Presentation
Assessing Drivers' Trust, Compliance, and Reliance in an Automated Flood Warning System: Effects of Errors Types and System Reliability
DescriptionFlooding is an emerging threat caused by precipitation changes and floodplain development. It is crucial to establish efficient real-time flood communication methods to report flood situations to drivers. Currently, these communication methods rely on automated, or machine-learning-algorithm based flood-warning systems. However, automated systems are susceptible to committing errors, such as false alarms and misses. Consequently, different error types can affect drivers’ trust levels and trustworthiness in the automated flood-warning system. This research examined how system accuracy and error type impacted drivers’ trust, compliance, and reliance on the system. Our results indicated that both false alarms and misses resulted in drivers’ lower perceived system reliability. Additionally, misses impacted drivers’ reliance more, while false alarms influenced drivers’ compliance more. Implications of this research can inform the design of the automated systems, especially in the flood-warning context, to improve drivers’ trust calibration in responding to potential error types and different system reliability levels.
Contributors
Event Type
Lecture
TimeTuesday, September 10th10:05am - 10:25am MST
LocationFLW Salon G
Surface Transportation
DEI