Benefits of Sound for Driving Simulation:
An Experimental Evaluation
(UMTRI-95-16)

Brian T. Davis, Paul Green

December, 1995

Sixteen drivers (eight under 30, eight over 60) drove a moderate fidelity driving simulator at approximately 45 mi/hr while performing a three-choice response-time task simulating use of an in-vehicle system. All driving was on two-lane roads with no traffic. There were no headwinds or any variations in drag, so maintaining a constant speed was extremely easy. Driving occurred under three conditions--no sound, speed-related sounds only (engine, drive train, and wind noise), and all sounds (speed sounds plus tire squeal and shoulder sounds).

The type of sounds provided (or their absence) had no significant effect on response time. In terms of driving performance, providing sound caused drivers to decrease their mean speed by 1-2 mi/hr and to reduce the variability in their speed, though these differences were not statistically significant. Providing sound had no significant effect on either mean (or standard deviation) of lateral position, though subjects tended to drive closer to the center of the lane when shoulder sound was provided. In contrast, conditions in which no sound was provided were rated as just as realistic as when sound was provided.

These data suggest that the benefits of speed-related sounds are relatively small for nonchallenging human-performance experiments conducted in driving simulators where speed maintenance is of interest. The development of high-fidelity sound systems cannot be justified for that purpose.

Close This Window