Abstract:Objective Conditional automated driving (SAE level 3) requires the driver to take over the vehicle if the automated system fails. The mental workload that can occur in these takeover situations is an important human factor that can directly affect driver behavior and safety, so it is important to predict it. Therefore, this study introduces a method to predict mental workload during takeover situations in automated driving, using the ACT-R (Adaptive Control of Thought-Rational) cognitive architecture. The mental workload prediction model proposed in this study is a computational model that can become the basis for emerging crash avoidance technologies in future autonomous driving situations.

Driver behavior and mental workload for takeover safety in automated driving: ACT-R prediction modeling approach