Behavior analysis of driving simulator students

J.R. Kuipers

Delft University of Technology / Green Dino Virtual realities, Wageningen, The Netherlands

The department Man-Machine Systems of the faculty Mechanical Engineering and Marine Technology started a cooperation with the company Green Dino Virtual Realities for research in the .eld of adaptive training systems. One of the topics is automatic instruction for driving simulators.

Green Dino Virtual Realities developed a sophisticated low cost driving simulator with an adaptive virtual instructor. Goal of the research is to improve the intelligence of the total system and the virtual instructor in specific. To measure the performance increase of the systems intelligence after adjustments we started a 0-measurement of the behavior of driving students. At the start the driving performance data of nearly 4000 students was available. The data analysis focused on the performance of driving students on driving tasks like not exceeding road speed, taking curves, approaching intersections etc. In the analysis the differences between genders got special attention. The results of gender differences in performance on driving tasks by students were compared with the differences in the performance of male and female experienced car drivers. The outcome showed a strong equality between the behaviors of drivers in the simulator as on the road. For instance men perform better on driving task like keeping lane and steering but women perform better on driving tasks like not exceeding the maximum speed and keeping a minimum following distance. These results show that the used simulator is valid alternative for measurement of driver behavior on speci.c driving tasks. Further more it gives more inside information on the performance difference between genders. The results will be used to improve the curriculum and to personalize the instruction regarding to gender differences for example.


Paper presented at Measuring Behavior 2005 , 5th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.

© 2005 Noldus Information Technology bv