Development of crash prediction models using real time safety surrogate measures

Typical engineering research on traffic safety focuses on identifying either dangerous locations or contributing factors through a post-crash analysis using aggregated traffic flow data and crash records. A recent development of transportation engineering technologies provides ample opportunities to enhance freeway traffic safety using individual vehicular information. However, methodologies on how to utilize and link such technologies to traffic safety analysis have not been thoroughly explored. Moreover, traffic safety research has not benefited from the use of hurdle-type models that treat excessive zeros in crash modeling. This study developed a new crash risk predictor, safe headway distance, to estimate traffic crash likelihood by using individual vehicular information and applying it to basic sections of interstates in Virginia. Individual vehicular data and crash data were used in the development of statistical crash prediction models including hurdle models. The results showed that safe highway distance measure was effective in predicting traffic crash occurrence and the hurdle negative binomial model outperformed other count data models including zero-inflated models.