Vehicle-Based Drowsy Driver Detection: Current Status and Future Prospects

Driver drowsiness is a major, though elusive, cause of traffic crashes. As part of its IVHS/human factors program, NHTSA is supporting research to develop in-vehicle systems . to continuously monitor driver alertness and performance. Scientific support for the feasibility of this countermeasure concept is provided by research showing that: l Drowsy drivers typically do not “drop off’ instantaneously. Instead, there is a preceding period of measurable performance decrement with associated psychophysiological signs. l Drowsiness can be detected with reasonable accuracy using driving performance measures such as “drift-and-jerk” steering and fluctuations in vehicle lateral lane position. l The use of direct, unobtrusive driver psychophysiological monitoring (e.g., of eye closure) could potentially enhance drowsiness detection significantly. l The use of secondary/subsidiary auditory tasks (e.g., auditory recognition tasks presented to the driver via recorded voice) could further enhance detection accuracy. The envisioned vehicle-based driver drowsiness detection system would continuously and unobtrusively monitor driver performance (and “micro-performance” such as minute steering movements) and driver psychophysiological status (in particular eye closure). The system may be programmed to provide an immediate warning signal when drowsiness is detected with high certainty, or, alternatively, to present a verbal secondary task via recorded voice as a second-stage probe of driver status in situations of possible drowsiness. The key requirements and R&D challenges for a successful countermeasure include low countermeasure cost, true unobtrusiveness, an acceptably-low false alarm rate, non-disruption of the primary driving task, compatibility and synergy with other IVHS crash avoidance countermeasures, and a warning strategy that truly sustains driver wakefulness or convinces him/her to stop for rest. Knipling and Wierwille Page 3