Assessing Safety Outcomes Related to Driver Visibility and Behaviour

摘要

Assessing the impact of vehicle safety systems designed to reduce the likelihood of crashes and near-crashes is difficult because crashes, thankfully, are infrequent, and near-crashes are hardly ever reported. Clearly, valid surrogate measures for driving safety are needed. Two efforts are described in the present paper to more clearly and definitively provide useful surrogate measures for evaluating whether reductions in crash risk are indeed associated with vehicle lighting systems. In one effort, a "transfer function" relating the magnitude of improvement in visual performance to the magnitude of observed reductions in crash frequencies is described. In another effort, driver behaviour, vehicle status data and real-time photometric data were combined to identify whether certain behaviours associated consistently with roadway locations having high-risk crash histories, were also found when drivers were exposed to high levels of oncoming headlamp illumination. The results of these efforts can help point toward meaningful quantification of the safety benefit of vehicle lighting.