Measuring Behavior 2005: Williams

Analysis of driver behavior at crossroads in urban area

M. Williams

Institute of Ergonomics, Darmstadt University of Technology, Darmstadt, Germany

The aim of the experiment was to analyze the driver behavior at different types of crossroads in order to find out if there are differences between different groups of subjects and if the drivers are confronted with different cognitive and motor demands which depend on the type of crossroad. Therefore, two types of crossroad in urban area have been considered: busy crossroad (minor road/major road with high traffic density) where the driver has to give way and crossroad with traffic light. In both cases the test subjects had to approach and pass the crossroad. Thirty-six persons participated in the experiment; the test car have been equipped with video cameras for behavioral data collection: one camera for the frontal view, one for the rear scenery and one for the eye glances. For analyzing driver behavior The Observer^® (Noldus Information Technology bv, The Netherlands) has been used. A configuration concerning the relevant behavioral classes had to be implemented in the beginning. The behavioral classes were: driving task, eye glances, head movements, and torso movements. Every behavioral task contains several elements. The subjective perceived strain has been rated using NASA-TLX ZEIS with 6 dimensions (physical demands, mental demands, time pressure, effort, performance, and frustration). Differences have been found between the behavior of different groups of subjects, e.g. the test subjects with less experience showed the tendency of using all information sources from the environment and experienced higher strain levels than the other subjects. Besides, the results of our study con.rmed the assumption that these two types of crossroads pose high cognitive demands to the driver, while the motor demands are high for the busy crossroad and rather low for the crossroad with trafic light.

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