A Comparative Evaluation of Thermographic Inspection Methods for Pv Plants: Manual, Single Uav, and Cooperative Multi-Drone Strategies

This paper presents a comparative evaluation of three thermographic inspection strategies for photovoltaic (PV) plants: manual ground inspection, single unmanned aerial vehicle (UAV) scanning, and a cooperative multi-UAV approach. Simulations were conducted in Webots, emulating infrared detection through color-based hotspot identification across an array of 80 PV panels. The cooperative method employs a leader-follower architecture, where a leader UAV rapidly scans the entire plant and delegates high-resolution verification tasks to four secondary UAVs via a consensus mechanism, optimizing routes with Dijkstra’s algorithm. Results show that the cooperative system reduces inspection time by approximately 92% compared to manual inspection and 41% compared to single-UAV scanning, while improving robustness against failures and ensuring non-overlapping coverage. The findings highlight the potential of coordinated UAV swarms to enhance the efficiency, scalability, and reliability of PV plant maintenance. Future work will explore real-world deployment, AI-driven hotspot classification, and integration with cloud-based analytics.