Ukázková aplikace využívající metod offloading

Abstract

The increasing deployment of drones in logistics for package delivery tasks involves complex computational challenges, such as real-time image processing for navigation. These activities often push the drone's computational and energy capacities to their limits, potentially compromising their flight speed and delivery efficiency. This work explores the concept of computational offloading as a viable solution to these constraints. This study involves developing a demonstration application that uses simulation to explore the impacts of offloading on such a system's behavior. The results reveal that effectively implemented offloading can increase delivery throughput by up to 88 % by reducing the computational load on drone processors and minimizing energy consumption. This enhancement allows drones to operate for extended periods and at higher speeds. Utilizing autonomous drone fleets further contributes to significant reductions in delivery costs, shorter delivery times, and improved customer satisfaction.