Item

Abstract

In an earlier study of lane changing with a fixed-based simulator, Wallis et al (1997 Perception 26 Supplement, 100c) demonstrated that the manoeuvre comprises two, roughly antisymmetric phases. Failure to provide visual information midway through the manoeuvre prevents initiation of the second phase, causing the vehicle to career off the road. In a real car, the windscreen and bonnet provide a visual reference frame, and vehicle movement provides both visual and haptic information. To test the importance of these cues we repeated the earlier study using a driving simulator comprising a motion platform, car chassis, and all-round vision.

Subjects drove the length of a two-lane highway and were asked to change lanes to the left or to the right. Lane-changing was conducted either with full vision or in complete darkness. As in the previous study, in the absence of visual information a clear relation between direction of lane change and final heading emerged, owing to a failure to complete the manoeuvre (p < 0.002).

We conclude that subjects have little appreciation of the effect movement of a steering wheel has on their heading, and that they rely heavily on intermittent visual feedback even in the presence of other sensory information.