Navigating These Mean Streets: Real-Time Mapping in Autonomous Vehicles

In this article the authors discuss the use of mobile mapping systems for applications in autonomous vehicles; the development of this is highlighted by the Defense Advanced Research Projects Agency (DARPA)-sponsored Urban Challenge showcase. Mobile mapping systems use a combination of georeferencing systems and visual imaging systems. Difficulties that these systems face in autonomous navigation include localization under Global Positioning System (GPS) blackouts, vehicle environment sensing, tracking other vehicles, safety, and long-term autonomy and robustness. The sensor fusion developed for the researcher’s Autonomous City Transport (ACT) vehicle is described. The objectives of the fusion were to sense and quantify local vehicle vicinity and also to provide terrain, man-made feature, and natural setting information for path planning. Position fusion was provided by DGPS, an inertial navigation system (INS), and compass. Sensor fusion was provided by LADAR, RADAR, monovision, stereovision, and SONAR. The vehicle equipped with this sensor array will be tested in DARPA’s Urban Challenge. Future developments of autonomous vehicles and navigation systems thereof are speculated upon in closing.

Language

  • English

Media Info

  • Media Type: Print
  • Features: Figures; Photos;
  • Pagination: pp 32-37
  • Serial:
    • GPS World
    • Volume: 18
    • Issue Number: 10
    • Publisher: Advanstar Communications, Incorporated
    • ISSN: 1048-5104

Subject/Index Terms

Filing Info

  • Accession Number: 01082296
  • Record Type: Publication
  • Source Agency: UC Berkeley Transportation Library
  • Files: BTRIS, TRIS
  • Created Date: Nov 30 2007 7:25AM