A novel method for imminent crash prediction and prevention

Abstract A crash prediction and prevention method was proposed to detect imminent crash risk and help recommend traffic control strategies to prevent crashes. The method consists of two modules, the crash prediction module and the crash prevention module. The crash prediction module detects crash-prone conditions when the predicted crash probability exceeds a specified threshold. Then the crash prevention module would simulate the safety effect of traffic control alternatives and recommend the optimal one. The proposed method was demonstrated in a case study with variable speed limit (VSL). Results showed that the proposed crash prediction and prevention method could effectively detect crash-prone conditions and evaluate the safety and mobility impacts of various safety countermeasures.