Screening Road Networks for Hazardous Locations – Addressing the Local Roads Issue

The first and most important step in road safety management is the identification of roads with safety needs. The most prevalent road screening technique utilizes the quality control method. Exposure data is used to estimate the expected number of crashes, which is then compared to the actual number of crashes in order to identify locations with safety needs. This approach must be modified when applied to county and city roads because of the lack of such data. This paper proposes a screening method suitable for local roads. The typical number of crashes is based on (1) segment connectivity with busier state-administered roads and (2) characteristics of the land development in the area. Two alternative models are developed and compared: one uses Negative Binomial regression with random effects and the second uses a classification tree. The developed models are discussed and their performance compared to point out the more promising one. Then, the selected prediction method is applied to identify Indiana county roads with safety needs.