Explaining Authoritative Control Interventions in Automated Driving to Support Driver Understanding, Trust, and Reliance

Abstract

Future automated driving systems (ADS) may be given the ability to fully block or takeaway driver control authority; for example, in cases of immediate danger. During these interventions, drivers will rely on the environment and the human–machine interface to explain the system’s behavior and to regain situational awareness. To examine how information can be structured to explain an authoritative control intervention, $N=46$ participants were exposed to three text message variations following interventions in two different driving scenarios in a simulator study. Text messages affected drivers’ objective and subjective responses to interventions. Including “Why” information about the reason for the automation’s behavior rather than “How” information about the control authority distribution increased understanding, trust, and appropriate reliance upon the automation. However, the evaluation of the automation’s behavior depended primarily on differences in the driving scenario, while text messages had only a minor influence. Driver gaze varied based on the driving scenario, with more glances to the HMI when driver control authority was taken away compared to when it was blocked. These findings have implications for human–machine interaction design, policy, and future research, highlighting the need for clear communication strategies during authoritative control interventions.