Bystander Interactions with Failing Vehicle Autonomy

Significant concern has been expressed about what will happen when autonomous vehicles fail while driving. This issue has been the source of many opinions and some suitable strategies have been proposed. However, little is known about proper ways for managing failure using methods that do not alienate potential purchasers of such systems. Inappropriate driver-vehicle interaction will lead to decreased safety and poor road user experiences. While the latter may not seem important, it will directly impact market adoption, public perceptions of autonomy quality, and other factors that could delay or derail otherwise successful market penetration of safety-enhancing technologies. Most of the research on vehicle autonomy failure centers on vehicle-to-driver handoffs. This is an important problem and significant resources are being devoted to this challenge. However, little attention is being spent on bystander interactions. How will pedestrians, bicyclists, and drivers of other vehicles react and interpret autonomous actions designed to keep the vehicle safe? For example, a likely failure mode when a successful handoff does not, or cannot, occur is for the vehicle to autonomously pull to the side of the road. Safety can be improved dramatically if the car is able to express the failure is occurring and a safety-oriented motion is underway. Drivers currently use hazard lights, extended turn signals, and other methods to inform nearby road users. Is it appropriate to use the same signals? Are new signals needed? How does wireless technology help or hinder such communication? This project will extend knowledge and findings from an existing, related project on appropriate human-robot interaction during autonomous robot failure. The project is focusing on new methods for autonomous robots to express failures so that operators and bystanders take appropriate action. Methodologies and lessons from this project will be applied to the proposed effort, thereby accelerating project activities.


    • English


    • Status: Active
    • Sponsor Organizations:

      Technologies for Safe and Efficient Transportation University Transportation Center

      Carnegie Mellon University
      Pittsburgh, PA  United States  15213

      Office of the Assistant Secretary for Research and Technology

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

      Carnegie Mellon University

      Pittsburgh, PA  United States 
    • Project Managers:

      Ehrlichman, Courtney

    • Performing Organizations:

      Carnegie Mellon University

      Pittsburgh, PA  United States 
    • Principal Investigators:

      Steinfeld, Aaron

    • Start Date: 20170101
    • Expected Completion Date: 20180730
    • Actual Completion Date: 0
    • USDOT Program: University Transportation Centers Program
    • Subprogram: DTRT-13G-UTC-26

    Subject/Index Terms

    Filing Info

    • Accession Number: 01645873
    • Record Type: Research project
    • Source Agency: Technologies for Safe and Efficient Transportation University Transportation Center
    • Files: UTC, RiP
    • Created Date: Sep 6 2017 7:44PM