Cooperative Mapping for Automated Vehicles

Localization is essential for automated vehicles, even for simple tasks such as lanekeeping. Some automated vehicle systems use their sensors to perceive their surroundings on-the-fly, such as the early variants of the Tesla Autopilot, while others such as the Waymo car navigate within a prior map. The latter approach is beneficial in that it helps the system to expect the expected, that is, it relieves the system of perceiving static features. However, making and updating such accurate prior maps using a specialized vehicle fleet is expensive and cumbersome. Techniques for Simultaneous Localization And Mapping (SLAM) are an attractive solution to this problem. SLAM uses visual and other sensors for creating and updating maps as the robot/vehicle navigates within the map. This project explores the possibility of using multiple vehicles to perform cooperative SLAM for improving and updating the map formed using optical cameras, radar, inertial measurement unit (IMU), and Global Navigation Satellite System (GNSS). It is assumed that raw data from these sensors can be shared among the vehicles over a wireless link either via vehicle-to-vehicle (V2V) or vehicle-to-infrastructure (V2I) communications.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; Photos; References;
  • Pagination: 27p

Subject/Index Terms

Filing Info

  • Accession Number: 01666073
  • Record Type: Publication
  • Report/Paper Numbers: D-STOP/2017/138, Report 138
  • Contract Numbers: DTRT13-G-UTC58
  • Files: UTC, NTL, TRIS, ATRI, USDOT
  • Created Date: Mar 15 2018 9:40AM