Traffic and Autonomy

This volume presents a coherent and high-quality collection of contributions addressing key challenges in traffic modeling and control. The chapters showcase a blend of theoretical rigor and practical relevance, with diverse methodologies ranging from PDE control to game theory and data-driven modeling. One chapter presents a retrospective on PDE backstepping techniques for stabilizing stop-and-go waves, extending from single-lane models to more complex multi-lane and multi-class systems. Another introduces a second-order time-to-collision metric that accounts for non-static acceleration and turning, improving predictive accuracy in safety analysis. The CGARZ model simplifies second-order traffic dynamics by collapsing free-flow regimes while preserving detail in congestion, striking a balance between realism and tractability. A multi-population mean field game framework is proposed to model the interaction between human-driven and autonomous vehicles, with a detailed stability analysis that sheds light on the benefits of CAV integration. Another contribution develops a multi-lane model where autonomous vehicles act as mobile bottlenecks, coupling macroscopic flow dynamics with controlled microscopic agents. The volume also includes a rigorous analysis of the Hughes model for pedestrian flow with affine cost functions, exploring uniqueness and sensitivity with respect to both initial data and model parameters.