Abstract

The purpose of the present study was to compare driving performance, subjective mental workload levels, and usability evaluation when driving conditions(normal navigation vs. front-to-rear end collision situation) and the information modality of in-vehicle navigation system(visual, auditory, and audio-visual) were varied. Driving performances were measured both by longitudinal(variabilities in accelerator input, driving speed, and brake input) and lateral vehicle controls(variabilities in lane position and steering wheel rate). NASA-TLX, which was adopted to measure subjective mental workload, and the drivers' subjective evaluation for the system usability were also measured. The results of Experiment 1 where normal and navigational driving was required can be summarized as the followings: (1) The drivers showed smaller lane position variability when the navigational information was provided through auditory or audio--visual modality than through visual modality. But (2) the other driving performance measures and subjective workload ratings were not significantly different among the types of information. And (3) the auditory or the audio-visual navigation systems were evaluated to be more usable than the visual system In Experiment 2, collision situation was simulated by manipulating the speed of the vehicle which the drivers were asked to follow. The results of the Experiment 2 showed the followings: (1) Time to initial and fun brake application after the warnings were shorter when the information was presented auditory or audio-visually than visually. (2) Subjectively reported workload levels were lowest in the auditory condition, bur: (3) the usability evaluation did not differ among the three types of warning. Finally, implication and issues related to this study were discussed.