Error Type, Risk, Performance, and Trust: Investigating the Different Impacts of false alarms and misses on Trust and Performance

Semi-autonomous vehicles are intended to give drivers multitasking flexibility and to improve driving safety. Yet, drivers have to trust the vehicle’s autonomy to fully leverage the vehicle’s capability. Prior research on driver’s trust in a vehicle’s autonomy has normally assumed that the autonomy was without error. Unfortunately, this may be at times an unrealistic assumption. To address this shortcoming, we seek to examine the impacts of automation errors on the relationship between drivers’ trust in automation and their performance on a non-driving secondary task. More specifically, we plan to investigate false alarms and misses in both low and high risk conditions. To accomplish this, we plan to utilize a 2 (risk conditions) × 4 (alarm conditions) mixed design. The findings of this study are intended to inform Autonomous Driving Systems (ADS) designers by permitting them to appropriately tune the sensitivity of alert systems by understanding the impacts of error type and varying risk conditions.This research is supported in part by the Automotive Research Center (ARC) at the University of Michigan, with funding from government contract DoD-DoA W56HZV-14-2-0001, through the U.S. Army Combat Capabilities Development Command (CCDC)/Ground Vehicle Systems Center (GVSC).Peer Reviewedhttps://deepblue.lib.umich.edu/bitstream/2027.42/149648/1/GVSETS 2019_FinalPaper.pdfDescription of GVSETS 2019_FinalPaper.pdf : Main Fil