From conflicts to crashes: Simulating macroscopic connected and automated driving vehicle safety

Area of Science:

• Transportation Engineering
• Road Safety
• Autonomous Systems

Background:

• Limited historical safety data for high-level Connected and Autonomous Vehicles (CAVs) necessitates advanced evaluation methods.
• Microscopic simulation offers a viable approach to generate and analyze vehicle trajectory and traffic conflict data.
• Developing robust techniques for analyzing microsimulation conflict data is crucial for road safety applications.

Purpose of the Study:

• To propose and evaluate a safety assessment approach for estimating Connected and Autonomous Vehicle (CAV) crash rates using microsimulation.
• To analyze the impact of different CAV market penetration rates (MPRs) and automation generations on road safety.

Main Methods:

• Modeled the city center of Athens using Aimsun Next software, ensuring calibration and validation with real traffic data.
• Simulated various scenarios with different CAV market penetration rates (MPRs) and two generations of automated vehicles.
• Utilized Surrogate Safety Assessment Model (SSAM) software to identify traffic conflicts and convert them into crash rates.

Main Results:

• Increased CAV market penetration rates (MPRs) correlate with significantly lower overall crash rates.
• Conflicts involving a second-generation CAV as the following vehicle demonstrated reduced crash rates.
• Lane-change conflicts resulted in higher crash rates compared to the lower rates observed for rear-end conflicts.

Conclusions:

• The proposed microsimulation approach effectively estimates CAV crash rates and supports road safety evaluations.
• Higher adoption of CAVs and the deployment of advanced automation generations are key to enhancing road safety.
• Understanding conflict types, such as lane-change versus rear-end, is crucial for targeted safety interventions.