Project Summary
Autonomous vehicle (AV) ridesharing has the potential to provide flexible, on-demand mobility, particularly to residents of rural communities that may not have access to more traditional public transportation. However, they still have trouble handling unpredictable “edge-case” scenarios, making human oversight necessary for safe operation. Rather than assign a dedicated safety driver to every vehicle, many AV rideshare companies have begun using remote assistants to help the AV computer interpret the world and decide how to act. However, making sure that these remote operators have the appropriate situation awareness to respond quickly and effectively to these assistance requests is an open problem. Similarly, the rider’s role becomes ambiguous when there is no onboard human driver. This project will also address the challenge of communicating user needs to the remote operators. In this REU project, students will design interfaces for both remote operators and end-users of autonomous transit systems. Undergraduates will identify user needs by leveraging publicly available datasets and a systematic literature review. They will use rapid prototyping (from paper sketches to Unity-based mockups) and scenario-driven simulation evaluations to design and evaluate interfaces to aid situation awareness and human-agent communication. The resulting interfaces should help operators maintain situational awareness without cognitive overload and empower riders to plan and adjust trips confidently.