Multisensory Processing in Simulated Driving

Abstract

Studies that explore integration of visual, auditory or vestibular cues, are derived from stimulus detection and discrimination tasks in which stimuli are selective and controlled. Multisensory processing is not as well understood in more dynamic and realistic tasks such as driving. As visual information is the dominant source of information when controlling a vehicle, we were interested in the contribution of auditory and physical motion (vestibular and proprioceptive) information to vehicle control. The simulated environment consisted of a straight, two-lane road and the task was to drive in the center of the right lane and maintain a constant speed, slowing down for occasional speed bumps. We examined differences in driving performance under four sets of sensory cues: visual only, visual and auditory, visual and physical motion, and visual, auditory and physical motion. The quality of visual information was manipulated across two experiments. In Experiment 1, participants drove in daylight in sunny weather, providing excellent visual information. In Experiment 2, visual information was compromised by providing dark and stormy weather conditions. In both experiments we observed an advantage of multisensory information, an effect that was enhanced when visual information was compromised. Auditory cues were especially effective in improving driver control.