On Threat Assessment and Decision-Making for Avoiding Automotive Vehicle Collisions

Road traffic accidents are one of the world’s largest public health problems. In the EU alone, traffic accidents cause approximately 1.8 million injuries and 43.000 fatalities each year. This thesis is concerned with the development of in-vehicle systems that can detect hazardous traffic situations and assist drivers in avoiding or mitigating accidents.An overview is given of different types of accidents and measures that are taken to reduce the number of accidents and their consequences. From this overview, certain types of accidents have been selected to be addressed in this research project. The contribution of this thesis is a number of algorithms that can assess traffic situations and make decisions to actively assist the driver in avoiding or mitigating these accident types.The approach that is used for making decisions on when and how to assist the driver, is to first estimate how a collision can be avoided by the driver. Secondly, the brakes are applied autonomously if hard braking is the only option to avoid or mitigate an accident. The algorithms proposed in this thesis are capable of estimating how collisions can be avoided in any type of collision scenario, such as rear-end collisions andintersection collisions. These algorithms can be used to avoid or mitigate collisions with all types of road users, such as pedestrians, cyclists and other vehicles.The algorithms have been evaluated using data from both real traffic conditions and collision situations on a test track. The results show that the algorithms can improve the performance of conventional rear-end collision avoidance systems, without significantly increasing the risk of unnecessary braking