Classification of Flight Segment Using Pilot and WSO Physiological Data

Abstract

To develop on-line metrics of operator workload, physiological measures were recorded from 10 F4 flight crews during air-to-ground training missions. Discriminant analysis techniques were used to classify eight flight segments for nineteen crew members using heart rate and eye blink data as variables. For the pilots 93% of the flight segments were correctly classified and 89% were correctly classified for WSOs. The advantage of collecting multiple physiological signals was demonstrated. Further, combined cardiac and eye blink data produced better classifications than when each was used alone. Application of this technique to the flight environment for use to evaluate mission/design effects and to provide for on-line monitoring of operator state is discussed.