Let Tesla Park Your Tesla: Driver Trust in a Semi-Automated Car

The reality of highly automated vehicles on every road seems increasingly possible. With companies such as Tesla, Google, Toyota, and many others racing to provide a fully autonomous vehicle, the need for research on self-driving cars has never been greater. Until recently, however, most of this research had been conducted in a sterile lab environment devoid of any real consequences. For that reason, we join a host of other researchers in evaluating human-automation interaction in the real world associated with miscalibrated trust. As previous research has shown, drivers can either over- or under trust a vehicle's automated features. To evaluate this in these in a realistic setting, we had participants use the Autopark feature in a Tesla Model X or park the car themselves in both parallel and perpendicular scenarios. Parking times, driver trust, self-confidence in their own ability to park, and workload were all evaluated throughout the experiment. Preliminary analyses into the data are reported. Trends for the interactions between parking condition (self versus auto) and the parking type (parallel versus perpendicular) emerged for both trust/self-confidence and workload. Data collection is still ongoing to evaluate whether these trends hold, and if they emerge as significant. In all, this study contributes to the growing body of literature which seeks to understand the complexities of human-automation interaction in the real world.