Investigating safety performance of the SAFESTAR system for route-based curve treatment

In order to improve curve safety, warning signs are used to alert drivers to unexpected changes in horizontal alignment and speed variations. While all countries have standards and guidelines for signs, markings and delineations to ensure a uniform system, they do not use them in the same way and choices and combinations may differ. SAFESTAR project promoted a method for establishing uniform curve signs on two lane rural roads to force drivers to adjust their speed to the actual risk. Although widely applied around the world and producing speed reduction benefits, as evidenced in the literature, the reliability in terms of statistical safety significance of the method has not been well established. The goal of this paper is to assess the safety fundamentals of SAFESTAR's risk ranking system and safety performance of the proposed signing schemes. Crash data and road characteristics from two Polish regions are used to carry out sound statistical analysis. Building on data from treated and untreated sites, the analysis first looks at relative changes in Crash Rates for different risk categories. Next, Crash Modification Factors are estimated. Results are interesting and suggest that system needs revision, both in terms of risk classification and on how treatments are implemented.