Most road safety studies rely on summary measures of exposure. The term summary measure denotes any aggregate indicator of exposure that does not directly identify and count the number of opportunities for accidents to occur. This paper shows how elementary units of exposure can be developed onthe basis of known aggregate measures, such as Annual Average Dailty Traffic (AADT). An elementary unit of exposure refers to any event that generates an opportunity for an accident to occur. Four such events are identified: (1) Encounters, i.e. vehicles passing each other in opposite directions of travel; (2) Simultaneous arrivals at points of intersection between potentially conflicting directions of travel, in particular vehicles entering intersections at the same time or within a very short time interval; (3) Change of travel lane on multi-lane highways; (4) Braking or stopping. These events describe traffic movements prior to a wide range of crash types. The only major group of accidents that is not directly related to particular events is running off the road, for which AADT can serve as an indicator of exposure. The number of events expected to occur for each of the four types identified is estimated by relying on the assumptions: (A) that AADT is known; (B) that vehicles or road users arrive at a point of potential conflict according to a Poisson process, and (C) that simultaneous arrivals within a very short time interval (between 1 second and 5 seconds) have the potential for generating a conflict. It is found that the number of events creating opportunities for conflicts increases considerably faster than AADT; as a rule of thumb the number of events equals AADT2. It is furthermore suggested that the rate of accidents per event-based unit of exposure declines considerably as the number of events increases, as a result of the fact that road users learn how to identify and control the hazards involved in each type of event.
Abstract