Towards modeling situation-specific driving behaviour – an element of Virtual Vehicle’s Foresight Safety Concept

Authors

B. Brandstätter, E. Santuccio, P. Pretto - Virtual Vehicle Reasearch GmbH

Summary

Driving is a complex task that requires several skills, such as vehicle control, trajectory planning, collision avoidance and interaction with other road users. The interaction among drivers happens through explicit, e.g., turn signals, and implicit signals, e.g., a swaying motion. Nevertheless, understanding the trajectories of other drivers has al-ways been a challenge and foreseeing them became a need. The scope of this re-search is to identify and estimate the influence of some cognitive conditions on the driving behaviour. Indeed, the current driving behaviour can reflect different physical and cognitive driver´s states, e.g., stress, distraction, fatigue. Specifically, we investigate the dependencies of driving behaviour from attentive vs. distracted state, as well as regular vs. extremely careful or aggressive driving attitude. The experiment is conducted in a realistic static driving simulator with an immersive projection system. The driving scenario consists of three sections: highway merging, highway driving and roundabout merging. Each driver is required to drive in all conditions: “Regular” – participants freely drive according to their skills; “Verbally distracted” – participants while driving talk with the experimenter; “Visually distracted” – participants while driving per-form in parallel a secondary task, i.e., tapping on screen; “Careful” – participants are asked to be extremely careful while driving; “Aggressive” – participants are asked to perform the scenario under time pressure. The results will contribute to the realization of an adaptable driver´s behavioural model, that will aim to the concept of foresight safety®, i.e., a system that will promptly inform the driver about foreseen trajectories of other vehicles from inferred driving behaviour.