How Does Explanation-Based Knowledge Influence Driver Take-Over in Conditional Driving Automation?

This article focuses on explanation-based knowledge about system limitations (SLs) under conditional driving automation (society of automotive engineers level 3) and aims to reveal how this influences driver intervention. By illustrating the relationships between environment, system, mental model, in dynamic decision-making processing for responding an issued request intervene (RtI), occurrence SL, concept RtI, scene(s) related SL are determined by knowledge-based learning. Based three concepts, is examined at five levels: 1) no explanation, 2) 4) some typical scenes 5) all above. Data collection conducted a simulator, 100 people with experience automated participated. The experimental results show that instructing drivers situations contributes greater increase rate successful intervention car control from 55% 95%. Furthermore, them RtI conducive significant reduction response time 5.48 3.62 s their first RtI. It also revealed learning effect dwindles but does not vanish even after number times. Compared explaining possible driver, introducing better take-over performances critical or unexplained scenarios. demonstrates importance necessity knowledge, especially explanation sample which drivers' behavior.