Defining and validating relationships between surrogate safety measures and road crashes (R-3161)

Road accidents constitute a huge economic loss to society. The annual estimated cost in Belgium amounts 12,5 billion euro, or 4,6% of the GDP. The traditional way of studying traffic safety makes use of accident data. However, this method has many disadvantages, like random variation related to the rare nature of crashes, underreporting and an ethical problem. Therefore there is a need for surrogate or complementary methods, using information about serious conflicts and normal interactive behavior. However, these methods currently still suffer from reliability and validity problems because of the use of subjective and suboptimal indicators. My research consists of three main steps. First, different types of conflicts are defined, quantitative continuous measures are implemented for each type and relationships with environmental characteristics are explored. In the second step, a framework based on taxonomic analysis is developed to subdivide driving behavior to identify critical sub-behaviors that lead to dangerous situations. The third step combines preconditions of a safe road environment with this framework and a model of human information processing, allowing the analysis and validation of the relations between the three levels of traffic situations (accidents, serious conflicts and interactive behavior).

Keywords:Traffic safety

Disciplines:Economics and business