The Impact of Expectations about Automated and Manual Vehicles on Drivers' Behavior: Insights from a Mixed Traffic Driving Simulator Study

Abstract

As drivers' expectations guide their perception and behavior, violated expectations can lead to mistakes and discomfort. In this work, the role of expectations regarding drivers' reactions to automated (AVs) and manual vehicles (MVs) was investigated. Interviews were conducted to explore expectations toward AVs and MVs. Then, drivers interacted with AVs and MVs in a multi-agent driving simulator. The vehicles yielded or insisted on the right-of-way, indicated by a lateral offset. Self-report data revealed that drivers expected AVs to drive second (yield) and MVs to drive first (insist) in narrow passages. Driving simulator data showed that driving behavior improved, i.e., faster passing time, higher average speed, and higher lateral position, when AVs yielded and matched drivers' expectations, compared to MVs that behaved the same way. No improvement was found when MVs (vs. AVs) insisted on the right-of-way. Overall, yielding was evaluated more trustworthy and cooperative than insisting for both vehicle categories.