Trust Dynamics in Driving Automation: The Effects of Driver Expectations and Error Consistency

The advancement of driving automation systems is transforming surface transportation. Understanding how drivers develop trust in these systems is crucial for their effective implementation. This study investigated the influence of drivers’ initial expectations and the consistency in system errors on trust in a Level-3 automated driving system (ADS). Participants read descriptions that characterized system capabilities to be either high or low, following which their initial levels of expectation were assessed. Error patterns (no, consistent, inconsistent errors) were manipulated across three simulated drives. Subjective trust ratings after each drive and reaction time to takeover requests (TORs) were measured. Results showed that initial expectations did not significantly impact overall trust or TOR performance; instead, trust was adjusted based on the system’s actual performance. These suggested a greater influence of direct experience over drivers’ preconceived expectations. The perceived predictability of the system partially mediated the effect of error consistency on trust, with inconsistent errors worsening TOR performance. The study highlighted the need for predictable ADS designs and driver-system interactions to enhance trust and road safety.