Estimating Pilot Fatigue in Commercial Flight Operations

Abstract

In order to assess airline pilot duty fatigue levels associated with normal operations, subjective fatigue, sleep cycles were unobtrusively monitored and compared to the estimates of a fatigue prediction algorithm (FADE). A group of 9 commercial airline pilots completed log sheets on which sleep, flight data and periodic estimates of fatigue levels were recorded over a 10-day period. The subjective fatigue scores indicated a significant increase during the 2000-0400 hours time block. The lowest reported fatigue scores occurred during the 0800-1200 hours. Hours of sleep predicted pilot fatigue levels better than circadian time, hours of flight, time zones crossed or hours of non-flying work. A fatigue-estimating algorithm (FADE) used logged sleep data and was well correlated with the subjective reports of fatigue. Use of fatigue algorithms may be useful to select the timing and crew rest considerations of comercial airline routes before they become part of normal operations.