Driving tasks may be represented at three hierarchical levels: navigation, guidance and stabilisation. One goal of safety research is to enable drivers to dedicate more attention to high levels of driving behaviour. Known simulator studies have mainly investigated skill-based performance. The key issue is validity of tasks and test results. Three types of validity may be distinguished: content validity, construct validity and criterion-related validity. Human factor requirements demand that outside forward view be presented in real time. For vehicle handling tests 60 Hz is appropriate. Screen resolution should be 1M pixels. Field of view should cover 160 degrees x 60 degrees horizontally and vertically, and resolution in two dimensions should be identical. Drivers can be asked to rate the representation of real world provided by the simulator. Validity is a key factor in simulation. Since there is a complete lack of consequence of driver actions (danger), some other means of stressing realistic behaviour and motivation, e.g. incentives, must be introduced.
Abstract