This internal project investigates methods for adaption of an adaptive restraint system of a passenger car to the accident scenario and the involved occupants. The final goal is a real-time algorithm which is applied at an automotive electronic control unit. The algorithm is divided in the following sub modules: The collision prediction model (ISC-P) calculates collision parameters, which in combination with the dynamic state of the ego-vehicle (vehicle model ISC-V) is input to the collision model (ISC-C). This model predicts the acceleration characteristics of the passenger compartment (crash pulse). The influence of Advanced Driver Assistance Systems is taken into account in the vehicle model. The crash pulse is input to the occupant model (ISC-O), which predicts the injury risk. In a next step the parameters of the adaptive restraint system are optimised with respect to minimisation of injury risk. The project integrates aspects of active and passive safety and aims at further reduction of injury risk in unavoidable accidents.