484 Crash risk identification at unsignalized intersections within the internet of vehicles

<h3>Background</h3> Accidents at intersections are common in developing countries like India. In addition, most intersections in India are unsignalized and have heterogeneous traffic, making vehicular operations more difficult due to merging and crossing traffic of multiple vehicle types. Over the past few years, the rapid advancements in sensor and wireless communication technologies have strongly driven the evolution of sophisticated cooperative driving systems, leading to the emergence of the Internet of Vehicles (IoV). <h3>Objective</h3> There are two-fold objectives in this study. The first is to investigate the traffic conflicts using Time to Collision (TTC) progression. Furthermore, if there are more observations about the potential conflict and the risk exposure lasts longer, the situation becomes riskier. The second objective is to come up with a conflict measure for evaluating traffic safety at every position at unsignalized intersections. <h3>Methods</h3> Initially, the combination of UAV technology for data collection and the AI Traffic video processing tool (DataFromSky) was used for visual tracking with accurate spatial and temporal traffic characteristics. By gathering data on vehicle trajectories at intersections, we have created a model to assess collision probability, which mirrors the degree of vehicular conflict occurring at these junctions. <h3>Results</h3> The proposed collision probability model effectively assesses the risk of accidents for vehicles at every position within the intersection. Numerical results show the high precision of our suggested model in terms of risk recognition when evaluating the collision probability at the study intersection. <h3>Conclusions</h3> The study methodology autonomously establishes conflict measures for traffic safety inspection. Either the driver or the autonomous control system can use the calculated collision probability to decide on subsequent actions to prevent potential accidents. Therefore, collaborative communication between vehicles can significantly enhance road safety within the IoV.