Optimising Management of Energy, Traffic and Sensor Data in Urban Aerial Systems

Abstract

Urban Air Mobility (UAM) presents a promising solution to alleviate ground congestion by integrating Unmanned Aerial Vehicles (UAVs) and Passenger Aerial Vehicles (PAVs) into urban transportation networks. However, the widespread adoption of UAM faces multiple challenges, such as battery capacities, urban airspace management complexity, and the efficient handling of high-volume sensor data generated by aerial platforms. This thesis addresses some of the key challenges by developing scalable, energy-aware, and computationally efficient frameworks using diverse optimisation techniques for traffic management and facilitate the efficient integration of UAVs and PAVs into urban airspace.

Firstly, we develop energy models using empirical studies on UAV battery charge-discharge dynamics by varying charging conditions, payloads, altitudes, manoeuvres, etc. These models enable accurate predictions of energy consumption, facilitating optimal scheduling and routing of drones to extend their operational range and mission duration. Our approach advances by holistically balancing energy constraints, sensing accuracy, and mission criticality.

Secondly, we propose novel optimisation strategies to improve vertiport terminal efficiency, which is crucial for managing UAM traffic in dense urban settings. Using a Mixed-Integer Linear Programming (MILP) formulation, we address scheduling and resource allocation at vertiports. Our methods enhance throughput by leveraging multi-directional flight paths and provide theoretical throughput bounds to guide infrastructure planning.

Lastly, we introduce a hybrid data compression technique designed for sensor data storage and transmission. This method significantly reduces data storage and transmission bottlenecks by compressing high-volume sensor data while preserving the fidelity necessary for reliable operation. The framework presented can be utilised for aircraft, UAVs, and other vehicles, helping in decision-making for efficient operations.