Trust Calibration through Reliability Displays in Automated Vehicles

Abstract

Automated driving is a growing trend in commercially available vehicles. These technologies have the potential to improve safety, but only if drivers understand how they work and when they may fail. Displays that dynamically show automation reliability (and likelihood of failure) have been suggested to improve performance, especially when the driver needs to take back control of the car. There remain many open questions about how best to design such displays. The proposed work investigates metric type, level of information, and modality of presentation to determine how best to improve driver outcomes of situation awareness, trust, reliance and compliance, and overall driving performance.