Investigating the accuracy rate of vehicle-vehicle conflicts by LIDAR technology and microsimulation in VISSIM and AIMSUN

Light Detection and Ranging (LiDAR) technology is a remote sensing technique which can be applied to determine the spectral signature and differential position of objects emitting radiation. In order to collect real-time traffic data in a signalized intersection, the LIDAR sensor was installed at Cold Spring Ln – Hillen Rd intersection in Baltimore city, USA. The installed LIDAR sensor can record the “Post Encroachment Time Threshold (PET)” and Time-to-Collision (TTC) indicators as two principal safety measurements between two motorized vehicles (including car-car, car-bus, car-truck, and bus-truck). PET implies a potential danger, while TTC describes an imminent danger. The study aims to investigate the accuracy of obtained PET and TTC from the Surrogate Safety Assessment Model (SSAM), which is a software application developed by the FHWA, and compare it with the results obtained using LIDAR technology. SSAM is a free open-source software to perform statistical analysis of vehicle trajectory data output from microscopic traffic simulation models. Hereupon, the intersection was modeled in VISSIM and AIMSUN, and the outputs of vehicles trajectories by microsimulation were imported to SSAM software to compute a number of surrogate measures of safety for each conflict. The results highlighted that 857, 966, and 959 conflicts were obtained by LIDAR sensor, VISSIM, and AIMSUN respectively in the same time interval. The Root Mean Square Error (RMSE) measure was used for evaluating the accuracy rate, and the result showed that the TTC and PET values by the trajectory of AIMSUN are 34% and 26% more accurate than TTC and PET values by the trajectory of VISSIM, respectively.