Detection of Human Fatigue
Abstract
Military personnel are frequently required to perform tasks for which fatigue is a common problem. Lapses in alertness associated with fatigue are major contributors to human error. We have proposed to create a system that will use unobtrusively obtained bio-behavioral measures to detect such lapses in alertness and thus lower the incidence of human errors. In the current phase of this work, we have developed and used a test task (EPVT) which incorporates the need for visual search activity as well as involves a cognitive component. For non-sleep deprived subjects, this task generates lapses in performance as demonstrated by significant changes in reaction time (RT) as a function of Time-on-Task. We have also discovered a number of bio-behavioral variables for which significant alteration in the variables is associated with long latency responses and thus with lapses in alertness. The most impressive result to-date demonstrates significant differences in pupil diameter associated with stimuli where RT is slow, normal and fast. Significant changes in a number of bio-behavioral measures also occur as a function of ToT. There are individual differences in the occurrence of bio-behavioral signs of impairment in alertness to the task. We are developing prototype software for monitoring operator alertness.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 05, 2005
- Accession Number
- ADA437371
Entities
People
- John A. Stern
- Timothy B. Brown