Abstract
It has been shown that distraction is a major factor contributing to roadcasualties. Detecting and limiting driver impairment could improve road safety. With the introduction of unobtrusive gaze-tracking systems gaze based real-time algorithms can be developed for driver state monitoring. The data reported here stem from an extended field study, during which seven drivers used an instrumented car for one month each. Eye tracking quality is best when the gaze is directed to the road center. The head tracking data show a bimodal distribution, which is interpreted as the drivers leaning the head against the head rest at times.