Allocation of Drivers’ Visual Attention During Preliminary Uses of Automated Driving: A Wizard-of-Oz Study

Abstract

In this experiment, we analyzed drivers’ visual attention during their first experiences with automated driving (AD). At the beginning of the study, participants ( n = 45) received one of three AD training modules: paper, video, or practice. They then drove two sessions on a public road with a Wizard-Of-Oz vehicle, each ending with a request to intervene (RTI). The first session (forward path) included 10 min of AD with a mandatory non-driving-related task (NDRT). The second session (return path) included 10 min of AD with a nonmandatory NDRT. Control checks (number and duration of gazes) toward the road and rear-view mirrors and three self-assessment questionnaires (trust, acceptability, and technophilia) were measured during both sessions. The results indicated a decrease in road and rear-view mirror gazes over time, mainly for the practice group. Drivers with a low level of trust glanced at the road and mirrors more often than participants with a high level of trust. In addition, participants with a high level of technophilia spent less time controlling the road traffic but only during the forward path. Visual attention on the road decreased rapidly during the first minutes of AD, mainly among drivers who followed the practice training and those who had a high level of trust and were tech-savvy. We observed that participants looked at the rear-view mirrors after 10 and 14 s on average after the RTI. These results raise questions about driver situation awareness in critical situations, including the capacity to intervene in an efficient and safe way during takeover requests.