Early Versus Late Traffic Management for Autonomous Agents
This dissertation investigates how the timing of centralized traffic management affects the performance of autonomous agents in autonomous vehicle and robotic fleets.
Overview
As autonomous vehicles and robotic fleets become a reality, safely managing how they navigate constrained spaces like intersections is a critical challenge. This dissertation investigates how the timing of centralized traffic management affects the performance of these autonomous agents. The study focuses on an early-versus-late intervention framework in which a predefined control zone determines when agents become subject to centralized coordination.
Utilizing Mixed-Integer Linear Programming (MILP) and Model Predictive Control (MPC), a framework is developed to coordinate agents safely while strictly satisfying operational constraints. This work evaluates how intervention timing influences system delay, throughput, controller resource utilization, platoon formation, and crossing-order fairness. The results provide critical insight into when earlier intervention improves system performance, when expanded control distances lead to diminishing returns, and how fairness constraints reshape the balance between efficiency and service-order preservation. Although motivated by autonomous traffic management, the proposed framework is widely applicable to other multi-agent systems, including urban air mobility and industrial robotic fleets.
Presenters
Brief Biography
Salman Ghori is a Ph.D. candidate in Electrical and Computer Engineering at King Abdullah University of Science and Technology (KAUST), working in the Aerospace and Transportation Systems research group under the supervision of Professor Eric Feron. Before joining KAUST, he earned a master’s degree in Aerospace Engineering from Sapienza University of Rome and worked as a Senior Engineer at ZF Friedrichshafen AG.
