter ihe password in iha Information and Password iab and prese the Allow Table Editing bution Operator Details Operating Day Enter New Delete Selected Operating Day Operating Day Labor Rates View Labor Rates Enter Hew Operator Delete Selected Information Details Operator Details Figure 24 Labor Data Entry Tab Pushing the Enter New Operator Details button cause the screen in Figure 25 to appear Here users identify operators by name and ID and pick the associated hiring date from a calendar Hiring dates are necessary because some labor rates depend upon seniority Operator Information Oj x Operator Operator Details OperatorlD 111 Operator Name Jay Date Hired Jan 2005 Jan 2005
7. XXXV XXXVI XXXVII XXXIX xl xli xlii CreatePrd2Cal CreateLnWdthCal CreateSndRt CreatePrwtSndRt CreateSItRt CreatePrwtSItRt CreatePrd1Rt CreatePrwtPrdl Rt CreatePrd2Rt CreatePrwtPrd2Rt CreateAntIceRt CreateSpnrRPM CreateT mprtr CreateSpeed CreateAlarm CreateWyPntID CreateAttFrontPlow CreateDmpBdy CreateAttLeftWing CreateAttRightWing CreateAttScraper CreateSprdr Create AttSpreader CreateAttSprayBar CreateAttTruck CreateBrdCrmbsData CreateEquipUsage b PopulateDatabase Calls all of the following procedures to populate the tables in each datafile s database using the arrays containing the decoded messages from decodemessagesfor12 1 11 111 IV V 1 FillPrwtMtrCal Xl Xll XVI FillBrdCrmbs FillOperator FillOprtngDay FillGPSStatus FillTnkVol Commented Out Right Now FillApplZone FillBlstCal FillVehicle FillRectnTime FillSpdCal FillSndCal FillSItCal FillAttFrontPlow FillDmpBdy FillAttLeftWing 47 10 xvii FillTmprtr FillSprdr xix Cal xx FillPrd2Cal xxi Fill WyPntID xxii FillAlarm xxii FillSpeed xxiv FillLaWdthCal xxv FillAntIceCal xxvi FillSndRt xxvii FillSItRt xxviii FillPrwtSndRt xxix HillPrwtSItRt HIIPrdlRt xxxi FullPrd2Rt xxxi FillPrwtPrdl Rt xxxiii FillPrwtPrd2Rt xxxiv FillAntlceRt xxxv FillSpnrRPM xxxvi FillAttScraper xxxvii FillEquipUsag
8. scales of 1 24 000 or larger These shapefiles must include patrol sections They must also be attributed with roadway functional class Custodianship of these databases is a difficult issue Options include 1 WIsDOT s local roads databases are maintained at headquarters in Madison The unit that does this work clearly has personnel with appropriate skills and technology for maintaining Wiscplow spatial databases However such activities are not within their mission WisDOT s district offices use GIS as a matter of routine and could develop the capability to maintain Wiscplow s spatial databases However WisDOT district offices typically do not maintain or manage data for county highway departments It is technically feasible for UW Madison to maintain the necessary spatial databases However the university is not in the practice of doing such things and some mechanism would have to be developed for continuity beyond the tenure of current Research Team members Some level of continued funding would have to be developed Consultants could be contracted to maintain the spatial databases Such an arrangement would perhaps be more institutionally stable than having UW Madison do the maintenance but a continuing funding source would have to be identified The most attractive option is to have either the county highway departments or the county land information offices take custodianship of the spatial databases This keeps both use a
9. AmitQ tryaltnext try2to4 try2to5 try 1t03 speedonepoint speedzero GetOther CheckSpeed GetMinDist QueryID PercentAlong PatrolDistance FindPath OpenShapeFile This procedure populates tables in the current database using data from the datafile database which was created earlier The tables in the current database are used directly by WiscPlow to compute performance measures The following procedures are called and they populate different tables in the current database using information extracted from the datafile s database a b oo Laborhours Calculates number of hours an operator worked and populates OperatorHours table of current database Laborpatrol Calculates normal and overtime hours worked and populates OperatorPatrolHours table in current database wntreport1 Populates PaveTemp table in current database wntreport2 Populates tables related to salt rate in current database wntreport3 Populates tables related to sand rate in current database wntreport4 Populates tables related to pre wet salt in current database wntreport5 Populates tables related to pre wet sand in current database wntreport6 Populates tables related to anti ice in current database 12 MMcall Populates tables in current database related to machinery management a MM table This procedure is called several times while populating the table called MMQuery E
10. Interviews of personnel within all eight participating counties and key personnel at WisDOT to identify needs 2 Development of a draft model for the data constituting a component of an underlying framework for applications 3 Development of core GIS functionality for conversion of point data as collected by differential Global Positioning System DGPS units aboard the vehicles to data associated with roadway centerlines and therefore with roadway names and patrol sections 4 Testing of the positional accuracy of the DGPS units Research during Year 2 included Vonderohe et al 2001 1 Establishment of a permanent DGPS test facility on the UW campus 2 Extension of a Microsoft Excel application initially written by Raven Inc for winter storm reporting This application was documented and distributed to participating County Highway Departments 3 Refinement and implementation of the model for detailed databases comprised of decomposed message sets and other data 4 Identification of goals objectives and performance measures for Machinery Management and Winter Storm Report applications 5 Development and implementation of a data model for summary databases founded upon the identified performance measures 6 Initial stage development of a GIS based prototype for Machinery Management and Winter Storm Report applications 7 Investigation of the feasibility of Real Time Monitoring application development 8 Ex
11. The feasibility of the shortest path between snapped points 2 and 4 is then tested If this path is not feasible the algorithm continues by Snapping preceding points 2 1 and 0 one at a time to alternative roadways and determining if feasible paths exist between these newly snapped points and snapped point 3 If none of the five consecutive points 0 through 4 aid in solving the spatial mismatch between the snapped points for points 2 and 3 then it is likely that no roadway centerlines within their buffers yield a feasible path and larger buffers and or more consecutive data points need to be utilized by the algorithm Once a feasible path is obtained intermediate points not used during the map matching process are snapped to the roadway along that feasible path On the other hand if a path is feasible between snapped points 3 and 4 then following the flow chart in Figure 6 alternative roadway centerlines are sought within the buffer for point 2 The algorithm in Figure 7 checks for feasibility of paths between the newly snapped point 2 and the previous and subsequent snapped points If there are no remaining alternative roadways within the buffer for point 2 or if there is no feasible path between point alt2 and the neighboring points then the algorithm in Figure 6 checks the feasibility of the path between snapped points 1 and 3 If this path is feasible point 2 is snapped to the roadway along that path and the spatial ambiguity is resol
12. monitoring capability and is not portable as it requires access to databases that are typically only available on client server networks Wiscplow presents users with a ten tab interface see Figure 20 The tabs control data entry data analysis and reporting and display functions 19 4 amp TART Procart 51 i z MEM E E EI Figure 19 Cada REPORT REPORT Information and Password Wainber Ewents Entry Matera Date Entry REPORT Labor THE UNIVERSITY WISCONSIN M A DISC M WiscPlow Version 1 0 Microsoft Comporations Visual Basic 6 08 programming environment ESRI s Environmental Syrians Research Inc MapObjects Lheveloped ai the Unrversity of Wisconsin Lreparimerit of Civil and Environmental E ngmeenng Madison Wireman Funded by the Department of Teanespestatuon aed Regional Draneporbatuom C enber THE UW MADISOM DISCLAIMS ALL WARRANTIES WHETHER EXPRESS IMPLIED OR STATUTORY IMCLUDIMG BUT HOT LIMITED TO WARRANTIES OF MERCHANTABILITY FITHESS FOR PARTICULAR PURPOSE OF THIRD PARTY RIGHTS EVENT SHALL THE BE LIABLE FOR INDIRECT SPECIAL HCIDENTAL EXEMPLARY GB CONSEQUENTIAL DAMAGES OF INCLUDING LOST PROFITS LOSS OF USE INTERRUPTION Or BUSINESS OR FOR LEGAL FEES ARISI
13. specified file selected by the user from a drop down menu from the Process List by removing its record from the Batch Process table 1 the current database Command12 Click Process File s in List Button When the user is ready to process datafiles they press the Process File s in List button and this procedure runs Everything that follows is looped through for every datafile to be processed Files to be processed are listed 1n BatchProcess table of current database 2 erasemessage erases all arrays containing decoded messages 1 e Messagel4 3 decodemessagefor12 Decodes raw data and stores the results in arrays 4 BuildDatabase First this procedure calls CreateDatabase to create a database for each datafile and appends tables to the database Then it calls PopulateDatabase to populate the tables using values stored in arrays a CreateDatabase Calls all of the following procedures to create specified tables then appends all of the tables to one database i CreateOperator CreateOprtngDay 11 CreateTnkVol iv CreateBrdCrmbs v CreateGPSStatus vi CreateApplZone vii CreateBlstCal viii CreateVehicle ix CreateRectnTime x CreatePrwtMtrCal xi CreateSpdCal CreateSndCal CreateSItCal xiv CreateAntIceCal xv CreatePrd1Cal 46 XVI XVII XXII XXIV XXV XXVI XXVII XXVIII XXIX XXX XXXI XXXIII XXXIV
14. 0 M17 8 17 45 c4 1500 0 200 200 600 800 1000 1200 1400 1600 1800 2000 12 M18 8 17 45 c5 1500 0 200 200 600 800 1000 1200 1400 1600 1800 2000 12 M20 8 17 45 c6 60 0 143 20 25 30 35 40 45 50 55 60 65 M21 8 17 46 c7 10 0 0 0 0 12 144 144 144 Figure 30 The First Few Records of a Typical Vehicle Data File 26 wii wd Parisoeword Wieder START of to Process hear ame Details ID Evert Date End Dale Begin Here file to process lis Select Eveni 500 Sion ID Select Operator 1234 John Doe vehicle ID 151 Dry Agent Remove File from Process List Select bo renee Clear file from process list Close WiscPlow Remove File from Database Select file to remove Clear Remove file from database Figure 31 Associating a Vehicle Data File with a Vehicle an Operator a Storm Event and Materials Pressing the Process file s in list button initiates a sequence of operations on each data file in the list The records are parsed and placed in a database that includes the associations with vehicle operator storm event and materials indicated earlier by the user Latitudes and longitudes of vehicle data points are transformed into the map proje
15. 1 1 Roadway Spatial Databases 9 1 2 Vehicle Data 9 2 Institutional Options for Maintenance of Wiscplow Source Code 10 Summary and Conclusions List of References Appendix A Wiscplow External Technical Documentation vi 2f 30 32 33 33 33 33 36 36 38 38 38 38 39 39 40 41 42 Executive Summary During the winter of 1999 2000 the Wisconsin DOT WisDOT in conjunction with Highway Commissioners from eight Wisconsin counties embarked upon a 5 year research and implementation program for Winter Maintenance Concept Vehicles In the first two years concept vehicles were deployed in the Counties of Columbia Florence Manitowoc Barron Kenosha Trempealeau Portage and Taylor These vehicles contain numerous on board sensing devices including AVL GPS material spreader sensors blade status sensors vehicle speed sensors and air and pavement temperature sensors Data from these sensors are not only recorded in message sets on board the vehicles they are also telemetered in real time to base stations in County Highway Offices for monitoring and operational purposes Data of this nature in kind and in quantity have never before been available to transportation agencies in Wisconsin All data coming from these vehicles are spatially and temporally referenced and are therefore ideal for management and analysis with geographic information systems GIS technology During years 1 4 of the W
16. 1 6 START Process Datafiles Tab The START Process Datafiles Tab appears in Figure 29 At this tab users select vehicle data files for processing and associate vehicle data with vehicles materials operators and winter events Vehicle data files are recorded on board the vehicles on PCMCIA cards that are later downloaded for processing by Wiscplow The first few records of a typical vehicle data file appear in Figure 30 The records contain information on time date latitude longitude vehicle speed material spreading rates equipment status and air and pavement temperature These records are formatted according to Raven s DCS710 or DCS710A console standard When the Add a file to process list button is pushed a browsing screen appears and the user navigates to the appropriate folder and selects the data file to add to the list This activates the Details box on the START Process Datafiles Tab see Figure 31 The user then associates information with the selected vehicle data file by choosing from pick lists in the Details box When all details have been picked the user presses the Apply Details button invoking indicated relationships among the data and displaying them in the updated list of data files to 29 process Any number of data files can be added to the process list in this manner Wiscplow is structured in this way because large data files require significant processing time After a long storm event analysts might
17. 36 REPORT Equipment Tab 30 Up to two analytical decision tools Figure 37 shows an example report and Figure 38 shows and example analytical decision tool 2 trick cost Ene earh event pairol aertion START Pr gr Zoom ros Columbia County Equipment Eeport Individual truck allachmenl unit cost per event per patrol seclion Attach Equip Type Class ID Event II Patrol Section Coat 5 attachment type Truck achment type Truck Truck Truck Truck Truck Truck Truck Figure 37 Example Equipment Report Weekly Cumulative Operating Hours for Units of an Attachment Class o 5 o E Q O Week Ending Figure 38 Example Equipment Analytical Decision Tool 31 6 1 10 REPORT Labor Tab Figure 39 shows the REPORT Labor Tab Users can request reports on up to five performance measures Figure 40 shows an example labor report Figure 39 REPORT Labor Tab 591 cost for each selected event 8 Columbia County Labor Report Labor cost for each selected event OperatorID Operator Normal Overtime 123 Blazquez 0 00 0 30 234 James 0 00 0 95 234 James 0 36 0 00 989 Sungchul 0 00 0 95 Figure 40 Example Labor Report 22 6 2 Testing and Validation of Wiscplow Two series of tests were applied to Wiscplow s functionality The first series was used to
18. 4 Material Data Entry 20 3 4 1 New Product 21 3 5 Labor Data Entry 22 3 5 1 New Operator Details 23 3 5 2 New Operating Day 24 3 5 3 New Labor Rates 25 4 Processing Data Files 27 4 1 Procedures to Process Data Files 27 4 2 Remove Data Files in List of Data Files to Process 30 4 3 Remove Data Files from Database 31 5 Report and Chart Generation 32 5 1 Material Report 32 5 1 1 Generating Reports 33 5 1 2 Generating Charts 34 5 2 Labor Report 38 5 2 1 Generating Reports 38 5 3 Equipment Report 40 5 3 1 Generating Reports 40 5 3 2 Generating Charts 42 6 Map Display 45 6 1 Roadway Centerlines and Patrol Section Map 45 6 2 Vehicle Route Display 46 6 3 Map Display Tools 47 Appendix A Wiscplow Tutorials 49 Appendix B Creating a Network for Wiscplow 107 Figure 43 Table of Contents for Wiscplow Users Manual 37 8 Installation and Training at County Highway Departments The Wiscplow system including user documentation tutorials test data and roadway centerline network shapefiles were installed at the Columbia and Portage County Highway Department Offices County Information Technology staff were on hand for the installation as were County Highway Department personnel Training was provided at each installation site Wiscplow is now in the hands of local level analysts for use in business operations The system was also installed at WisDOT Headquarters in the Bureau of Highway Operations
19. Private Sub Command1 Click OK button On Error Resume Next Dim addedInformation As Boolean addedInformation False Set newdbinfo DBEngine Workspaces 0 OpenDatabase DirPathInfoDbase Set newtblinfo1 newdbinfo OpenRecordset AgentLiquid Set newtblinfo2 newdbinfo OpenRecordset AgentDry Check to make sure that liquid agent and code both exist if one exists If DataCombo2 Text And DataCombo3 Text lt gt Then MsgBox You must select liquid agent voExclamation WiscPlow Exit Sub End If If DataCombo3 Text And DataCombo2 Text lt gt Then MsgBox You must select a code for the liquid agent vbExclamation WiscPlow Exit Sub End If dry agent has been selected add it to database If DataCombo1 Text lt gt Then With newtblinfo2 AddNew Fields DryAgent DataCombo1 Text Figure 41 Example Internal Source Code Documentation for OK Button Comments are in Bold Type Winder Events Products To make changes in the tables you must enter the password in the Information and Password tab and press ihe Allow Table Editing bution E BGrid2 Dry Agent Liquid Agent Digan Condes b 9055 sond 10 Ory Cacl2 Brine 5 send diss Magnesium Chl nde DBGrid1 fits sand 3025 sa Zall Bune 1055 3 MSU Road Deicer RD nene nan Enter Mew Product Delete Select
20. does not endorse products or manufacturers Trade and manufacturers names appear in this report only because they are considered essential to the object of the document ii Acknowledgements The authors wish to thank the Highway Departments and Land Information Offices of Columbia Kenosha and Portage Counties all of whom provided data and much assistance throughout this project The Marathon County Land Information Office provided data The Bureau of Highway Operations at the Wisconsin Department of Transportation is thanked for its continuing enthusiastic support of the work reported herein The Midwest Regional University Transportation Center at the University of Wisconsin Madison provided funding and administrative support 11 EXHIBIT B Technical Report Documentation Page 1 Report No 2 Government Accession No 3 Recipient s Catalog No 04 01 CFDA 20 701 4 Title and Subtitle 5 Report Date August 2006 GIS Tool to Measure Performance of Winter Highway Operations VEM 6 Performing Organization Code 7 Author s 8 Performing Organization Report No Alan Vonderohe Carola Blazquez Nick Brezovar Sungchul Hong Jason Lauters Teresa Adams MRUTC 04 01 9 Performing Organization Name and Address 10 Work Unit No TRAIS Midwest Regional University Transportation Center University of Wisconsin Madison CB NO 1415 Engineering Drive Madison WI 53706 0092 05 22 12 Sponsoring Organization Name and Addres
21. each allowing only one linear feature with bidirectional traffic unidirectional flow flow Figure 18 Ramp Requires Re Digitizing to be Topologically Correct After editing of the cartography by one of the Research Team members a second Team member visually inspected the entire data set for registration with the countywide orthophoto The network integrity of the final network shapefile was confirmed with a pathfinding engine and with Wiscplow The cartography of the final Portage County network shapefile is shown in Figure 19 There are two layers 1 Roadways and 2 Patrol Sections In Figure 19 patrol sections are highlighted in bold red As with Columbia and Kenosha Counties metadata for the network shapefile including descriptions of all attributes were prepared according to Federal Geographic Data Committee Content Standards for Geospatial Metadata These metadata are embedded within the network shapefile database The complete Portage County network shapefile was installed along with Wiscplow at the Portage County Highway Department Office see Section 8 The network shapefile along with its metadata is included on a CD attached to the master copy of this report 6 Completion and Testing of Wiscplow Functionality 6 1 Wiscplow Functionality Wiscplow is intended for use by analysts and other technical staff in a post processing environment It is a standalone application not an enterprise solution It has no real time
22. maintenance at the local level Those who need the data become responsible for it Impediments include lack of skills in spatial database maintenance at county highway departments and lack of resources at county land information offices 9 1 2 Vehicle Data Management of large volumes of vehicle data over time might become problematic Current file naming conventions include acquisition date and vehicle identifier This convention should be adequate for identification of individual files However as the number of instrumented vehicles grows and the stream of in coming data increases the shear volume of information might become unmanageable without careful procedures It is recommended that vehicle data be archived on a seasonal basis so that data at hand for Wiscplow operations is for the current season only If there is need to analyze data for multiple seasons for example to compare performance measures from year to year the necessary data can be removed from its archive on an as needed basis 9 2 Institutional Options for Maintenance of Wiscplow Source Code On going maintenance of Wiscplow s source code beyond Year 5 is problematic Options include 6 WisDOT s Bureau of Automation Services has the skills necessary for maintenance of Wiscplow However BAS has a long standing policy of not maintaining software developed by other parties 7 The source code could be maintained at the county level by land information offices or
23. of Year 4 Therefore it was proposed that the Research Team in conjunction with selected counties develop the data during Year 5 5 Description of options for short and long term institutional arrangements for management of winter maintenance vehicle data roadway spatial databases and the Wiscplow system These options are re articulated near the conclusion of this report 3 Year 5 Research Objectives Year 5 was envisioned as the concluding effort in 1 development of concepts metrics data and technologies for management and exploitation of the rich data streams arising from winter maintenance concept vehicles and 2 planning for extended implementation use and management of the developed data and technologies into the future Overall objectives of the Year 5 research were to 1 Develop and implement a solution to the spatial ambiguity problem 2 Develop and test the necessary roadway spatial databases for selected participating counties 3 Complete development and testing of Wiscplow functionality including performance measures decision management tools and spatial database management tools 4 Develop complete technical and user documentation for the Wiscplow system 5 Install and provide training on Wiscplow and the spatial databases at each selected participating County Highway Department 6 Re articulate the options for short and long term institutional arrangements for management of winter maintenance vehicle
24. of a state highway extending north into Marathon County The Marathon County Land Information Office provided a digital cartographic representation of this roadway segment which was then transformed into the Portage County coordinate system and merged with the Portage County roadway spatial database Patrol sections were embedded in the network shapefile by manually inserting their boundaries based upon the hardcopy map Attributes for patrol section identifiers were manually entered Other attributes and the topological structure required by Wiscplow were manually entered Topological testing with a pathfinding engine and with Wiscplow revealed numerous problems that had to be addressed by editing to ensure network integrity One type of problem existed at a number of highway interchanges Crossing highway segments might represent overpasses underpasses or at grade intersections Nodes must be inserted at at grade intersections to allow turns Nodes must be removed at overpasses and underpasses to disallow turns The orthophoto served as a visual reference in resolving these problems In some places roadway segments were missing from the database and were manually digitized from the orthophoto see Figure 16 In other places all topologically necessary roadway segments existed but did not allow for all possible ways of travel through the network These segments had to be modified or re digitized to correct the problems see Figure 17 16
25. GIS TOOL MEASURE PERFORMANCE OF WINTER HIGHWAY OPERATIONS x 49 Project04 01 li d August 2006 Midwest Regional University Transportation Center College of Engineering Department of Civil and Environmental Engineering University of Wisconsin Madison W Authors Alan Vonderohe Carola Blazquez Nick Brezovar Sungchul Hong ason Lauters Teresa Adams University of Wisconsin Madison Investigator Alan Vonderohe Professor Department of Civil and Environmental Engineering University of Wisconsin Madison DISCLAIMER This research was funded by the Midwest Regional University Transportation Center The contents of this report reflect the views of the authors who are responsible for the facts and the accuracy of the information presented herein This document is disseminated under the sponsorship of the Department of Transportation University Transportation Centers Program in the interest of information exchange The U S Government assumes no liability for the contents or use thereof The contents do not necessarily reflect the official views of the Midwest Regional University Transportation Center the University of Wisconsin the Wisconsin Department of Transportation or the USDOT s RITA at the time of publication The United States Government assumes no liability for its contents or use thereof This report does not constitute a standard specification or regulation The United States Government
26. IT units However they would have to agree to do this Furthermore each installation of Wiscplow would evolve differently over time as each would be maintained separately Eventually data models and therefore databases would become incompatible across installations 8 UW Madison could potentially maintain the source code Some mechanism would have to be developed for continuity beyond the tenure of current Research Team members Some level of continued funding would have to be developed 9 Consultants could be contracted to maintain the source code A continuing funding source would have to be identified 39 10 The source code could maintained on retainer basis by member members of the Research Team A continuing low level funding source would have to be identified 10 Summary and Conclusions GlS based decision support system for winter highway maintenance has been developed tested documented and installed at two county highway department offices and WisDOT headquarters Wiscplow computes performance measures and produces reports and decision management tools based upon them Testing included both computational validation and integrity of the user interface Wiscplow requires accurate network shapefiles of roadway centerlines and patrol sections These were developed along with associated metadata tested and delivered to three Wisconsin counties Control procedures were used to ensure quality in the net
27. NG OUT OF THE USE OF THIS LICENSE AGREEMENT REGARDLESS OF THE FORM OF ACTION WHETHER IN CONTRACT OR T INCLUDING HECLIGEHCE Password Intonation Help Help Chek to obtain Uber Allow table editing Tables curently locked for eating amier password to edri Change Password Figure 20 Wiscplow User Interface with Ten Tabs 20 Portage County Network Shapefile Patrol Sections in Bold Red Labor Dein Enory 6 1 1 Information and Password Tab Figure 20 shows the Information and Password Tab The tab displays acknowledgements and disclaimers presents users with a Help button and has functions for entering and changing passwords The Help button opens a 109 page Microsoft Word file of user documentation see Section 7 2 There is no searchable index or help wizard Entry of a password allows entry and modification of material equipment and labor data After entry or modification of these data users may lock the database for security by returning to the Information and Password Tab and pushing the appropriate button 6 1 2 Winter Events Entry Tab The Winter Events Entry Tab appears in Figure 21 Here users enter data on storm events incidents and anti ice events Wiscplow automatically assigns identifiers to all events For each type of event displayed data tables contain start dates and times end dates and times and codes describin
28. OT WisDOT in conjunction with Highway Commissioners from eight Wisconsin counties embarked upon a 5 year research and implementation program for Winter Maintenance Concept Vehicles In the first two years concept vehicles were deployed in the Counties of Columbia Florence Manitowoc Barron Kenosha Trempealeau Portage and Taylor These vehicles contain numerous on board sensing devices including AVL GPS material spreader sensors blade status sensors vehicle speed sensors and air and pavement temperature sensors Data from these sensors are not only recorded in message sets on board the vehicles they are also telemetered in real time to base stations in County Highway Offices for monitoring and operational purposes Data of this nature in kind and in quantity have never before been available to transportation agencies in Wisconsin All data coming from these vehicles are spatially and temporally referenced and are therefore ideal for management and analysis with geographic information systems GIS technology During years 1 4 of the Winter Maintenance Concept Vehicle Program WisDOT contracted with the University of Wisconsin Madison to 1 conduct research on the nature of the data its quality and its potential applications and 2 develop and test technology that exploits the data to support decision making at local and state levels 2 Summary of Years 1 4 Research Research during Year 1 included Vonderohe et al 2000 1
29. Office Due to hardware problems Kenosha County has never collected any vehicle data files Moreover near the conclusion of this project all data collection equipment had been removed from Kenosha s concept vehicle and there were no immediate prospects for having it re installed Consequently it was decided to postpone installation of Wiscplow at Kenosha County until a later date The network shapefile developed by the Research Team was delivered to the Kenosha County Land Information Office The installation CD for Wiscplow includes a wizard that guides installers through the process That CD is included along with the master copy of this report 9 Options for Maintenance of Wiscplow and Its Data There are a number of options for future maintenance of Wiscplow and the required databases There are technical requirements and institutional needs Full scale implementation of Wiscplow across participating counties is expected to be incremental as more instrumented vehicles become available over time 9 1 Data Requirements and Options 9 1 1 Roadway Spatial Databases Wiscplow requires roadway network shapefiles with source scales of 1 24 000 or larger These shapefiles must include patrol sections They must also be attributed with roadway functional class Such spatial databases were developed for Columbia Kenosha and Portage counties but to remain useful they must be maintained over time to keep them current For example modifica
30. a7 Recordset Fields LiquidAgent Value lt gt none And Data6 Recordset Fields DryAgent Value lt gt None And Data Recordset Fields DryAgent Value lt gt none Then Delete record s from DBGrid1 If DBGridl SelBookmarks count lt gt 0 Then For n 0 To DBGridl SelBookmarks count 1 Data Recordset Bookmark DBGridl SelBookmarks n Data6 Recordset Delete Next n 35 DBGrid1 Refresh Data6 Refresh End If Delete record s from DBGrid2 If DBGrid2 SelBookmarks count lt gt 0 Then For n 0 To DBGrid2 SelBookmarks count 1 Data7 Recordset Bookmark DBGrid2 SelBookmarks n Data7 Recordset Delete Next n DBGrid2 Refresh Data7 Refresh End If Display message that record s were deleted MsgBox The selected record s have been deleted vbInformation WiscPlow End If Message to notify none value in Dryagent and Liquid agent is default value If Data7 Recordset Fields LiquidAgent Value None Or Data7 Recordset Fields LiquidAgent Value none Or Data Recordset Fields DryAgent Value none Or Data6 Recordset Fields DryAgent Value None Then MsgBox None is default value vbExclamation Wiscplow Data7 Refresh Data6 Refresh End If End If End Sub Figure 42 continued Source Code Related to Material Data Entry Tab Comments in Bold Type Face 3 Word for access to the users manual 7 1 2 External Technical Documentation Wiscplow s external technical document
31. able Editing button Equipment Configuration TruckUnit TruckClass FrontPlowUnit FrontPlowClass LeftwingUnit LeftWingClass Rig gUnit RightWingClass_ 1688 18 3161 633 561 65 3188 3260 3261 3263 3266 3278 3289 Enter New Equipment Delete Selected Configuration Equipment Configuration Equipment Cost AttachmentClassID EquipType Description CostPerHour EffectiveStartDate EffectiveEndD ate 632 FrontPlow V shape 20 1 1 2005 12 31 2005 FrontPlow rsible minimum 23 00 1 1 2005 12 31 2005 FrontPlow One way 1 1 2005 12 31 2005 Wing minimum 23 000 1 1 2005 12 31 2005 Enter New Equipment Cost Scraper _ er body blade single 655 1 1 2005 12 31 2015 E _ Cil 53 1412200 12312201 Scraper body blade double 1 1 2005 12 31 2005 Spreader controlled wi 1 1 2005 12 31 2005 Delete Selected Equipment Spreader material to power 1 1 2005 12 31 2005 Cost m _ _ x _ Spreader eeding material to pc 1 1 2005 12 31 2005 Spreader mounted 1 1 2005 12 31 2005 Spreader mounted computer c 1 1 2005 12 31 2005 Spreader yr feeding material to 1 1 2005 12 31 2005 Spreader ader includes prewe 1 1 2005 12 31 2005 Figure 28 Equipment Data Entry Tab 6
32. ach time it 1s called it takes one of many tables beginning with Att e g AttSpreader as its argument MMReportCost Populates MMReportCost table using MMQuery and AllWinterEvents tables MMReportAvgProd Populates the MMReportA vgProd table using the MMQtery table MMReportPercentUse Populates the MMPUQuery table using MMQuery i Pucontinue Populates MMPercentUse amp MMPercentUseState using the other MM tables 13 DatafilesUpdate Updates the Datafiles table in the current database to include the datafile just processed and the information associated with it 49 2 GRAPHICAL REPRESENTATION OF PROCEDURE Buisn jejoonuy pue squiopJg Buisn guipusiddo RjPusug pue squu pJg Buisn 13191480 pue JHpUuS pue squugpig Buisn WINS pue squuopuag jutodejepyjes Jud pue squuopig Buisn pue eo nuy pue ayeindog pue upuesydo Buisn andog duo ojejndod OLHOd3sLNM SL HOd3sS LNM PLYOdAYLNM CLYOdaYLNM CLYOdAYLNM LLHOd3M LNM so qe sJnopoegojeJedo ayeyi0ge7 UOISJ9AUOO lOuJlvgdsogvl SM OHHOSYI 9jeuipJ
33. adways such as ramps see Figure 3 At complex interchanges such as in Figure 4 multiple ambiguities can arise The problem of resolving these ambiguities is known as map matching and has been the subject of research by a number of investigators for more than a decade Blazquez and Vonderohe 2005 Patrol Section 2 gt Direction of Vehicle Travel Data Point Assigned to Incorrect Patrol Section a Patrol Section 1 hd Figure 2 Incorrectly Assigned Data Point at Intersection p gt Direction of Vehicle Travel e Patrol Section 1 f 7 Data Points Assigned to Incorrect Patrol Section Patrol Section 2 Figure 3 Incorrectly Assigned Data Points at Ramp 4 Figure 4 Multiple Ambiguities at a Complex Interchange Many performance measures computed by Wiscplow are associated with patrol sections Clearly if a vehicle data point is snapped to an incorrect patrol section associated performance measures will be erroneous In addition calculation of some performance measures requires determination of traveled distances A calculated traveled distance is grossly in error if one of a pair of consecutive vehicle data points is snapped to an incorrect roadway 4 1 Map Matching Algorithm The map matching problem was addressed by one of the Research Team members Blazquez Her algorithm examines strings of consecutive vehicle data points and uses recorded vehicle speeds computed trave
34. ages in Columbia County s case by seniority are typed into appropriate boxes The Labor Rate Setup screen is tailored to each county s labor contract m Laker Infomation Colonia Labor Kate Effective Dates Start Dara Emi Dane Jan 2005 sun 2005 oes Hourly Wages E Tu Firs mantis 20 Asa 24 mors After B manths 55 months Aftor 12 months 20 180 months Figure 27 Labor Rate Setup Screen for Columbia County 24 6 1 5 Equipment Data Entry Tab The Equipment Data Entry Tab appears in Figure 28 Tables of data on equipment configurations and costs are displayed These tables contain data on how vehicles are configured by equipment class and costs per hour for equipment classes respectively The costs per hour table includes effective start and end dates because these costs vary over time Existing data can be modified by deleting full records and entering new data Pushing the Enter New Data button for either configuration or cost causes an appropriate data entry screen to appear Si WiscPlow _ Information and Password Material Data Entry REPORT Labor Labor Data Entry START Process Datafile s Equipment Data Entry REPORT Material REPORT Equipment Map Display To make changes in the tables you must enter the password in the Information and Password tab and press the Allow T
35. amination of Local Roads Databases being developed by WisDOT with an eye towards their ability to support GIS based winter maintenance concept vehicle applications Research during Year 3 included Vonderohe et al 2002 1 Development of a methodology for building patrol section route systems in roadway spatial databases and application of that methodology to state and US highway representations in the Columbia County database 2 Extension and refinement of the software to include a set of GIS based applications for winter maintenance operations based upon the computation of performance measures from raw data coming from the vehicles The software system was dubbed Wiscplow 3 Field testing of Wiscplow in two Wisconsin counties 4 Development of management decision tools for interpretation of computed performance measures and their interrelationships 5 Analysis of the sensitivity of selected performance measures for Winter Storm Reports to temporal resolution and spatial noise in raw data coming from the vehicles 6 Description of editing functions necessary for management of spatial and aspatial data supporting Wiscplow Research during Year 4 included Vonderohe et al 2003 1 Identification and classification of spatial ambiguities between reported vehicle coordinates and digital roadway centerline representations Such ambiguities can lead to incorrect determination of the roadway upon which a vehicle is tra
36. and directed links is necessary for pathfinding between vehicle data points calculation of distances along paths and representation of allowable turns and directions of travel Network shapefiles must be attributed with roadway functional class because US and state highways must be distinguished from county and local roads They must also be attributed with roadway names to support map based queries Network shapefiles must also include patrol sections with appropriate identifiers In addition earlier analysis had indicated that the positional accuracy requirements of Wiscplow necessitated a source scale for the cartography of 1 24 000 or larger No participating county maintained a topologically structured network shapefile of its roadways although all of them had some digital roadway representation The counties roadway Spatial databases were of varying accuracy and currency Subsequently the Research Team undertook development of network shapefiles building upon data available from the counties including existing roadway spatial databases and digital orthophotos Initially this work was expected to entail development of network shapefiles for interstate US and state highways only These are the roadways upon which the counties perform reimbursable winter maintenance for WisDOT Five counties were initially selected for development Columbia Kenosha Manitowoc Portage and Trempealeau Shortly after work began Manitowoc and Trempealeau Cou
37. ase and the topology necessary to support Wiscplow A spatial join function was used to initiate this Spatial joins merge attributes from two spatial databases but only for features that coincide Since there were non coincident features in the two databases due to the updates attributes of the updated features were added manually by interpretation of a hard copy roadway map and overlay of the network shapefile with a county wide digital orthophoto During this process the entire network shapefile was visually examined for registration with the orthophoto Testing of the topology or network integrity of the network shapefile with a pathfinding engine and with Wiscplow revealed a few topological inconsistencies that were subsequently corrected The cartography of the final Columbia County network shapefile is shown in Figure 11 There are two layers 1 Roadways and 2 Patrol Sections In Figure 11 patrol sections are highlighted in bold red Metadata for the network shapefile including descriptions of all attributes were prepared according to Federal Geographic Data Committee Content Standards for Geospatial Metadata These metadata are embedded within the network shapefile database see Figure 12 The complete Columbia County network shapefile was installed along with Wiscplow at 12 Hey ESN x YE AD aR un Ue LLLA E
38. ated an appropriate buffer size of 45 to 50 ft The source scale of the Portage County roadway centerline spatial database is 1 12 000 while that of Columbia County is 1 4 800 The map matching algorithm is very effective However it is based upon certain assumptions which if violated will cause it to fail Spatial ambiguities due to the vehicle moving against allowable directions of travel or making disallowed turns cannot be resolved It is also assumed that all actual roadways are represented in the roadway spatial database and that vehicles must travel on roadways not for example parking lots If the roadway spatial database is incomplete or incorrect or if the vehicle travels other than on roadways ambiguities cannot be resolved and the associated collected vehicle data are ignored by Wiscplow See Blazquez 2005 for an in depth analysis of the algorithm s controlling parameters i e buffer size speed comparison tolerance and number of consecutive data points to be examined 10 Measured Correct snap Incorrect snap Figure 9 All Eleven Spatial Ambiguities at This Interchange Were Resolved 5 Development and Testing of Roadway Spatial Databases The map matching algorithm requires roadway spatial databases to be structured as network shapefiles that contain not only cartographic lines but also have connected topological structures with no gaps or overlaps Such an underlying network structure consisting of nodes
39. ation consists of 1 A procedure map a graphical representation of the procedure map and the Wiscplow folder structure These appear in Appendix A of this report 2 A data dictionary that describes each item in each table in each Wiscplow database The data dictionary runs to 42 pages and is included on a CD along with the master copy of this report 7 2 User Documentation The users manual for Wiscplow runs to 109 pages and is not included in the text of this report It exists as a Microsoft Word document embedded within Wisplow and accessible through the system s Help button The users manual is also included on a CD that accompanies the master copy of this report 36 The table of contents for the Wiscplow users manual appears in Figure 43 full Wiscplow installation includes a tutorial with associated vehicle data files databases and network and patrol section shapefiles A walk through of the tutorial is included in Appendix A of the users manual Wiscplow Users Manual Table of Contents 1 Introduction 4 2 Installation and Setup 5 2 1 System Requirements 5 2 2 Installing Wiscplow 5 2 3 Initial Setup 7 3 12 3 1 Information Password 12 3 2 Winter Events Entry 13 3 2 1 New Storm Information 14 3 2 2 New Incident Information 15 3 2 3 New Anti Ice Information 16 3 3 Equipment Data Entry 17 3 3 1 New Equipment Costs 18 3 3 2 New Equipment Configuration 19 3
40. cation of this report 20 Security Classification of this page 2 No Of Pages 22 Price Unclassified Unclassified 0 Form DOT F 1700 7 8 72 Reproduction of form and completed page is authorized 1V Table of Contents Section Disclaimer Acknowledgements Exhibit B Table of Contents Executive Summary 1 Background 2 Summary of Years 1 4 Research 3 Objectives of Year 5 Research 4 The Spatial Ambiguity Problem and Its Solution 4 1 Map Matching Algorithm 4 2 The Algorithm Applied to Actual Data 5 Development and Testing of Roadway Spatial Databases 5 1 Columbia County 5 2 Kenosha County 5 3 Portage County 6 Completion and Testing of Wiscplow Functionality 6 1 Wiscplow Functionality 6 1 1 Information and Password Tab 6 1 2 Winter Events Entry Tab 6 1 3 Material Data Entry Tab 6 1 4 Labor Data Entry Tab 6 1 5 Equipment Data Entry Tab 6 1 6 START Process Datafiles Tab 6 1 7 Map Display Tab 11 12 14 16 19 19 21 21 22 23 25 25 2f 6 1 8 REPORT Material 6 1 9 REPORT Equipment Tab 6 1 10 REPORT Labor Tab 6 2 Testing and Validation of Wiscplow 7 Technical and User Documentation 7 1 Technical Documentation 7 1 1 Internal Technical Documentation 7 2 2 External Technical Documentation 7 2 User Documentation 8 Installation and Training at County Highway Departments 9 Options for Maintenance of Wiscplow and Its Data 9 1 Data Requirements and Options 9
41. ce liquid Check all Inventory REPORT Labor REPORT Equipment Map Display Labor Data Entry Hourly average material application rate for each patrol section Tons of salt Pounds lane mile of salt Cubic yards of sand Pounds lane mile of sand Gals of prewetting liquid ton of salt Gals of prewetting liquid ton of sand Check all Quantity of material used for each event and patrol section Gals of prewetting liquid Gals of anti ice liquid Quantity of material used for all events and each patrol section Tons of salt Cubic yards of sand Gals of liquid material prewetting and anti ice Check all Gals lane mile of anti ice liquid Select all to include Number of blasts for each operator and event 75 Pavement Temperature Hourly average pavement temperature for each patrol section Jan 1999 ha __ 12 hs Date Information Analytical Decision Tools Hourly avg salt application rate and pavement temperature by vehicle ID over a day midnight to midnight Seasonal cumulative salt use by patrol section Average pavement temperature salt and sand rates by selected storm event for all patrol sections Pre wet salt application rate quideline compliance by selected storm event and all patrol sections for initial application Pre wet salt application rate guideline compliance by selected storm event and all patrol sections for repeat applicati
42. countywide orthophoto The network integrity or topological structure of the network shapefile was checked with a pathfinding engine and with Wiscplow The cartography of the final Kenosha County network shapefile is shown in Figure 15 There are two layers 1 Roadways and 2 Patrol Sections In Figure 15 patrol sections are highlighted in bold red Metadata for the network shapefile including descriptions of all attributes were prepared according to Federal Geographic Data Committee Content Standards for Geospatial Metadata These metadata are embedded within the network shapefile database The complete Kenosha County network shapefile was delivered to the Kenosha County Land Information Office The network shapefile along with its metadata is included on a CD attached to the master copy of this report 5 3 Portage County The original Portage County roadway spatial database was digitized from United States Geological Survey Digital Orthophoto Quarter Quads at a source scale of 1 12 000 Since initial data capture alignment changes had taken place on one of the major highways and the County Land Information Office had recently completed the corresponding updates of the spatial database The Research Team was provided with a copy of the updated database and with a countywide high resolution digital orthophoto The County Highway Department provided a hardcopy map of patrol sections Portage County performs winter maintenance on a stretch
43. ction coordinates of the roadway network shapefile The map matching algorithm then associates the vehicle data points with roadways and patrol sections These associations are stored in the database A point shapefile of the vehicle data points is also produced The vehicle data points can then be viewed and queried at the Map Display Tab 6 1 7 Map Display Tab At the Map Display Tab users can display the roadway network shapefile the patrol section shapefile and any number of vehicle data shapefiles Each shapefile is considered to be a ayer that can be added turned on and off or deleted from the display Figure 32 shows the Map Display Tab with Columbia County s roadway network and patrol sections There are a set of function buttons just above and to the left of the map These include pan and zoom add and remove layers print measure distance and identify functions The identify function can be used to query any feature in the map display for its underlying attributes For example in Figure 33 the user has added a vehicle data shapefile zoomed in identified a particular vehicle data point and obtained the names of the roadway and patrol section upon which the vehicle was traveling vehicle speed operator name environmental data date and time equipment status and material spreading rates Multiple vehicle data files can be viewed simultaneously showing relationships between vehicle tracks and activities 6 1 8 REPORT Mater
44. data roadway spatial databases and the Wiscplow system 4 The Spatial Ambiguity Problem and Its Solution Automated vehicle tracking involves determination of the roadway upon which a vehicle is traveling by registration of two dimensional vehicle coordinate locations to cartographic representations of roadway centerlines i e digital roadway maps Problems sometimes arise due to errors in 1 the vehicle coordinates collected by differential GPS and 2 the cartographic representations typically having a source scale that limits their positional quality Because of these errors vehicle coordinates do not exactly register with roadway centerlines see Figure 1 This phenomenon can lead to ambiguities in that by examination of single pairs of coordinates the vehicle can appear to be on the incorrect roadway p Direction of Vehicle Travel e Vehicle Data Point Patrol Section 1 Patrol Section 2 e Figure 1 Vehicle Data Points Do Not Register with Cartoaraphic Centerlines Perhaps the simplest example occurs at at grade intersections see Figure 2 Spatial analysis software typically associates a two dimensional data point with the centerline that is nearest by perpendicular measure The mathematical process 15 referred to as Snapping In Figure 2 the vehicle appears to be on the wrong roadway for the data point nearest the intersection The problem is exacerbated at diverging or converging ro
45. dimensional coordinates that must be associated with roadways and patrol sections by overlaying them with digital maps Because there are errors in both the coordinates and the maps vehicles can be computed to be on incorrect roadways This results in incorrect values for performance measures that are often computed by patrol section and based upon traveled distance The map matching problem was successfully addressed during year 5 through development testing and implementation of a new decision rule algorithm Wiscplow is intended for use by analysts and other technical staff in a post processing environment It is a standalone application not an enterprise solution It has no real time monitoring capability and is not portable as it requires access to databases that are typically only available on client server networks Its reporting functions are categorized by labor equipment materials and map display Internal and external technical documentation were developed for Wiscplow FGDC standard metadata were developed for all spatial databases Wiscplow s source code all documentation and all spatial databases are included on a CD Vil attached to the master copy of this report A separate installation CD is also included with the master copy of this report On going maintenance of both Wiscplow and the necessary spatial databases raises a number of institutional issues and options Wiscplow requires roadway network shapefiles with source
46. e CombineBrdcrmbEquip Combines data from BrdCrmbsData table and EquipUsage table to create BrdCrmbs table in each datafile database DoProjection Converts GPS coordinates to county coordinates and then populates the X and Y columns in BrdCrmbs table in database It calls the following procedure in order to do so a projection This procedure actually converts the GPS coordinates into the appropriate county map coordinates using the projection parameters stored in the CoordinateConversion table of the Info database addpnteventshp Adds the breadcrumbs from BrdCrmbs table into a new map layer totalsecfeet Populates the SectionFeet table in current database showing total distance of each patrol section in county poppatroldist Populates the PatrolDistance table in current database showing distances traversed in each patrol section for processed data addmeasure This procedure is the map matching algorithm used by WiscPlow to determine what road a given GPS point is on It contains calls to the following procedures and sometimes these procedures call one another to perform the map matching algorithm Ultimately this procedure determines what road the given GPS points are on and uses that data to populate the patrol section roadway class and distance fields in BrdCrmbs table a tryltolast b correctWrongSnap C passingArrays d getpath e tryaltprev 48 eS CBS 8 11
47. echnical and User Documentation Detailed technical documentation is necessary not only to describe the internal workings of software to interested readers but also to ensure that the system can be maintained in the future by other than the original developers Detailed user documentation is necessary to ensure utility of software in the business place 7 1 Technical Documentation Wiscplow s technical documentation has an internal component and an external component 7 1 1 Internal Technical Documentation Wiscplow s internal technical documentation consists of extensive comments inserted within the Visual Basic source code Figure 41 shows a small portion of Wiscplow s source code with comments highlighted in bold type face Figure 42 shows how comments and source code relate to Wiscplow s user interface The source code consists of 23 rich text formatted files that are included on a CD along with the master copy of this report Wiscplow s spatial analytical functionality is based upon Environmental Systems Research Institute s ESRI Map Objects The installed version of Wiscplow consists of a large executable module and a number of dynamic linked libraries text files and bitmaps A full installation CD for Wiscplow is included with the master copy of this report Wiscplow requires access to the following Microsoft products 1 Excel for decision management tools 2 Access for database management 33 1
48. ed Agents Command28 Figure 42 Source Code Related to Material Data Entry Tab 34 Command23 Enter New Product button that takes into effect changes to the dry and liquid agent Products tab Private Sub Command23 Click Check the Password by calling function If CheckPassword False Then Exit Sub End If Bring up material entry form Form5 Show Refresh Grids after done entering records DBGrid1 Refresh DBGrid2 Refresh End Sub Command28 Delete Selected Agent s Delete selected agent in Products tab Private Sub Command28_Click Check the Password by calling function If CheckPassword False Then Exit Sub End If Dim response As Integer Dim count As Integer count DBGrid1 SelBookmarks count DBGrid2 SelBookmarks count Show message if no record is selected If count 2 0 Then response MsgBox You must first select a record to delete vbExclamation WiscPlow Exit Sub End If Ask the user if they are sure they want to delete record s If count 0 Then response MsgBox Are you sure you would like to delete the selected record s vbExclamation vbYesNo WiscPlow End If If they answer no then exit procedure If response 7 Then Exit Sub End If If they answer yes delete records and display message that record s were deleted If response 6 Then If none value is not selected delete the selected value If Data7 Recordset Fields LiquidAgent Value lt gt None And Dat
49. eed comparison yields a large disparity then the algorithm follows the flow chart illustrated in Figure 6 The algorithm determines if feasible routes exist between the preceding and subsequent points bounding the data points of concern If a feasible route is found using either of the bounding points the intermediate data points not used to obtain the shortest path are snapped to the roadways along the determined feasible path yes yes yes yes Alternative Road Search and Feasibility Path Check Algorithm yes yes End End End LEGEND Is the path between snapped points i and j feasible Start algorithm End algorithm with spatial Is there an alternative roadway ambiguity not solved centerline within the buffer End algorithm with spatial around point k ambiguity solved Figure 6 Map Matching Algorithm For example if there is no feasible path between S2 and S3 then the algorithm looks ahead by Snapping point 4 to the nearest roadway centerline within its buffer and determines if the shortest path between snapped points S3 and S4 is feasible If the tested path is not feasible the algorithm snaps point 3 to the next nearest roadway centerline within its buffer obtaining point alt3 The algorithm illustrated in Figure 7 is i
50. g events Modifications to existing data require deletion of appropriate records and entry of new data Data in the tables can be selected for deletion but they cannot be edited by typing on a keyboard 3 WiscPlow START Process Datafile s Equipment Data Entry REPORT Material REPORT Equipment Map Display Information and Password Winter Events Entry Material Data Entry REPORT Labor Labor Data Entry To make changes in the tables you must enter the password in the Information and Password tab and press the Allow Table Editing button Storm Information Storm Events TT NR RUN DEZ ea Enter New Storm S006 1 0000 1 12 2005 4 20 00 PM 1 13 2005 50018 50019 50020 955 004M 1 5 2005 12 00 00 1 28 2005 10 40 00 AM 1 6 2005 3 10 00 4M 1 26 2005 Delete Selected Storm 12 00 00 AM 1 13 2005 5 00 00 1713 2005 Enter New Incident Delete Selected Incident Anti Ice Information 12 00 00 AM 8 12 2005 1 00 00 8 12 2005 Enter New Anti Ice Delete Selected Anti Ice Figure 21 Winter Events Entry Tab When new data are entered by pushing for example the Enter New Storm button a data entry screen appears see Figure 22 Here dates are picked from calendars times are picked from scrolled lists and codes are chosen from descript
51. have many large data files to process These can be placed ina long list that can then be processed in batch mode overnight Data files can be deleted from the process list It is also possible to remove from the existing database other data files that have been previously processed Datafiles Process Details iO Eve StatOaste End Date Add a file to process list Select Event ID Select Operator Apply Details Select Vehicle ID Ory Agent gt Liquid Agent Remove File from Process List Select to remove C Clear Remove file from process list Close WiscPlow Remove File from Database Select to remove Clear Remove file from database Figure 29 START Process Datafiles Tab 6 5 0 104A DCS710A M10 ResDrvC ResDrv AgDrvC AgDrv CommDrvC CommDrv Bridge Culvert StopSgn OtrSgn NoPass Rd Rpr ShidRpr Asph Seal Estimate 12 89 M13 8 17 41 0 00000 0 00000 0 0 A N 1 8 17 41 0 0 0 00000 0 00000 0 0 M19 8 17 41 mt 52 52 0 off off off 0 off off off off off off M22 8 17 41 s81 off off off off off 0 off off off off off 5 5 0 2 M23 8 17 41 ps down no no down down down down down down down no no no no 0 00000 0 00000 0 0 M13 8 17 41 0 00000 0 00000 0 0 A N M6 8 17 41 c1 2500 1200 20 260 20 20 3350 143 136 M7 8 17 41 c2 200 200 600 800 1000 1200 1400 1600 1800 2000 12 M8 8 17 41 c3 100 200 300 400 500 600 700 800 900 1000 12
52. ial Tab The REPORT Material Tab appears in Figure 34 At this tab users can request reports on up to 19 performance measures ten for application rates seven for inventory and one each for 21 Wter Events Enty _ Material Data Entry REPORT Labor Labor Data Entry START Process Datafile s Equipment Data Entry REPORT Material REPORT Equipment Map Display 5 PATROL SECTION SHP v COLUMBIA ROAD jt 1 a Ne pe toon na 510254 04 294962 08 4 39 PM Figure 32 Map Display Tab with Function Buttons and Columbia County Roadway Network and Patrol Sections sand Wither Everts Eniry Mate ial Cala Erdry START Proceas adr xj Loco 52990152 400955 95 e S ce axe EX Ex e 2 toured X520 700 83 402408 48 Figure 33 Zoomed Map Display with Vehicle Data Points and Result of a Query Using the Identify Function 28 3 WiscPlow Information and Password Wrter Events Enty Material Data Entry START Process Datafile s Equipment Data Entry Application Rates REPORT Material Average material application rate for each operator and event Pounds lane mile of salt Pounds lane mile of sand Gals of prewetting liquid ton of salt Gals of prewetting liquid ton of sand Gals lane mile of anti i
53. in files are located as well as the buffer tolerance to be used When county parameters are changed this file is rewritten password ini file This file contains the current password used in WiscPlow When the password is changed this file is Bitmaps Folder Contains images used in WiscPlow Help Folder Contains user manual for WiscPlow which includes a tutorial The manual 15 titled UserManualVersion3 doc Network Shapefile Folders One folder will exist for each roadway network e g columbiaroads nws In addition a temporary network xxxxtmp nws will also exist These network folders should never be modified County Shapefiles Two shapefiles exist for each county to be used with WiscPlow e g columbiaroads and columbiapatrolsections Each shapefile consists of 6 7 files with a variety of extensions shp dbf shx sbx Sbn prj These files should never be modified or deleted
54. ing roadway is resolved and the correct roadway for point 2 is Interstate 39 Point S1 is then obtained by snapping point 1 to the Interstate 39 centerline Table 1 Speed Comparison for Determining Feasibility of Shortest Paths Data Points Shortest Path Calculated Average Feasible Path Distance Speed mi hr Recorded Speed feet mi hr NO 50 52 26 31 YES 52 gt 53 53 54 35 YES 50 alt2 31 YES alt2 gt 53 33 YES The algorithm resolves spatial ambiguities at converging roads divided highways and intersections in a similar manner Figure 9 presents results for vehicle data points at an interchange in Columbia County Wisconsin Circles indicate original measured vehicle data points Incorrectly snapped locations for these points are depicted as asterisks and correctly Snapped locations determined after the algorithm was executed are shown as rectangles Bold arrows represent shortest feasible paths between correctly snapped points Eleven of 28 vehicle data points were initially snapped to incorrect roadways All of these spatial mismatches were resolved by the map matching algorithm The appropriate buffer size for the algorithm depends on the quality and geometry of the spatial data i e vehicle data points and roadway centerline spatial databases Testing of Columbia County data indicated that buffer sizes from 30 to 45 feet correctly snapped all vehicle data points Separate testing of Portage County Wisconsin data indic
55. inter Maintenance Concept Vehicle Program WisDOT contracted with the University of Wisconsin Madison to 1 conduct research on the nature of the data its quality and its potential applications and 2 develop and test technology that exploits the data to support decision making at local and state levels The research reported herein was funded during year 5 by WisDOT through the Midwest Regional University Transportation Center Year 5 was the concluding year of the overall project Year 5 work included refinement and comprehensive documentation of a GIS application dubbed Wiscplow The application computes performance measures and produces decision management tools by post processing data recorded on board the vehicles and integrating it with spatial databases depicted roadways patrol sections and their attributes The necessary spatial databases did not exist at the county level so year 5 work included development and testing of them Ultimately Wiscplow its documentation including tutorials and all necessary spatial databases were installed at the Columbia and Portage County Highway Departments and at WisDOT headquarters On site training was provided to staff Successful development of Wiscplow required solutions for a number of technical problems not the least of which was the map matching problem arising from spatial ambiguities in the data caused by positional and representational errors Vehicle locations are captured as two
56. ion FGDC STD 001 1998 Time convention used in this document local time Metadata profiles defining additonal information ESRI Metadata Profile htto www esri com metadata esriprof80 html Figure 12 A Page of Metadata for the Columbia County Network Shapefile the Columbia County Highway Department Office see Section 8 The network shapefile along with its metadata are included on a CD attached to the master copy of this report 5 2 Kenosha County The original Kenosha County roadway centerline spatial databases had been compiled from large scale cadastral maps by digitizing centerlines For a number of years the Kenosha County Land Information Office had been updating the spatial database using survey plans and 14 coordinate geometry COGO software That office estimated the accuracy of the roadway centerlines to be 5 10 feet The roadway centerlines were provided in five shapefiles one each for interstate highways other US highways state trunk highways county trunk highways and minor roads Roadway names were provided as annotation not as attributes The network shapefile was generated by merging and editing the five separate shapefiles manually entering roadway names from the annotation and other necessary attributes for Wiscplow and manually entering topological structures for allowable turns and travel directions Editing consisted of 1 adjusting or re digitizing centerlines that did not register with the cou
57. ive pick lists Distinctions are made between routine and forecasted anti ice events 21 5 5 Information Storm Information Stat Date and Time SEE 1200 00 ess Storm Code 1 Jan 2005 Jan 2005 Dry Snow Mon Tue Wed Thu Fri Sat Sun Storm Code 2 None __ 5 SER Storm Code 3 17 __ 18 None 24 __ 25 J gt Storm Code 4 Mone End Date and Time zu m Amount of snow in inches Jan 2005 Jan 2005 4 3 NOTE Entering a start and end date time is required Entering storm codes and amount of snow is optional Figure 22 Storm Data Entry Screen 6 1 3 Material Data Entry Tab The Material Data Entry Tab appears in Figure 23 The tab displays data for dry agents and liquid agents Existing data can be modified by deletion of existing and entry of new full data records Pushing the Enter New Product button causes a data entry screen to appear where new agents and associated codes can be chosen from pick lists Gl WiterP live SLANT Process 5 Display neo Paseo Winker Everts Entry Plater ial Dale Labor Dala Products make changes in the tables you must enter the password in the Information and Password tab and press ihe Allow Table Editing button Dry Agent Liquid Agent Drphgent
58. l arrangements concerning maintenance of Wiscplow source code might be the most difficult problem to address while ensuring longevity of the products of this research Five options were described Perhaps the most feasible is a low level retainer for a member or members of the Research Team to take custodianship of the source code This report describes activities and accomplishments during the concluding fifth year of a significant research and development relationship among WisDOT a number of county highway departments and land information offices and the Midwest Regional University Transportation Center and the Department of Civil and Environmental Engineering at the University of Wisconsin Madison seems valid to assert that all parties have benefited WisDOT and the 40 counties received advanced technology and training in analysis and management of winter maintenance concept vehicle data They participated in development of an understanding of these data and in development of tools for exploitation of the data in decision making The University of Wisconsin Madison saw this research form the basis for seven M S degrees and two PhD degrees The research produced six published papers and nine formal presentations at national and statewide conferences and workshops List of References Blazquez C 2005 Decision Rule Algorithm for Map Matching in Transportation PhD Dissertation University of Wisconsin Madison Blazq
59. led distances and allowable turns and directions of travel for the roadways Blazquez 2005 The algorithm selects all roadways within a buffer of specified distance around a vehicle data point and snaps the point to the closest roadway by determining the minimum perpendicular distance from the data point to each roadway Consider a set of five consecutive vehicle data points 0 4 As shown in Figure 5 points 1 and 2 are snapped to ramp 2 because it is the closest roadway contained within the buffers around the points Subsequently the shortest path indicated with a bold arrow is computed between the two snapped data points 51 and S2 using network topology and turn restrictions The speed of travel between 51 and 52 is computed from the length of the shortest path and the difference in time stamps for the points The computed speed is then compared to the average of the vehicle speeds recorded at the data points If the computed speed is within a specified tolerance of the average recorded speed then the obtained shortest path is viable and the snapped locations 51 and S2 are accepted as correct Figure 5 Two Vehicle Data Points 1 and 2 and their Correctly Snapped Locations S1 and S2 The algorithm advances to the next data point 3 snaps it to the nearest roadway centerline within its buffer and calculates the shortest path between S2 and the newly snapped data point 53 If the path between 52 and 53 is not feasible because the sp
60. les in the current database and a few tables in the info database by making calls to the following procedures 1 CombineAllEvents Populates the AllWinterEvents table in the current database using Storm Incident and Antiicing tables 2 CombineAllIndexes Populates the AllEventIndexes table in the info database using StormIndex AntilcingIndex and IncidentIndex tables START Process Datafile s Tab Most of WiscPlow s most complicated code 15 contained within procedures called from the user interface in the START Process Datafile s tab contained in the main form Each bold header below contains the name of the initial procedure called in Visual Basic as well as the caption contained on the button in the WiscPlow user interface AddNewFileButton Click New File To Process List Button This procedure runs when the user clicks the New file to Process List button The procedure clears all the variables used to read raw datafiles before calling the following procedure 1 Readfilename Goes through many checks of the data then decodes the new datafile It checks if a database already exists for the specified file and also checks to make sure that the file contains information meaningful to WiscPlow It contains calls to the following procedures a DecodeMessages This procedure 15 called if all of the above checks of the data are passed It extracts information from the raw datafile and store
61. m included internal and external technical documentation for the software Final development of the application required refinement of performance measures decision management tools and the user interface A number of previously unsolved technical problems also needed to be addressed These included the map matching problem in which moving vehicles must be tracked by roadway and patrol section by registering strings of two dimensional vehicle coordinates to digital maps spatial databases The problem is exacerbated by errors in the coordinates and in the maps A decision rule algorithm was developed and tested against a number of available data sets The algorithm resolves nearly all ambiguities encountered in the data This algorithm is embedded in the installed version of the software Testing revealed the limit 1 24 000 on source scale of the spatial databases needed to support the application Future maintenance of both the software and the data raise technical and institutional issues that were identified and described by the research team Recommendations concerning these issues are included in the final sections of this report 17 Key Words 18 Distribution Statement Winter Operations Concept Vehicle AVL GPS Application GIS Application Performance Measures Map Matching No restrictions This report is available through the Transportation Research Information Services of the National Transportation Library 19 Security Classifi
62. nd maintenance at the local level Those who need the data become responsible for it Impediments include lack of skills in spatial database maintenance at county highway departments and lack of resources at county land information offices Options for maintenance of Wiscplow s source code include 1 WisDOT s Bureau of Automation Services BAS has the skills necessary for maintenance of Wiscplow However BAS has a long standing policy of not maintaining software developed by other parties The source code could be maintained at the county level by land information offices or IT units However they would have to agree to do this Furthermore each installation of Wiscplow would evolve differently over time as each would be maintained separately Eventually data models and therefore databases would become incompatible across installations UW Madison could potentially maintain the source code Some mechanism would have to be developed for continuity beyond the tenure of current Research Team members Some level of continued funding would have to be developed viii 4 Consultants could be contracted to maintain the source code A continuing funding source would have to be identified 5 he source code could be maintained on a retainer basis by a member or members of the Research Team A continuing low level funding source would have to be identified 1X 1 Background During the winter of 1999 2000 the Wisconsin D
63. nitiated if an alternative roadway centerline exists within the buffer This algorithm verifies if a path is feasible between the alternative snapped location for point 3 where k 3 and former and succeeding neighboring snapped points 2 and 4 where k 1 2 and k 1 4 If the shortest paths between these three points are not feasible because the speed comparison fails the algorithm searches for other roadway centerlines within the buffer around point 3 that have not already been used in a feasibility path check Finding a new candidate point 3 is then snapped to it and the feasibility of shortest paths between snapped points 2 3 and 4 k 1 k k 1 is checked again If these paths are feasible then the spatial ambiguity is resolved and the algorithm in Figure 7 terminates End LEGEND Is the path between snapped Start algorithm points i and j feasible End algorithm with spatial ambiguity not solved Do other alternative roadway centerlines exist End algorithm with spatial within the buffer around ambiguity solved point k Figure 7 Alternative Road Search and Feasibility Path Check Algorithm If no alternative roadway centerline exists within the buffer for point 3 the algorithm in Figure 7 stops without resolving the ambiguity using the alternative snapped location for point 3 and exits back to the algorithm presented in Figure 6
64. nties asked to be excused from the project citing concerns over potential future costs for vehicle and spatial database maintenance Given a 11 reduction in the number of participating counties it was decided to expand the scope of network shapefile development to include not only interstate US and state highways but also all local roads within each remaining county Columbia Kenosha and Portage Work for each county entailed 1 data capture and editing 2 data structuring 3 quality control 4 testing with Wiscplow and 5 development of metadata 5 1 Columbia County The original Columbia County roadway centerline spatial database had been compiled photogrammetrically at a source scale of 1 4 800 During 2004 and early 2005 the Columbia County Land Information Office made major updates to the database resulting from on the ground additions and changes to roadways They also inserted patrol sections with identifiers in the updated spatial database An example of updates made by the Columbia County Land Information Office appears in Figure 10 Figure 10 Old Roadway Centerlines Left and Updated Roadway Centerlines Right in Columbia County Spatial Database The Research Team had been working with the original roadway spatial database having appropriate attributes for use with Wiscplow Therefore it was necessary to develop a new network shapefile that had the updated cartography the attributes of the original spatial datab
65. ntywide high resolution orthophoto and 2 making small corrections for topological consistency Figure 13 shows three examples requiring adjustment or re digitizing of roadway centerlines Figure 14 shows two examples requiring editing for topological consistency Figure 13 Roadway Centerlines Yellow Requiring Adjustment or Re Digitizing Figure 14 Topological Problems An Overshoot Left and An Undershoot Right The spatial reference system used by the Kenosha County Land Information Office is the State Plane Coordinate System referenced to the North American Datum of 1927 NAD 1927 Concept vehicle coordinates captured by differential GPS are latitude and longitude referenced to the North American Datum of 1983 NAD 1983 The vehicle coordinates are transformed by Wiscplow from latitude and longitude into state plane map projection coordinates but Wiscplow cannot perform datum transformations Thus the network shapefile was transformed into NAD 1983 HARN State Plane Wisconsin South placing it on the same datum as the vehicle coordinates 15 Patrol sections were embedded in the network shapefile by manually inserting their boundaries based upon a hardcopy map provided by the County Highway Department Attributes for patrol section identifiers were manually entered After editing of the cartography by one of the Research Team members a second Team member visually inspected the entire data set for registration with the
66. on Mon Tue Wed Thu Fri Sat Sun pe 8 1 2 ja Start Date Jan 1999 End Date Oct 2005 26 8 23 ia 9 2 ps ps 7 _ oet 2005 Mon Tue Wed Thu Fri Sat Sun List of Datafiles Figure 34 REPORT Material Tab Columbia blasts and pavement temperature Users may also request up to five analytical decision tools i e charts that show relationships among performance measures over time and space Start date and end date calendars are included for selection of time periods to be covered by the reports and analytical decision tools An example report is shown in Figure 35 and an example analytical decision tool is shown in Figure 36 Reports can be printed or stored as rich text format rtf files Automated Winter Storm Reports Total Salt used for each Patrol Section for each Storm For the Period of 1 9 2001 through 2 19 2001 StormID 50012 50012 50012 50023 50023 50023 50031 5006 5006 5006 PtrlSctnID 106 107 110 106 107 110 107 106 107 110 Figure 35 Example Material Report 20 Total Salt used in Tons 16 79 0 32 0 82 Material Application Rates and Pavement Temperature N e e Ibs lanemile 9 Q lt O
67. ooo asp sonquye uonoes joped Buisn 92uejsip pue so jedeus squuopug squuopig ejeindod Jed seouejsip UBISSV 155 juonoos ayeindog suonoes joned ppy 0 esn eyejndod MOS 1Isld1OulVvddOd 1444945 A dHSJIN3A31NdQCdV NOI LO3f OHd A 1 g s jqe Buisn SJUu9 3J91UlAAJ V pue aqe 150 Asenbuyy polgBbayyoday WA exejndod qosudoAvLJoda3sWiwN SJUS 3J91UIMIIV pue ayeindog 5 v 318V DININ yuy pug 3915 Buisn TIVOWW dsvavlvd ALVINdOd JISVEvVLvd OL 5 dav eyepdr 31vadnsa JIsvavlvd JIsvavlvd ALVAYD anng y
68. points SO and S2 is computed Subsequently the speed comparison shown in Table 1 is performed to determine if this path is feasible In this case the obtained path is feasible since the difference between the average calculated and recorded speeds 26 and 31 mph respectively is within tolerance 20 mph Therefore the current Snapped positions for points 0 and 2 are initially assumed to be correct The algorithm continues by finding the shortest path between the next pair of snapped points 52 and 53 This path is not feasible because if the vehicle was at S2 it would have to exit down the ramp and travel approximately 5 126 feet in 5 seconds at an average speed of 699 mph to reach S3 Thus 52 53 or both were snapped to an incorrect roadway centerline The algorithm now looks ahead to S3 and S4 and determines that the difference between calculated 29 mph and average recorded speeds 35 mph is within tolerance Therefore an alternative roadway centerline is sought within the buffer around point 2 Interstate 39 is found to be the next nearest roadway resulting in point alt2 shown as a triangle in Figure 8 Feasibility is now checked for paths between the preceding SO and alt2 and between alt2 and its successor S3 As indicated in Table 1 both computed shortest paths are feasible Calculated speeds along these paths are within 20 mph of their respective average recorded speeds for the vehicle Therefore the spatial ambiguity at the diverg
69. s 13 Type of Report and Period Covered Wisconsin Department of Transportation Final Report 01 01 05 Hill Farms State Transportation Building 08 31 06 4802 Sheboygan Avenue 14 Sponsoring Agency Code Madison WI 53707 15 Supplementary Notes Project completed for the Midwest Regional University Transportation Center with support from the Wisconsin Department of Transportation 16 Abstract A five year research effort fifth year funding through MRUTC culminated in development implementation and installation of a GIS application for assessing performance of winter highway applications The software accepts data recorded on board winter maintenance vehicles during operations and combines it with spatial data representing roadways and vehicle patrol sections Analysts can then select among a number of performance measures and decision management tools for outputs from the system Outputs are categorized according to labor equipment materials and map displays that indicate vehicle routes and data collected along the way The software full user documentation and necessary spatial databases were installed in two Wisconsin county highway department offices and at Wisconsin DOT headquarters Training was provided to staff The spatial databases were developed and scrutinized for quality by the research team from data provided by the counties FGDC standard metadata were included with the spatial databases Documentation of the full syste
70. s it in arrays b bFileExists sFile As String This procedure is called to determine if a file exists It is used a few times throughout the Readfilename procedure 45 Commandl Click Apply Details Button After the user has entered information in the appropriate combo boxes they push the Apply Details button and this procedure runs calls the following procedure to make sure that all the information added by the user is valid 1 verifyInformation This procedure checks to make sure that all the appropriate data has been entered by the user and then checks to make sure that all the information necessary to process the datafile exists 1n the databases For example it checks to see that there 1s a labor rate for the specified operator and checks to make sure that costs exist in the database for all attachments on the specified truck Commandl11 Click Remove file from database Button This procedure runs when you remove a file selected from a drop down menu by the user from the database It searches through the current database and deletes all records associated with the specified datafile It also deletes the database for that specific file from the databases subfolder of WiscPlow Lastly it deletes the shapefile used in the map display for the specified file from the datafile shapefiles subfolder of WiscPlow Command8_Click Remove file from process list Button Removes the
71. tion Center December 2005 Ot TRE 43 lt Table of Contents gt 1 Procedure 2 Graphical Representation of Procedure Map 3 WiscPlow Folder Structure e 44 1 PROCEDURE MAP PURPOSE This document is intended to aid in the understanding of WiscPlow s source code It attempts to show the hierarchical structure of WiscPlow s procedures Each section beginning with a bold header corresponds to a button pressed by the user in the WiscPlow interface What follows 15 the hierarchical progression of procedures within WiscPlow that are called when that button is pushed Only the most complicated procedures which involve calls to other subsequent procedures are mapped out in this document Other commands and procedures in WiscPlow that are not discussed here can be easily understood by anyone with programming experience in Visual Basic and therefore are not required in this document Loading the Main Form The Form Load procedure runs automatically every time WiscPlow s main form 15 loaded Form Load For the Main Form This procedure reads in all the data required on the main form It reads in global variables from the current ini file in the pathways subfolder of WiscPlow It also creates and configures the map objects used to create the Map Display in WiscPlow It populates the Storm Incident and Antiicing tab
72. tions will be necessary whenever alignments change new roadways are built or patrol section boundaries are changed Custodianship of these databases is a difficult issue Options include 6 WisDOT s local roads databases are maintained at headquarters in Madison The unit that does this work clearly has personnel with appropriate skills and technology for maintaining Wiscplow spatial databases However such activities are not within their mission 7 WIisDOT s district offices use GIS as a matter of routine and could develop the capability to maintain Wiscplow s spatial databases However WisDOT district offices typically do not maintain or manage data for county highway departments 8 Itis technically feasible for UW Madison to maintain the necessary spatial databases However the university is not in the practice of doing such things and some mechanism would have to be developed for continuity beyond the tenure of 38 current Research Team members Some level of continued funding would have to be developed 9 Consultants could be contracted to maintain the spatial databases Such an arrangement would perhaps be more institutionally stable than having UW Madison do the maintenance but a continuing funding source would have to be identified 10 The most attractive option is to have either the county highway departments or the county land information offices take custodianship of the spatial databases This keeps both use and
73. uez C and A Vonderohe 2005 Simple Map Matching Algorithm Applied to Intelligent Winter Maintenance Vehicle Data Transportation Research Record Journal of the Transportation Research Board No 1935 pages 68 76 Vonderohe A Malhotra A Sheth V Mezera D and T Adams 2000 Wisconsin Winter Maintenance Concept Vehicle Data Management Year 1 Final Project Report Submitted to the Wisconsin Department of Transportation Vonderohe A Adams T Malhotra A Sheth V Manchikanti K and D Mezera 2001 Wisconsin Winter Maintenance Concept Vehicle Data Management Year 2 Final Project Report Submitted to the Wisconsin Department of Transportation Vonderohe A Adams T Malhotra A Stanuch G and C Blazquez 2002 Wisconsin Winter Maintenance Concept Vehicle Data Management Year 3 Final Project Report Submitted to the Wisconsin Department of Transportation Vonderohe A Adams T Blazquez C and J Maloney 2003 Wisconsin Winter Maintenance Concept Vehicle Data Management Year 4 Final Project Report Submitted to the Wisconsin Department of Transportation 4 Appendix Wiscplow External Technical Documentation 42 WiscPlow Version 3 0 Technical Documentation Developed at The Department of Civil and Environmental Engineering University of Wisconsin Madison Funded by the Wisconsin Department of Transportation amp Midwest Regional University Transporta
74. validate the system s computational integrity to ensure that performance measures were being computed correctly and that decision management tools were producing correct results This was done by selecting a few small vehicle data files for Columbia and Portage Counties Wiscplow was used to parse the data files into database tables Then the data in the tables were used in hand computations and spreadsheets to produce check values for performance measures and inputs to decision management tools These check values were then compared to corresponding values produced automatically by Wiscplow Due to hardware difficulties Kenosha County had not collected any vehicle data files Therefore the Research Team generated fictitious but representative vehicle data files for Kenosha developed check values from them and compared the check values to Wisplow s results when using the Kenosha County Roadway network shapefile and patrol sections The second series of tests was used to ensure the integrity of the user interface to prevent or trap user errors and to ensure that execution time errors due to invalid data values would not occur This was done by having a member of the Research Team who was not involved in development of Wiscplow source code attempt to make the program fail by simulating as many user mistakes as could be identified When error trapping or user interface flaws were detected the source code was modified to correct them 7 T
75. ved Otherwise the algorithm loops back to point 3 and begins repetitions by examining the next nearest roadway within the buffer for point 3 The combination of both algorithms illustrated in Figures 6 and 7 calculates shortest paths between neighboring data points and alternative snapped locations for data points until a feasible path is obtained or until a specified number of consecutive points have been examined 4 2 The Algorithm Applied to Actual Data The example illustrated in Figure 8 includes a set of vehicle data points collected every five seconds by a winter maintenance concept vehicle during the 2002 2003 winter season in Columbia County Wisconsin Columbia County s roadway centerline spatial database has a nominal scale of 1 4 800 The spatial mismatch occurring at the diverging roadways in Figure 8 is resolved by the map matching algorithm In this example points 0 2 3 and 4 are Snapped to the nearest roadway within their buffers resulting in points SO S2 S3 and S4 shown as rectangles Points SO S3 and S4 are on the Interstate 39 centerline while point S2 is on the ramp centerline No roadways are contained within the buffer for point 1 thus this point is not used in determining the feasible path However point 1 is snapped to the correct roadway once the shortest path between the adjacent data points is determined to be feasible Figure 8 Resolution of Spatial Ambiguities at a Ramp The shortest path between
76. veling They arise from positional errors in the reported coordinates and in the digital cartography Without resolution they result in incorrectly computed performance measures 2 Conduct of two workshops for project participants The first workshop introduced participants to Wiscplow functionality and developed detailed concepts for refined performance measures and decision management tools The second workshop provided hands on use of Wiscplow after it had been revised according to needs identified at the first workshop The second workshop also stressed implementation issues such as the need for accurate roadway spatial databases and institutional support Wiscplow in business operations 3 Evaluation of WisDOT videolog GPS coordinate strings collected in 2001 and 2002 as potential roadway centerline representations for use with Wiscplow Analysis revealed gaps overlaps errors and missing segments e g ramps that rendered these data unusable for roadway centerline representations It was clear that the data would require extensive editing before they could be made sufficient for use with Wiscplow 4 Description of data requirements for the Wiscplow system Roadway spatial databases with 1 24 000 or larger source scale are required These spatial databases must be maintained and kept current They must include patrol sections and functional class attributes None of the participating counties possessed such data at the end
77. work shapefiles The map matching problem arising from errors in DGPS vehicle coordinates and in roadway centerline representations was addressed through development testing and implementation of a decision rule algorithm Wiscplow now associates nearly all vehicle data points with correct roadway centerlines Detailed internal and external technical documentation was developed for Wiscplow Such documentation is necessary for on going maintenance of Wiscplow beyond the tenure of the Research Team Detailed user documentation including tutorials and test data was developed and delivered to participating parties Such documentation is necessary to ensure Wiscplow s utility in the workplace Training was provided to county highway department personnel at the two installation sites Wiscplow was developed as a single user standalone system because its use at the local level was envisioned this way Wiscplow is not an enterprise solution Under its current configuration data from more than one county cannot be combined Five options for maintenance of roadway network shapefiles were described The most attractive of these options is to have county highway departments or land information offices take custodianship of the data However there are impediments including lack of resources and inexperience with GIS and therefore lack of track record with land information offices on the part of many county highway departments Institutiona

Final Report - Wisconsin Transportation Center

Contents

Download Pdf Manuals

Related Search

Related Contents