Pervasive Sensing to Correlate Vehicle Driving Behavior With City-Scale Traffic Dynamics

Individual driving behavior is a pivotal element that shapes the overall traffic dynamics in a city. In this work, we study and analyze the complex web of relationships between individual driving behaviors and their impact on the overall traffic dynamics of a smart city with two primary objectives: first, understanding the spatial interaction between individual vehicles and their impact on each other, and second, finding anomalous driving behaviors, which lead to congestion and traffic incidents. Specifically, we introduce an overarching modular framework investigating human factors of driver characteristics, vehicle attributes, geographical terrain surrounding the road infrastructure, and environmental conditions. Analyzing driving maneuvers considering such diverse factors emphasizes the removal of the bias in driving behavior prediction. Our study serves as an essential foundation to delve into the problem of traffic dynamics in a city and further motivates the need for a deeper understanding of the correlation between driving behavior and traffic management.