Drowsy Driver Monitor and Warning System

Richard Grace, Sonya Steward
2001 Proceedings of the First International Driving Symposium on Human Factors in Driver Assessment, Training and Vehicle Design: driving assessment 2001   unpublished
The design and use of a low-cost drowsy driver monitor, the Copilot, and proposed driver interfaces are presented. The Copilot consists of a digital camera integrated with a low-cost digital signal processor (DSP). The Copilot is a functionally enhanced version of a previous monitor that has been successfully used in a variety of research projects in simulators and in over-the-road vehicles. Compared to its predecessor, the new monitor is small and easy to use, providing an effective research
more » ... ol for the field or in the laboratory. The Copilot measures slow eyelid closures as represented by PERCLOS (Percent Eyelid Closure). PERCLOS is defined as the proportion of time that a subject's eyes are closed over a specified period. PERCLOS has been separate-ly validated in two independent laboratories as an accurate predictor of performance degradation in sleep-deprived subjects. The current driver interface is based on recent experimental results that drowsiness feedback can reduce drowsiness and improve driver performance for sleep deprived truck drivers operating a truck simulator. This controlled experiment was undertaken with n = 16 Commercial Driving License (CDL) holders driving a high-fidelity truck simulator (TruckSim) to establish the effects of drowsiness feedback on: (1) driver alertnessdrowsiness; (2) driving performance and (3) driver-initiated behaviors. Subjects served as their own controls, driving one simulated 4-hr night drive without drowsiness feedback (control condition) and one simulated 4-hr night drive with drowsiness feedback (feedback condition). Although there was significant between-subject variability in drowsiness and consequently in the number of drowsiness-based alarms and warning alerts, drowsiness feedback tended to have consistent effects on key classes of outcome variables, including reduced drowsiness levels, improved driver performance and self alerting activities (driver movements). The warning triggers are associated with PERCLOS calculated over three minutes. The current interface consists of an audible tone that is associated with the
doi:10.17077/drivingassessment.1010 fatcat:2xss7pzsmjhizdkjh2s3pgrzxe