Contributions of the EU Projects UnCoVerCPS and Enable-S3 to Highly Automated Driving in Conflict Situations

Abstract

Almost 50% of all urban traffic accidents involving personal injury occur in intersection and turnaround area. Left turn through oncoming traffic is challenging for both human and automated vehicles because of a variety of time critical factors to consider. In particular, the position and speed of oncoming and crossing road users and the consideration of their existential, measurement and behavioral uncertainties influence the collision-free outcome of such a risky scenario. Highly automated, connected vehicles can help to reduce the number of accidents in inner-city intersection situations if these uncertainties are taken into account. We describe the integration of a model-predictive planning method, possibilities for modeling and calculating conflict areas based on geometric lane information, the calculation of conflict times based on behavioral predictions of observed road users and the interaction with the driver of the highly automated vehicle. The function is evaluated in field tests for the two EU funded projects Enable-S3 and UnCoVerCPS.