The need for consumer information on car safety is growing steadily. Up to now, the consumer information about crash tested cars has been provided, but this should be supplemented by safety ratings based on real world accidents which reflect all types of accident circumstances. The possibilities and limits of the necessary "Quality Criteria for the Safety Assessment of Cars based on Real-World Crashes" are analysed in an EU project in co-operation with research institutes, industry and universities. The paper summarises the results of three major objectives within this "SARAC" project: (1) The existing rating procedures world wide have been analysed and are described by their methods and the assessment criteria/crash parameters used. The output results of the different ratings are compared with respect to the correspondence of the rankings produced by each method; (2) The interrelationship between prospective (for example EuroNCAP) and retrospective rating systems is being investigated. Results from Australia, US and Europe represent the first experiences available, suggesting it is likely that correlations can be established; and (3) The paper finally outlines possibilities to integrate the aspects of crash compatibility and vehicle aggressivity in a comprehensive overall rating system. The methods for measuring aggressivity have been tested on databases and the necessary parameters are outlined. Possibilities for an improved international comparison of safety rating results based on real accidents and necessary future steps are discussed.
Abstract