Design: paired trials without and with naps and bright light. Setting: Real driving on a private road circuit. Environmental controlled car. Participants: 9 shift workers on tree shifts (morning-afternoon-night) Measurements: Sleepiness at the wheel was measured by ambulatory polysomnography and assessed using 30 seconds segments of recordings when the percentage of theta EEG was at least 50% (15 seconds) of the period recorded. Subjects were also called to rate their sleepiness on the 7-point Stanford Sleepiness Scale (SSS).

Intervention: Participants drove the same car on two similar 24 hours periods of work, with three pilots in each shift (morning, afternoon, night), separated by three weeks. During the baseline period, the subjects were told to manage their rest as they usually do in the real life. During the second experimental period, they had to rest lied in a dark room during two naps of 20 minutes and then exposed to bright light pulse (5000 lux) during 10 minutes.