CAV Testing Scenario Design and Implementation using Naturalistic Driving Data and Augmented Reality

Testing and evaluation is a critical step in development and deployment of connected and automated vehicle (CAV) technology. Testing standards for human driven vehicles, such as Federal Motor Vehicle Safety Standards (FMVSS), have been established a long time ago. However current standards can’t be applied to CAVs, because they often assume the presence of a human driver, who conducts the driving tasks. It is very important to develop test procedures and identify applicable testing scenarios (user cases) for CAVS to evaluate the “intelligence” of the vehicle. The intelligence level indicates whether a CAV can drive safely and efficiently without human intervention. The newly released Automated Driving Systems Guideline 2 has made it very clearly that the new automated driving systems needs validation methods and needs be tested by incorporating behavior competencies. In this project, the project team will investigate how to design such testing scenarios by looking into crash and naturalistic driving databases, and how to implement the defined scenarios in the augmented reality (AR) testing environment. The team will focus on testing higher levels of automation defined by SAE (level 3 or higher), in which human behaviors are much less involved in the driving tasks. As a result, the team will exclude crash or near-crash events due to human errors (e.g., distraction or fatigue). The team assumes the on-board sensors and vehicle control systems work as designed and will only select scenarios that place challenges to sensor perception (e.g. a pedestrian/bicyclist suddenly appears on the roadway) and vehicle decision making. This study will also consider the scenarios that require more sophisticated human driving skills and decision making procedures. For example, turn left from a driveway with stop sign to a major arterial. The driver needs to seek gaps from both directions of the traffic. The driver also has alternatives including make a stop in median lane first or directly drive to the lane of the opposite direction. This brings challenges to both vehicles sensors and decision making algorithms. Several representative testing scenarios will be identified and implemented in the augmented reality (AR) testing environment. The AR environment integrates a real-world testing facility and a simulation platform, so that CAVs can be tested and evaluated with realistic virtual traffic in a safe and cost-effective way. The identified testing scenarios will first be constructed in the simulation platform with realistic driver/pedestrian/bicyclists behaviors calibrated from naturalistic driving data (NDD). A real CAV will be tested under the scenarios and its performance will be recorded and evaluated in terms of safety and efficiency.

Language

  • English

Project

  • Status: Active
  • Funding: $135025
  • Contract Numbers:

    69A3551747105

  • Sponsor Organizations:

    Office of the Assistant Secretary for Research and Technology

    University Transportation Centers Program
    Department of Transportation
    Washington, DC  United States  20590
  • Managing Organizations:

    Center for Connected and Automated Transportation

    University of Michigan, Ann Arbor
    Ann Arbor, MI  United States  48109
  • Performing Organizations:

    University of Michigan Transportation Research Institute

    2901 Baxter Road
    Ann Arbor, Michigan  United States  48109
  • Principal Investigators:

    Feng, Yiheng

    Liu, Henry

  • Start Date: 20170110
  • Expected Completion Date: 20190330
  • Actual Completion Date: 0
  • USDOT Program: University Transportation Centers Program

Subject/Index Terms

Filing Info

  • Accession Number: 01646273
  • Record Type: Research project
  • Source Agency: Center for Connected and Automated Transportation
  • Contract Numbers: 69A3551747105
  • Files: UTC, RiP
  • Created Date: Sep 22 2017 3:53PM