Obstacle criticality evaluation in multi­vehicle scenario

Abstract

Great progress has been done to reduce the severity and frequency of road traffic accidents in the last decades. However, they are still a major cause of casualties worldwide. Advanced driving assistance systems show great potential in further decreasing the number and gravity of traffic accidents. Active safety systems have received a lot of attention in recent years. Passive safety systems, on the other side, have received less focus. The activation of passive safety systems before the crash instant is a promising strategy for the reduction of injuries due to traffic collisions. For this to be possible, sensor data needs to be processed as to detect a crash before it happens. In this context, the present work focusses on the development of a criticality measure for the classification of obstacles in the surroundings of the vehicle regarding the danger they represent. A measure that uses the obstacle’s distance and radial velocity is selected based on previous works regarding similar problems. It is then adapted to better suit the intended application. The resultant measure is called “approach time” and it is discussed in detail in the work.