An exploratory simulator experiment was performed in which subjects drove through 45 degree turns of different curvature (100 or 200 m) and different directions (left or right) towards an intersection point with the course of a vehicle approaching from the right. There were four possible configurations defined by the respective vehicles' speed and positions at the onset of a trial. These were: a (control) configuration in which the opposing vehicle remained stationary, one in which it would cross the intersection later than the observer's vehicle, one in which a collision would ensue, and one in which the opposing vehicle would cross first, in all cases assuming that neither vehicle changed its speed. Drivers showed that they could discriminate between these configurations, as is apparent from the pattern in the instantaneous decelerations at curve entry, the parameter describing longitudinal control in curves. Moreover, entry decelerations were, on the average, larger in left than in right turns and larger in 100 m than in 200 m curves. The results suggested the hypothesis - to be tested in an independent experiment - that the perceptual cue subjects used as the input to their longitudinal control strategy was the bearing angle at which they saw the opposing vehicle. I.e. subjects appeared to base the amount of deceleration to be applied at curve entry on a purely positional cue.
Abstract