Energy-Aware Drone Path Finding with a Fixed-Trajectory Ground Vehicle

Rotary-wing Unmanned Aerial Vehicles (commonly referred to as drones) are versatile autonomous transportation platforms that can be used for a variety of data collection applications including emergency response, environmental monitoring, surveillance, and many others.

In this work, we investigate how to plan efficient paths that minimize mission completion time for drone data collection where the drone must rendezvous with a moving ground vehicle (GV) that cannot stop and wait for the drone. Moreover, we address the limited onboard energy storage issue by adapting drone speed.

We propose a mixed-integer nonlinear program (MINLP) solution to solve this problem to optimality and provide two heuristics-based alternative solutions (the k -TSP and D -TSP approaches) that are more computationally tractable.

We evaluate these approaches in extensive simulations using real drone characteristics to highlight their trade-offs. Our results show that the k -TSP algorithm performs well when data collection points are closer to the GV, averaging within 4.5% of the optimal solution, while the D -TSP approach is more versatile, finding solutions in situations where the k -TSP algorithm tends to fail. Furthermore, we show that adapting drone speed can improve solution quality by up to 47.1% compared to fixed-speed approaches. In summary, this article serves as an exploratory study in energy-aware planning and scheduling for drones and other autonomous transportation systems.