Eye-Com Research Validates PERCLOS for Drowsiness Detection: Part 2

This is the second of a series of five posts detailing research on the PERCLOS measure as an indicator of drowsiness conducted by Eye-Com Corporation in cooperation with the U.S. Department of Defense and the U.S. Department of Transportation.

Part 2: The PERCLOS MEASURE

The PERCLOS Measure is a Reliable Indicator of Driver Drowsiness and Fatigue

PERCLOS stands for the percentage of eyelid closure over the pupil over time. A PERCLOS-monitoring device provides video taken of the eyes of the operator in order to score the percentage of a specified time that the eyelid is at least 80% closed over the pupil. First studied in 1994, PERCLOS soon proved to be more effective at drowsiness detection than all other available technology at the time. Later PERCLOS studies prompted the National Highway Traffic Association Safety Administration (NHTSA) and the US Department of Transportation (DOT) to declare PERCLOS “the most reliable and valid bio-physiological determinant of a driver’s alertness level.” The NHTSA Drowsy Driver Technology program has since produced many validation studies of the PERCLOS metric in laboratory, simulator, and applied settings.

After a review of these studies, it was determined that the ideal drowsiness-detection system would conform to NHTSA’s standard that an in-vehicle monitoring system should be unobtrusive to the vision or activity of the operator. The challenges to developing such a technology included designing a system that operates under all lighting conditions and environmental domains, especially where the driver’s head or face is constantly moving or where the driver’s eyes are obscured by sunglasses, lenses and other eye-ware, glare, or oncoming headlights. Keeping this in mind, Eye-Com Corporation developed the ambulatory and wearable Eye-Com Biosensor, Communicator and Controller System, also known as the EC-6.

Continued in Part 3: Detecting Drowsiness with the Eye-Com Eye-Com Biosensor, Communicator and Controller

Comments