Exploring congestion phase transitions in vehicular traffic via topology and driver behavior modeling

Traffic congestion is a world-wide problem with far-reaching consequences. Due to road topology as well as driver behavior, extreme degrees of congestion can occur. Just as in biological swarms, stochastic properties of individual agents contribute to the emergence of macroscopic network properties. In this paper we apply approaches for swarm behavior to road traffic, expanding them to include road topology and driver behavior. By varying traffic density, we study the phase transitions arising in traffic when moving from a free-flow to congested state. Ideally, understanding how phase transitions occur in realistic traffic will allow us to identify early signs and prevent such transitions from happening in real traffic, before the critical point is reached.