Beca is developing a new generation of crash prediction models for rural roads based on research to better understand the relationship between rural road crashes and roadside hazards.
The research aims to fill a gap in industry knowledge. Current models and crash rates are somewhat deficient in that they do not include all the key factors that influence crash occurrence. The typical approach for the identification of road safety problems uses analysis of historical crash patterns. This approach is not always sufficient to enable a suitable diagnosis of road safety deficiencies, especially for low volume roads where the scarcity of crashes may mask the potential to reduce the risk of crashes. Crash prediction models can be developed to better evaluate crash risk and identify the benefits of road improvement treatments to target and reduce this risk.
Beca's research involves a pilot study in which both manual inventory data and video data is collected for a sample of rural roads. The purpose of the pilot study is to establish the sample size for the main study to follow later and develop some preliminary crash prediction models. A key innovation is the testing of video based analysis methodology. Findings from this research will be used to develop crash prediction models that will address rural road safety and provide policymakers in the Ministry of Transport and NZ Transport Agency with tools to evaluate road safety investment.