gusts in the manual drive. Heading error is greater immediately following automation failure than following wind gusts in the manual drive during early, intermediate and late phases of driving performance. Thus, performance recovery appears to be better in the manual driving mode than in the automated mode during the first 20-second period (t(39=5.07, p<.001).
In the second study, 11 drivers completed automated and manual drives as in Study 1. The aim of Study 2 was to test whether the impairment in performance found in the automated drive could be reduced or eliminated if drivers were physically engaged in the driving task throughout the drive. Thus, in the second study, drivers in the automated drive were instructed to keep their hands on the steering wheel while the automated system was operational. Study I revealed that automation-induced impairment in lateral control occurred within the first 20 seconds following system failure. Thus, in the second study, drivers' performance was analyzed for 30 seconds following each perturbing event. Figure 3 shows heading error across six 5-second periods for manual and automated driving conditions. The results replicate the findings of Study 1: heading error is greater immediately following automation failure than following wind gusts in the manual drive.
The findings from the subjective state measures in the first study suggest that monitoring an automated driving system may be just as effective in inducing subjective fatigue and stress states as prolonged driving under monotonous driving conditions. At a performance level, the research highlights some of the potential