The aim of the research presented here is to assess what benefits may be derived from new devices in terms of accidents avoidance. We focused on the driving aids specially designed to improve safety.

A first research project, based on in-depth accident analyses, showed that in most accidents a driver "weakness" could be identified in the behavioural chain. We determined which were the needs corresponding to every class of situations, and hence what driving aids would be likely to reduce or eradicate the accidents concemed. This qualitative approach allowed us to point out the major categories of driving situations, accidents, needs and potential aids, but had to be corroborated by statistical analyses based on national accident records.

Thus, we extended this type of approach to a representative sample of police reports, tried to make a quick reconstruction of the accident, and attempted to determine in which way various driving aids could have been efficient. We worked on 350 accidents, and then we added the results of the analyses to a computer file which already contained the description of these accidents. By appropriate means (correspondence analysis), we specified the efficiency of each driving aid, in each situation and condition, etc... in order to determine what is globally at stake, but also to get an idea of more precise targets:​ that is driving aids particularly fitted to specific problems. A typology of accidents, drivers' needs and efficient aids is analysed. Conclusions are drawn about devices that appear the most promising.