MANAGING FATIGUE BY DROWSINESS DETECTION: CAN TECHNOLOGICAL PROMISES BE REALIZED?

Abstract

This chapter reviews issues and research underway on the validation of technologies purporting to monitor motor vehicle operator alertness. Although such efforts have a long history, they have increased markedly in recent years, owing to the prevalence and seriousness of fatigue-related crashes, the unreliability of subjective estimates of sleepiness impairment, the potential of drowsiness-detection technologies as a component in alternatives to proscriptive hours of service, and the fact that technological advances have made the goal of on-line drowsiness detection feasible. Fatigue-management technologies fall into four broad categories: (1) readiness-to-perform and fitness-for-duty; (2) mathematical models joined with ambulatory technologies; (3) vehicle-based performance technologies; and (4) in-vehicle, online, operator status monitoring technologies. The latter have the greatest number of initiatives, spanning a wide range of biobehavioral variables. Scientific, implementation, and legal/policy criteria for drowsiness detection technologies are reviewed, with a discussion and brief review of a study illustrating the fundamental design and methodological requirements needed to establish scientific validity and reliability for drowsiness-detection technologies. (a) For the covering entry of this conference, please see IRRD abstract no. 895120.