Inexperienced as well as experienced drivers drove in the simulator on a straight freeway. Results were compared with the data of a similar full-scale study wyth an instrumented car on the road. In both experiments the driver task was manipulated by specific instructions for lateral and longitudinal vehicle control. Especially, in lateral vehicle control the simulator appeared to differ from real driving in an absolute as well as a relative way. It seems necessary to measure the dynamics of the simulator and instrumented car accurately and to modify the dynamics of the simulator in closer correspondence with the instrumented car. Only then would it be worthwhile to continue closed-loop (validation) experiments.
Samenvatting