Improving Automated Vehicle Safety with Collective Perception

  • Author / Creator
    Siqi Yan
  • Automated Vehicle (AV) is a trending technology being developed with the promise to reduce traffic accidents caused by human errors. Perception plays a crucial role for Automated Driving Systems (ADS) to make safe decisions. However, local sensory data is insufficient to capture comprehensive information for ADS due to occlusion and limited perception range. Vehicle-to-vehicle (V2V) communication technology provides an alternative solution by enabling vehicles to exchange perception information (referred to as Collective Perception or CP), increasing environment awareness for all vehicles in the network, and thus improving driving safety. In this research, we demonstrate the safety benefits of CP through the simulation of three driving scenarios in the Carla simulator. We use pose graph optimization (PGO) to fuse high-level sensory data between the local sensor and the data coming from the V2V network to reduce state estimation uncertainty. After that, a connected ADS was developed to make driving decisions based on the fused information to prevent potential traffic hazards. We show that the time-to-collision (TTC) metrics between the ego vehicle and unconnected road users (URU) are improved for all three scenarios compared to the baseline, indicating that CP can improve safety for AV in the simulated scenarios.

  • Subjects / Keywords
  • Graduation date
    Spring 2022
  • Type of Item
  • Degree
    Master of Science
  • DOI
  • License
    This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.