Is the NASA nap the solution for sleep inertia in automated driving? Unveiling effects of deep sleep on driving behavior, vigilance and subjective experience

Sleeping during automated driving is likely to become a common future practice. When drivers must resume control due to system boundaries, sleep inertia—post-sleep grogginess and impaired performance—affects their takeover and driving behavior. Research outside of driving indicates that deep sleep (AASM stage N3) can intensify sleep inertia. Operational strategies, like the NASA nap (capping nap duration to 40 min), aim to avoid deep sleep, though effects in driving contexts remain unclear. This study investigated how deep sleep impacts driving behavior, vigilance, and drivers’ subjective states.

In a driving simulator study, 20 participants completed two test sessions at different times of the day. Participants slept during an automated drive and drove manually both before and after sleep. Regression analyses assessed changes in driving behavior, vigilance, and subjective states in response to varying amounts of deep sleep.

With more deep sleep, takeover times and reaction times to auditory cues increased, and driving speed decreased. Lane keeping and stimuli detection rates were not significantly affected. Participants reported less motivation, delays in perception, discomfort, and struggles to stay awake after increased deep sleep.

Findings show that speed-related performance aspects were more affected by deep sleep than accuracy, suggesting drivers compensated for reduced cognitive abilities under sleep inertia. The results support the use of proactive strategies to avoid deep sleep. However, since deep sleep onset occurred rapidly, restricting nap duration might be insufficient to prevent sleep inertia.