Pilot workload prediction

Abstract

A predicting model of pilot workload is developed using a time-based algorithm, work-load values from previous research, and experimental data obtained by a group of experienced pilots on a Singer-Link Gat-1 instrument trainer with three degrees of motion (roll, pitch, and yaw). Each pilot performed three experimental flights presented in a counterbalanced order; each flight consisted of short, medium, or long cruise and initial approach segments. Results strongly suggest that pilots were more sensitive to the rate at which work was done than to the total amount of work accomplished. The result of predictions obtained with the model showed that the time-weighted average of the component work-load ratings were able to predict the obtained work-load ratings accurately.