After having reconstructed a traffic accident on basis of the evidence (traces, final position and damage of the vehicle) in consideration of physical circumstances, there are two questions which remain open: On one side, the question of cause (for example, a driving mistake or technical failure), on the other side, if the accident could have been prevented.

Until now, assumptions were made based on plausibility and experience. Nowadays, however, physiological sequences of the viewing process are very well examined and extensively depicted in literature. Former attempts to apply this knowledge have failed because only in exceptional cases closed solutions are possible; graphical methods dash against the required level of preciseness.

The target of this paper is to discuss the question of how and when a driver can recognize a hazard on the road. Answering this question is very important for accident reconstructions since it may provide details of the cause and preventability of a collision. The paper addresses human perception, especially the visual perception of motion. It tries to establish an objective criterion of analytically locating the point where a potential danger could have been avoided. In order to do this, one has to empirically find the lower threshold value for the angular change of the angle of vision at a given distance, a value that has to be reached to enable the brain to detect the relative motion of another vehicle and, potentially, a hazard.

(EVU-members can download the full article)