Download Full Text (1.3 MB)
The monitoring of photovoltaic panels is an essential task to maintain the efficiency of the power plant. The conventional form of inspection consists of manual inspection with a thermal camera which takes several days as fields contain thousands of panels. Solar panel installation is costly, and the installation does not guarantee maximum power generation without proper maintenance and handling. A drone with a thermal camera payload was developed to reduce labor costs as well as increase efficiency of the inspection as it can survey the field with a large field of view. Several image processing methods were used to process the incoming thermal camera feed to effectively detect hotspots. The purpose of this research is to investigate an alternative inspection method to speed up the process. A proof-of-concept field was generated with several pieces of wood and heat pads in certain sections which represent hotspots in the panels. The drone effectively followed a pre-planned trajectory to inspect the sections in the field and notify the user if a hotspot was detected. Once this hotspot is detected, the crew can provide the necessary maintenance to allow the panel to function efficiently. The drone was tested in a controlled environment, therefore it is expected for occlusions and detection errors to happen in a field with actual panels, yet that is planned to be addressed in future work.
Aerial vehicles; Thermal imaging; Robotics; Energy
Vega, Fausto, "Toward the Use of Drones for Photovoltaic Panel Inspection" (2021). Undergraduate Research Symposium Posters. 28.