According to reconstructions drivers had the chance of avoiding an accident by choosing the right manoeuvre in 35% of cases. But is it possible to talk of choice when time is so limited? Emergency manoeuvres undertaken by drivers involved in an accident seem for the most part to be ill-advised or in more favourable cases badly performed. At junctions for example it is noticeable that lateral swerving is underused in comparison with emergency braking. It is nevertheless doubtful whether a greater use of swerving as a result of training in particular will achieve greater safety, for the consequences resulting from failed manoeuvres have also to be taken into account. Even insufficient braking is likely to reduce the seriousness of an accident whereas this is not true of swerving. An experiment was undertaken in order to determine whether drivers were able to assess to what extent these two manoeuvres were possible without carrying them through for safety reasons and also to eliminate the learning factor. It was shown that the advantages of swerving were as a whole well perceived but that the judgements made showed a great deal of variation between and within experimental subjects which leads us to believe that they were not trained to rely on this type of judgement in order to react to an emergency situation. Another laboratory experiment, showed that individuals placed in simulated emergency situations preferred and justified in most cases a two- stage strategy, involving firstly braking even if it meant undertaking a second action (swerving in front or behind an obstacle) if their judgement of the situation made it seem necessary. It is certainly strategies of this type which are often noticed in accidents, the initial braking generally leads to a locking of the wheels which makes any subsequent action on the steering wheel ineffective. Swerving on its own only happens when visibility and the space available are crucial or on the contrary as a manoeuvre of last resort when braking is obviously no longer possible. This indicates that training drivers to swerve may not be effective.
Abstract