Real-Time Detector of Human Fatigue

We propose to develop a computer based system for the online analysis of bio-behavioral signals associated with momentary loss of alertness. We will continue our development of bio-behavioral measures indicative of lapses in alertness. Our system will use only measures that can be obtained without attaching sensors to the operator. We will concentrate on monitoring the oculomotor system using cameras and monitoring the cardio respiratory and neuromuscular systems using a Laser-Doppler Vibrometer. Our results to date suggest that many of the measures included in our battery will need to be adapted for individual subjects. Thus we will develop algorithms for which customization to account for individual differences can be an automatic process that is transparent to the operator. The system will be tested and validated by introducing probe stimuli during laboratory tasks at points where the system identifies a lapse in alertness. We also propose to transition the system to evaluate alertness in sleep deprived subjects working in military simulation environments.