Driving Simulator and Scenario Effects on Driver Response

Driving simulators are becoming a common research tool for investigating driver behavior in safe, controlled environments. Wide ranges of simulator configurations are available for research—this wide range, as well as and the complexity of the virtual worlds being driven, has not been studied for specific effects on driver response. This study examined driver behaviors in virtual scenarios with varying complexity (difficulty of driving maneuvers) while using driving simulators of different levels of fidelity to assess how complexity and fidelity can affect physiological and subjective mental workload ratings. Participants were equipped with physiological data collection equipment and drove each scenario twice on each simulator; as a baseline drive and again while completing a oneback recitation task. Participants completed NASA-TLX surveys to rate self-reported overall mental demand after each scenario. ANOVA analysis showed that scenario complexity and secondary task were significant factors for all NASA-TLX subscales as well as for driver pupil radius; simulator complexity had no effect. These findings indicate that lower-fidelity simulators may be as useful a research tool as more expensive higher-fidelity machines for physiological and subjective mental workload ratings; a useful finding for research labs without access to highfidelity simulators.