Safety analytics for integrating crash frequency and real-time risk modeling for expressways

Area of Science:

• Transportation Engineering
• Traffic Safety Analysis
• Statistical Modeling

Background:

• Crash contributing factor studies have traditionally been conducted independently, analyzing crash frequency and real-time safety risks separately.
• Previous research has not simultaneously analyzed crash frequency and real-time crash risk to determine if integration enhances crash occurrence explanations.

Purpose of the Study:

• To integrate crash frequency and real-time safety analyses using expressway data.
• To develop and compare an integrated Bayesian model with a non-integrated model for traffic safety analysis.

Main Methods:

• A Bayesian integrated model was developed by incorporating the logarithm of estimated expected crash frequency into the real-time safety model.
• A non-integrated model was also developed, independently estimating crash frequency and real-time crash risk.
• Expressway data was utilized for both model estimations.

Main Results:

• The integrated model demonstrated superior performance compared to the non-integrated model for both crash frequency and real-time safety predictions.
• The inclusion of the logarithm of expected crash frequency effectively linked the two models and provided valuable information.
• Novel variables, such as the average daily standard deviation of speed, were identified as positively influencing crash frequency.

Conclusions:

• Integrating crash frequency and real-time safety analyses offers a more robust approach to understanding and predicting traffic crashes.
• The proposed integrated methodology can enhance the accuracy of traffic safety models.
• This study provides a foundation for future research into complex crash mechanisms and improved road safety strategies.