UNMANNED AERIAL VEHICLES: TRAJECTORY PLANNING AND ROUTING IN THE ERA OF ADVANCED AIR MOBILITY

Advanced air mobility (AAM) is a revolutionary concept that enables on-demand air mobility, cargo delivery, and emergency services via an integrated and connected multimodal transportation network. In the era of AAM, unmanned aerial vehicles are envisioned as the primary tool for transporting people and cargo from point A to point B. This thesis focuses on the development of a core decision-making engine for strategic vehicle routing and trajectory planning of autonomous vehicles (AVs) with the goal of enhancing the system-wide safety, efficiency, and scalability. Part I of the thesis addresses the routing and coordination of a drone-truck pairing, where the drone travels to multiple locations to perform specified observation tasks and rendezvous periodically with the truck to swap its batteries. Drones, as an alternative mode of transportation, have advantages in terms of lower costs, better service, or the potential to provide new services that were previously not possible. Typically, those services involve routing a fleet of drones to meet specific demands. Despite the potential benefits, the drone has a natural limitation on the flight range due to its battery capacity. As a result, enabling the combination of a drone with a ground vehicle, which can serve as a mobile charging platform for the drone, is an important opportunity for practical impact and research challenges. We first propose a Mixed Integer Quadratically-constrained Programming driven by critical operational constraints. Given the NP-hard nature of the so called Nested-VRP, we analyze the complexity of the MIQCP model and propose both enhanced exact approach and efficient heuristic for solving the Nested-VRP model. We envision that this framework will facilitate the planning and operations of combined drone-truck missions and further improve the scalability and efficiency of the AAM system. Part II of the thesis focuses on the survivability reasoning and trajectory planning of UAVs under uncertainty. Maintaining the survivability of an UAV requires that it precisely perceives and transitions between safe states in the airspace. We first propose a methodology to construct a survivability map for an UAV as a function of the vehicle's maneuverability, remaining lifetime, availability of landing sites, and the volume of air traffic. The issue of trajectory planning under uncertainty has received a lot of attention in the robotics and control communities. Traditional trajectory planning approaches rely primarily on the premise that the uncertainty of dynamic obstacles is either bounded or can be statistically modeled. This is not the case in the urban environment, where the sources of uncertainty are diverse, and their uncertain behavior is typically unpredictable, making precise modeling impossible. Motivated by this, we present a receding horizon control method with innovative trajectory planning policies that enable dynamic updating of planned trajectories in the presence of partially known and unknown uncertainty. The findings of this study have significant implications for achieving safe aviation autonomy within the AAM system.Ph.D