Criticality estimation of Pre-Crash scenarios

The activation of Pre-Crash systems like a reversible seat belt pretensioner can reduce the injury level of an occupant during a collision. The time-to-collision is a common criterion for discriminating between critical and non-critical situations. This paper investigates the driver´s behavior as well as his/her criticality estimation of Pre-Crash scenarios and models the criticality estimation to a classification with features of the situation. Experimental data is analyzed regarding its possibility to model the subjective driver estimation by measured features. A situation model is developed that predicts a classification criterion, which can be used for Pre-Crash activation functions.