Our existing countermeasures are based on an underlying strategy of separating drinking from driving through behavioural controls. Consequently, they are person-centred rather than accident-centred;​ little consideration has been given to the potential of measures specific to particular types of alcohol-related accident.

Three in every four alcohol-related accidents occur at night, one third of them on rural roads. Of these, the vast majority (80 per cent) involve a single vehicle which loses control. After controlling for exposure, the major difference between alcohol-related and 'sober' run-off-road accidents is that the former occur significantly more often on curves. Thus, 'run-off-road-oncurve' accidents constitute an important target for sepcific countermeasure action.

Previous in-depth accident research suggests that the majority of alcohol-related accidents can be traced to failures, of one form or another, in visual perception. Previous research on curve negotiation has identified perception of curvature and, through it, course and speed selection as being of paramount importance. At night, the natural cues to the presence and nature of a curve are severely degrated and the effects of alcohol compound the perceptual difficulties faced. One possible approach, therefore, is to provide additional information on curvature through special delineation treatments. It is hypothesised that, if the signals are sufficiently potent, such treatments will sufficiently offset the perceptual deficits of the drinking driver as to significantly improve his curve negotiation performance