Microscopic Analysis of the Impact of Congestion on Traffic Safety and Efficiency at a Signalized Intersection: A Case Study

Traffic congestion is a major issue in urban areas around the world, causing significant delays, economic losses, and environmental pollution. Congestion can also have a negative impact on traffic safety and efficiency, especially at signalized intersections where traffic flows intersect and interact. In this work, we analyze the trajectory data of road users at an urban intersection in Germany and compare their behavior during a 2-hour period of traffic congestion with four weeks of normal traffic conditions. Our evaluation of safety shows that in 9 of 33 crossing traffic flows the minimum Post-Encroachment Time decreases from more than 3 s to less than 3 s during congestion implying an increase in the level of risk. To assess the impact on efficiency, we examine two key parameters: traffic volume and traffic delay. A z-score of -0.04 for the traffic volume at the entire intersection indicates a minimal impact of congestion. In contrast, the z-score of -1.22 for the traffic volume at the congested exit highlights a considerably higher impact of congestion at this specific location. Both negative z-scores collectively signify an adverse effect of congestion on traffic efficiency. This finding is further supported by a significant increase in the mean traffic delay, which increases from 19 s to 35 s across the entire intersection during congestion. Furthermore, we propose a novel methodology to classify road users' behavior during congestion at intersections as either "good" or "bad" using trajectory data. Application of this methodology demonstrates that depending on the route 40-50% of traffic participants exhibit bad behavior during periods of congestion. The implications of our findings are significant for traffic management and engineering and can advance the development of strategies to improve traffic safety and efficiency at signalized intersections.