A multivariate investigation of driver performance changes during extended driving periods in a computer-controlled driving simulator

Abstract

An investigation was conducted to determine the effects of driving time on nine continuous measures of driver performance and alertness, and to determine the extent to which such variables could be used to predict driver performance degradations in driving emergencies.

Subjects were required to respond to one of three types of driving emergencies presented after each had driven continuously for periods of 30, 60 or 150 minutes (determined by random group assignment).

The general findings of continuous measures indicated that in driving periods of 60 minutes and longer: (a) drivers exhibited linear increases in a tendency to cause the vehicle to drift laterally (within the lane boundaries) and yaw over time; (b) drivers exhibited significant linear increases in the number of large steering reversals, and significant decreases in the number of small steering reversals over time; and (c) drivers exhibited significant increases in heart beat interval standard deviation and in body movement frequency over time.

In all driving duration conditions, drivers exhibited significant degradations in their ability to respond to emergencies relative to their baseline driving performance levels. However, the amount of degradation did not differ among duration conditions.

Correlational relationships of continuous and emergency performance data, possible methodological problems of previous investigations (in light of present findings), effects of steering gain on continuous and emergency performance, and suggestions of possible future investigations were discussed.