MEASURING PERFORMANCE DURING A FALLBACK PROCEDURE IN AUTONOMOUS VEHICLES

The aim of this paper was to assess the functionality of an experiment that was developed for the DRILL project. The experiment used a computer program and a distracting task to measure a driver’s ability to take control of an autonomous vehicle (fallback performance). Switching theory and tachistoscopic traffic tests served as the foundation for this experiment. The experiment measured the subject's reaction time and the correctness of the reaction. The final sample consisted of N = 48 participants, who were self-selected. The results of the statistical analyses suggest that the experiment does successfully induce and register the switching effect. It confirmed the linear relationship between the results of the experiment and the results of the ATAVT test. Further, it analyzed the reaction time on the item level (e. g. reaction times for an item level depending on its order, the effect of the correctness of the item´s response on the reaction time). However, the small size of the task-switching effect and the unresolved method of evaluation has not yet allowed us to draw a clear conclusion about the functionality of the experiment. Possibilities for improvements to the experiment design are also presented.