Supervising the self-driving car: situation awareness and fatigue during automated driving

Abstract

The capacity for human drivers to resume control from an automated vehicle remains a central focus of human factors research. Physiological measures promise to allow the vehicle system to determine when a driver is in a ready-state for transition of control, particularly for level 3 automation and above. We employ an adapted measure of Situation Awareness (SA) to assess the quality of driver SA following an extended period of simulated level 3 automated driving. It is hypothesised that a within-subjects design will demonstrate increasing passive fatigue to be predictive of reduced SA following a takeover request. Participants were also randomly allocated to one of two separate conditions in which supervising drivers were either permitted to, or prohibited from the use of non-driving related tasks (NDRT) during automated driving, to investigate a potential avenue for targeted SA enhancement through deliberate NDRT engagement. Preliminary results provide tentative support for our hypotheses.