Sleep in automated driving - Effects of time of day and chronotype on sleepiness and sleep inertia

Sleep is a desired use case in highly automated driving and promises drivers relief from sleepiness. Benefits of sleeping on subjective mood and performance in the driving task may be opposed by sleep inertia, the post-sleep grogginess and performance impairment. The magnitude of both, sleepiness and sleep inertia may depend on sleep-homeostatic and circadian factors. We therefore investigated how the effects of sleep inertia are affected by the time of day a nap is taken during automated driving and the role of individual chronotypes both in comparison to the effects of sleepiness. In the presented driving simulator study, 10 distinct morningness and 10 eveningness chronotypes were invited to two test sessions, one in the early morning (starting at 6 a.m.) and one in the late evening (starting at 9 p.m.). Participants slept during an automated drive and were requested to drive manually before and after an instructed nap. We examined the time course of the subjective state, driving behavior and vigilance in dependence of time of day and individual chronotype under sleepiness and sleep inertia. Effects of both, sleepiness and sleep inertia on subjective arousal, wellbeing, motivation, and driving speed were stronger when the time of day did not fit the individual chronotype. This counts particularly for morningness types. Subjective arousal, wellbeing, and motivation were more impaired after the instructed nap in the evening than in the morning. Analogously, driving speed was slower and take-over times were longer after napping in the evening compared to in the morning. Compared to sleepiness, lane keeping and reaction times were not significantly impaired under sleep inertia, probably because the task induced workload could be compensated by reducing speed and prolonging take-over times. Under sleepiness, driving was characterized by distinct time-on-task effects, which caused mood, lane keeping, and vigilance to deteriorate with driving time. In contrast, the investigated measures were more stable with time on task under sleep inertia, possibly due to a compensation by speed reduction and a balance between beneficial and detrimental effects of the intermediate sleep.