Proactive Safety Analysis Using Roadside LiDAR Based Vehicle Trajectory Data: A Study of Rear-End Crashes

This paper presents a methodology to detect rear-end conflicts at signalized intersections with the help of roadside LiDAR sensors. Raw data collected in the point cloud format from the sensors was processed using a series of data processing algorithms to obtain vehicle trajectories. Time-based (MTTC), deceleration-based (SDI), and severity-based (CSI) surrogate safety indices were calculated from the vehicle trajectories to identify the conflict threats at every frame of the dataset, which were further aggregated together to evaluate the risk exposure and risk severity at different temporal segments of the leader/follower car-following period to obtain a rear-end conflict index. The identified conflicts were compared with the historical crash records using negative binomial models. The results indicate correlation between the identified conflicts and the crashes, and further provide new information about the rear-end crash risks at the intersection which could support the proactive approach of traffic safety analysis.

Automated summary

This paper presents a methodology for detecting rear-end conflicts at signalized intersections using LiDAR sensors. Vehicle trajectories were obtained from raw data collected by the sensors and processed using data processing algorithms. Surrogate safety indices were calculated from the trajectories to identify conflict threats and evaluate the risk exposure and severity during different temporal segments. The identified conflicts were compared with historical crash records, showing a correlation and providing new information about rear-end crash risks at intersections.