Identification of black spots on highways using fault tree analysis and vehicle safety boundaries

In previous studies investigating the association between probability of crash occurrence and highway alignment, vehicle skidding, which is an unsafe state of tire cornering, was considered along with tire cornering in crash type classification. In addition, the rear-end crash was often neglected. With regard to the index of each crash type, the safety boundaries of each index, which is the critical value of the index beyond which the corresponding type of crash occurs, were not precise either. To address these issues, this paper presents a method to calculate the average probability of crash occurrence on highway sections based on fault tree analysis (FTA) and an in-depth investigation into vehicle safety boundaries. The proposed method was tested taking an expressway between Bengbu and Nanjing (a section of the G36 expressway) in China as a case study, and the results were validated against the data of crashes occurred on the expressway from December 2005 to February 2017. The results of the proposed method agree better with the crash data than the methods employed in previous studies in terms of goodness of fit and correlation analysis. This method provides a new tool for black-spot identification on highways during the design phase and while in service.