Uncrewed aerial vehicles (UAVs) are widely used in various applications, including computation offloading (COF) for Internet of Things (IoT) devices. However, their limited battery life presents a major challenge. We propose a novel energy management system allowing a UAV to procure energy— at a given price—from locally deployed laser beam directors (LBDs) and renewable energy (RE) sources. To offset energy procurement costs (EPCs), the UAV earns revenue by providing COF services and wirelessly charging low-power IoT devices at a fixed price. A lightweight reinforcement learning (RL)- based trajectory design approach is developed to maximize the number of offloaded IoT devices by selecting an optimal sequence of regions to visit within a limited mission time. We then formulate a minimization problem to manage the UAV’s EPC while ensuring sufficient energy for COF along the optimized path. The proposed system and trajectory design aim to enhance UAV efficiency in IoT COF applications and reduce EPCs. Local RE sources further improve system sustainability. Simulation results show that over 28% of IoT devices (32% of regions) can offload tasks within a 180-second mission, while UAV EPC is significantly reduced compared to baseline energy procurement without service pricing.
