Stop-and-Go Wave Propagation in Real Highway Traffic: Insights from Microscopic Trajectories and Macroscopic Sensor Data

Traffic congestion is a complex phenomenon that displays wave-like behavior where even a clear road can be disrupted by the actions of a single driver, causing the formation of stop-and-go waves. Studying these unprecedented hurdles is necessary to understand traffic dynamics, improve AI-based traffic management systems, and enhance overall efficiency in transportation. This study analyzes data that uses real-world highway traffic at an urban city using METR-LA sensor data and NGSIM vehicle trajectories to compute stop-and-go wave propagation. Looking at each car, the speeds and distances between vehicles are analyzed with the principles of statistical mechanics, revealing regular patterns in collective traffic behavior. Speed variations in car platoons tend to grow as they spread in a non-linear fashion, just like chaotic dynamics found in other complex systems (“Butterfly effect”). The results, combining wave theory and statistical mechanics to understand and model the traffic, provide meaningful information that could help both traffic management and future physics-based studies of the same or similar complex systems.