DIGITAL MAP REQUIREMENTS FOR AUTOMATIC VEHICLE LOCATION

New Jersey Transit (NJT) is currently investigating acquisition of an automated vehicle locator (AVL) system. The purpose of the AVL system is to monitor the location of buses. Knowing the location of a bus enables the agency to manage the bus fleet more efficiently and to provide their customers with up to the minute information on bus arrivals and departures. To monitor the location of the buses, their positional information (as determined by the AVL) is displayed on a digital map such as a GIS. To ensure accurate information, the location (coordinates) of the bus must be consistent (or within a small tolerance) with those of the digital map. If this is not the case, the system may yield incorrect information. This problem may become especially critical in an urban area where the system would be most valuable. In this project, a methodology was developed for testing and evaluating the accuracies of an AVL system and supporting digital maps. The AVL system analyzed in this project was the Continuous Positioning System (CPS) by Anrew Corporation. Digital mapping products evaluated were TIGER/LINE, NAVTECH and Digital Orthophotos. The above data sets were evaluated with an accurate network of control points measured by Global Positioning System (GPS). Following the analysis of this study, some recommendations on the appropriateness of the tested AVL system and NJT’s digital mapping data were made. ACKNOWLEDGEMENT This study was supported by a grant from the National Center for Transportation and Industrial Productivity (NCTIP), and the New Jersey Transit (NJT). The Principal Investigator wishes to thank Dr. Louis Pignataro and Dr. Lazar Spasovic the former and present directors of NCTIP respectively, Mr. James W. Kemp, Louis Millan, Glenn D.Newman and Williams Shortes of New Jersey Transit for supporting and assisting this research. Additional thanks are in order to Mr. Robert Kidwell and Paul Krahmer of Andrew Sensor Products for conducting the CPS/GPS data acquisition runs and for providing the data output used in this study. Thanks are also due to the companies that provided data and equipment for this study. These companies include: 1. Andrew Corporation for the use of their CPS equipment and CPS/GPS data output. 2. Navigation Technologies Inc. for the Navtech Navigable Map Database. 3. Leica GeoSystems (in particular Mr. John White) for providing the MX8600-RT GPS receiver for measuring the GPS control points. 4. NJ Department of Environmental Protection (Mr. Larry Thorton) for providing Digital Orthophotos. 5. NJ TRANSIT Bus Operations, for use of a commuter bus and related operating support. TABLE OF CONTENTS ABSTRACT ACKNOWLEDGEMENT TABLE OF CONTENTS LIST OF FIGURES LIST OF TABLES 1.0 INTRODUCTION 2.0 DESCRIPTION OF EQUIPMENT USED IN THIS PROJECT 2.1 GPS Equipment For Position Determination Of The Control Points. 2.2 The Continuous Positioning System (Cps) 3.0 DESIGNING AN AVL / DIGITAL MAPPING SYSTEM EVALUATION 3.1 Selecting Routes For AVL (And GIS) Testing 3.2 Selecting Control Along The Tested Routes 3.21 Digital Orthophoto Quarter Quad (DOQQ). 3.3 Control Point Measurements With GPS 4.0 DATA ANALYSIS 4.01 Route Name Convention 4.02 Defining An Acceptable Error Tolerance. 4.03 The Data Analysis Process 4.04 Terms and Parameters Used in the Following Tables and Figures. 4.1 Comparison Between Recorded GPS Data And The Reference Data. 4.2 Comparison Between Recorded CPS Data And The Reference Data. 4.3 Comparison Between TIGER/Line And Navtech Data Sets, And The Reference Data 4.4 Accuracy Analysis With Respect To GPS Control Points. 4.41 Comparison Between GPS Control Points And Digital Map Data (TIGER/Line And Navtech). 4.42 Comparison Between GPS Control Points And AVL Data (CPS And GPS). 4.5 Consistency (Repeatability) Analysis Of Two Runs Of The Same Route 4.6 Location Error CPS Marked Points With Respect to GPS Control 4.7 Location Error as a Function of Environment Ranking 5.0 SUMMARY OF INACCURATE DATA. 6.0 SUMMARY OF FINDINGS AND CONCLUSIONS Appendix A Appendix B Appendix C LIST OF FIGURES Figure 1 The Leica MX-8600-RT Figure 2. Continuous Positioning System (CPS) block diagram Figure 3. The CPS equipment enclosure Figure 4. Bus Route of August 18 Figure 5. Bus Route of August 19 Figure 6. Newark Bus Route of August 19 Figure 7. File name convention for CPS and GPS data Figure 8. Positional errors of GPS only. Figure 9. The distribution of GPS errors. Figure 10. Positional error of CPS only. Figure 11. The distribution of CPS errors. Figure 12. Combined CPS and GPS error Figure 13. Distances between Navtech and TIGER/Line data sets, and GPS control points. Figure 14. The accuracy of CPS and GPS with respect to GPS control (8/18 run). Figure 15. The accuracy of CPS and GPS with respect to GPS control (8/19 run). Figure 16. Error determination with respect to GPS control. Figure 17. Location error the CPS with respect to the GPS control. Figure 18. Location error in various operation environments of the AVL system with respect to the GPS control. Figure 19. Index of error locations 8/18 Figure 20. Index of error locations 8/19 Figure 21. Location 1, CPS Figure 22. Location 1, GPS Figure 23. Location 2, CPS Figure 24. Locations 2, GPS Figure 25. Location 3, CPS Figure 26. Location 3, GPS Figure 27. Location 4, CPS Figure 28. Location 4, GPS Figure 29. Location 5, CPS Figure 30. Location 5, GPS Figure 31. Location 6, CPS Figure 32. Location 6, GPS Figure 33. Location 7, CPS Figure 34. Location 7, GPS Figure 35. Location 8, CPS Figure 36. Location 8, GPS Figure 37. Location 9, CPS Figure 38. Location 9, GPS Figure 39. Location 10, Navtech Figure 40. Location 11, Navtech Figure 41. GPS point 1, North Broad St. and Ridgeway, Hillside Figure 42. GPS point 2, Westfield Av. And Elmora Av., Elizabeth Figure 43. GPS point 3, Chestnut St. and West Westfield Ave., Roselle Park Figure 44. GPS point 4, Aiden St. and Route 28, Cranford Figure 45. GPS point 5, Elm St. and North Ave., Westfield Figure 46. GPS point 6, North Martine Ave. and Midway Ave., Fanwood Figure 47. GPS point 6 (view II), North Martine Ave. and Midway Ave., Fanwood Figure 48. GPS point 7, Wiley Ave and East Front St., Scotch Plains Figure 49. GPS point 8, West End Ave. and West Front St., Plainfield Figure 50. GPS point 9, Jackson Ave. and North Ave., Dunellen Figure 51. GPS point 10, South Washington Ave. and North Ave., Dunellen Figure 52. GPS point 10 (view II), South Washington Ave. and North Ave., Dunellen Figure 53. GPS point 11, Jackson Ave. and North Ave., Dunellen Figure 54. GPS point 12, West End Ave. and West Front St., P lainfield Figure 55. GPS point 13, Central Ave. and West 2nd St., Plainfield Figure 56. GPS point 14, Watchung Ave. and East 5th St., Plainfield Figure 57. GPS point 15, Elm St. and North Ave., Westfield Figure 58. GPS point 16, Aiden St. and West Westfield Ave., Roselle Park Figure 59. GPS point 17, Chestnut St. and West Grant Ave. Roselle Park Figure 60. GPS point 18, Chestnut St. and west Lincoln Ave., Roselle Park Figure 61. GPS point 19, Marshall Ave and Salem Rd., Roselle Park Figure 62. GPS point 20, Bloy St. and Princeton Ave., Hillside Figure 63. GPS point 31, Raymond Plz W and Raymond Blvd., Newark Figure 64. GPS point 32, Washington St. and Raymond Blvd., Newark Figure 65. GPS point 33, Lock St. and New St., Newark Figure 66. GPS point 34, Dr. Martin Luther King Blvd. And New St., Newark Figure 67. GPS point 36, Washington St. and Hill St., Newark Figure 68. GPS point 37, Broad St. and Market St., Newark Figure 69. GPS point 40, Commerce Ct. and Commerce St., Newark Figure 70. GPS point 42, Broad St. and Count St., Newark LIST OF TABLES Table 1. Continuous Positioning System Characteristics Table 2. The location and the ranking of the control points. (*) Indicate "Marked Points" Table 3. The results of the GPS control survey. The units are in meters. Table 4. Bus routes on which the AVL system was tested. Table 5. Positional error of GPS only. Table 6. Positional error of CPS only. Table 7. Combined CPS and GPS errors. Table 8. The accuracy of the TIGER/Line and Navtech data sets Table 9. Distances between Navtech and TIGER/Line data sets, and GPS control points Table 10. The accuracy of CPS and GPS with respect to GPS control. Table 11. Consistency analysis of the AVL system. Table 12. Location error in various operation environments of the AVL system with respect to the GPS control. Table 13. The locations where CPS, GPS and Navtech data errors were found. DIGITAL MAP REQUIREMENTS FOR AUTOMATIC VEHICLE LOCATION 1.0 INTRODUCTION Automated Vehicle Locator (AVL) is a technology that enables a fleet operator to track and monitor the location of its vehicles at any given time. It is being used mainly in Transit and in commercial vehicle operation (management) systems. In Transit applications, the information on the exact location of a vehicle enables the operator to provide more accurate information to its customer. Furthermore, knowing the exact location of the buses enables speedy reaction to operation related problems. Most AVL systems provide location information in terms of coordinates (latitude and longitude or northing and easting). These coordinates have to be matchable with coordinates on a map so that the location of the vehicle can be uniquely identified. In order for the matching process to be successful, the coordinate systems of the map and those of the AVL system must be consistent, compatible and sufficiently accurate. The accuracy requirement is such that the AVL location should correspond to a unique and unambiguous point on the map. For many AVL applications, where the various stops or destinations of the vehicle are spread out over a large area this matching process can be achieved with rather moderate accuracies of the AVL and the digital map systems. However, in a Transit application that operates in densely built urban and suburban areas with frequent bus and traffic control stop