Sustainable last-mile delivery is essential for businesses aiming to stay competitive while reducing environmental impact. Retailers are increasingly adopting green, efficient methods like drones for last-mile operations to protect ecosystems. Recent research highlights key challenges in drone delivery, including routing, cargo optimization, battery management, data communication, and environmental protection—interconnected issues critical to achieving eco-friendly and efficient delivery. This research focuses on addressing these challenges, particularly battery limitations, routing optimization, and real-time adaptability. This research proposes a framework integrating Mobile Wireless Charging Stations (MWCS) with Model Predictive Control (MPC), reinforcement learning, and genetic algorithms to enhance drone range and efficiency. A Genetic Algorithm (GA) optimizes delivery routes based on delivery points and MWCS locations, while MPC adjusts drone trajectories and MWCS placement dynamically. Simulations show this approach significantly improves delivery times, energy use, and system efficiency over traditional methods, supporting sustainable and reliable last-mile drone delivery.

Keywords: drone, last-mile drone delivery, routing, genetic algorithm, model predictive control, reinforcement learning.