In the area of safety-oriented driver assistance systems there is a trendto increase the accident mitigation capabilities by adding or strengthening autonomous system reactions. However, this also increases the potentialfor involuntary accidents in the case of malfunction. Due to product liability regulations these high risk functions require an increased development effort as well as more reliable sensor platforms, which drive up their costs. The accident mitigation capabilities of autonomously acting systemscan also be achieved by an alternative strategy avoiding the high risk system reactions. The key is an early and reliable warning giving the drivertime to react to the situation, combined with functions supporting the driver in his reactions, e.g. emergency braking. Early system reactions withlow false activation rates can only be achieved by an advanced understanding of the traffic situation and an interpretation of the driver's actionsin this context. To achieve this, the traditional approach of assessing the criticality of one potential collision object is extended towards observing and assessing multi-object scenarios. An analysis of accident statistics shows that in a high percentage of accidents the multi-object constellation provides additional information enabling early criticality assessments of the traffic situation. Using this information, the driver can be supported in an optimal way by an early, low-risk system reaction. This approach is the key for the vision "safety for everybody", i.e. providing cost-effective collision mitigation functions with high collision mitigation capabilities to the mass market. The full text of this paper may be found at: http://www-nrd.nhtsa.dot.gov/pdf/esv/esv21/09-0153.pdf For the covering abstract see ITRD E145407.
Abstract