Driving behavior in a real and a simulated road tunnel--a validation study.

The aim was to validate driving behaviour in a simulated road tunnel. Speed and lateral position of 20 subjects were measured in a real tunnel and in the same tunnel implemented in the VTI driving simulator. In both situations a left-hand steered and manually geared passenger car was used. Driving speed was higher in the simulated tunnel than in the real tunnel. Elimination of speed information from the speedometer caused a similar small speed increase in both situations. Also, the difference in speed between driving lanes was similar in both cases. The effects on speed variation were similar to that for speed level. Regarding lateral position, subjects positioned the car somewhat further away from the nearest tunnel wall in the real tunnel than in the simulated tunnel. In both situations the distance to the nearest wall was greater when it was located to the left of the driver than on the opposite side. Lateral position deviation was about the same when the road was straight, but in a curved section it was somewhat greater for the simulated tunnel. It is concluded that behavioural validity in absolute terms was not quite satisfactory, especially regarding choice of speed, whereas relative validity was good for both speed and lateral position.