What leads to reliance on automated vehicles? An inferential analysis of responses to variable AV performance

The appropriate use of automated vehicle technology is pivotal for reducing latent security risks associated with automated driving. Appropriate use demands variability in reliance on the automation—specifically, relying on the automation only when it is capable of similar or better performance than human drivers. The central role of reliance in appropriate use suggests a pressing need to understand the factors that contribute to automation reliance. We address this need through a driving simulator study of 49 participants, where drivers were asked to make reliance decisions with partial vehicle automation technology across four events representative of reliance decisions made in current automated vehicles. We used step-wise logistic regression analysis to assess the role of reliance inertia, situational trust, dispositional trust, situation awareness, and driver demographics on reliance. The results suggest that the factors that influence reliance vary by the traffic scenario, that reliance inertia has the strongest influence on subsequent reliance decisions, and that reliance was more strongly related to situational trust than dispositional trust. These findings suggest a need for technologies that calibrate reliance rather than trust in AVs and for additional broader studies of driver reliance on AV.