Advancing Urban Air Mobility: Strategic, Tactical, and Human-Centered Approaches to Traffic Management
Open Access DepositedUrban Air Mobility (UAM) offers a transformative solution for urban transportation, enabling safe and rapid air travel. However, its integration faces unprecedented regulatory and operational challenges. Existing airspace frameworks are ill-equipped for high-density urban operations, requiring stringent safety measures to manage eVTOL aircraft over populated areas. Meanwhile, commercial viability depends on scaling traffic throughput and reducing costs to make UAM services affordable and accessible. Addressing these challenges demands a focus on both regulatory adaptation and operational efficiency. To address these challenges, this dissertation presents a comprehensive traffic management system for UAM that balances safety and efficiency through increased autonomy in both vertiport and air traffic management systems. Central to this system is an integrated framework combining strategic conflict management with tactical deconfliction to mitigate risks and optimize performance. The strategic conflict management is embodied in the demand capacity balancing (DCB) algorithm, a mixed-integer linear programming (MILP)-based method that computes optimal departure times via ground delay to ensure traffic volumes around bottlenecks remain within safe thresholds. Recognizing that speed adjustments in tactical deconfliction can introduce arrival time uncertainties, I introduce the rolling-horizon DCB approach, which incorporates a learning-based estimated time of arrival estimator to dynamically adjust departure scheduling. This refinement reduces conflicts and enhances overall efficiency compared to static, single-round planning. On the other hand, a multi-agent reinforcement learning (MARL)-based tactical deconfliction method provides real-time speed advisories to maintain safe separations between aircraft. To improve adaptability across diverse operational scenarios, I propose a large language model (LLM)--augmented MARL framework that dynamically adjusts reward structures, enabling more efficient resolution of conflicts under varying traffic conditions and different human preferences. The integration of strategic and tactical conflict management is critical for scalable and efficient UAM operations. Strategic DCB preconditions traffic flow, ensuring that tactical deconfliction remains effective, while advanced tactical deconfliction increases the effective capacity of DCB, reduces ground delays, and enhances overall operational efficiency. A well-coordinated balance between these layers prevents excessive delays from rigid strategic planning and mitigates last-minute conflicts from purely reactive control, providing a robust and adaptive framework for high-density UAM operations. To bridge the gap between theoretical development and practical implementation, I also introduce the Vertiport Human-Automation Teaming Toolbox ( V-HATT). This framework integrates human operators with automated systems for scheduling and real-time control within terminal airspace and vertiports, which are among the primary bottlenecks in UAM operations. V-HATT addresses this challenge by providing a detailed decision-support tool for optimizing vertiport capacity management, ensuring efficient turnaround times, and reducing operational delays. It leverages advanced optimization techniques for arrival and departure coordination, a co-simulation architecture for airspace and taxiway deconfliction, and a vertiport management interface for effective human-machine teaming. Together, these contributions promise significant improvements in the efficiency, safety, and reliability of UAM operations, thereby supporting the sustainable expansion of urban air transportation systems.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.