A review of machine learning and drone-based solar panel inspection techniques

This paper aims to improve defect identification, operational efficiency, and cost-effectiveness of drone-based photovoltaic (PV) solar panel inspection methods by leveraging artificial intelligence (AI) algorithms and modern imaging technologies. As the usage of renewable energy sources develops, it is vital to maintain solar panels in a timely and accurate way to guarantee their maximum perfor- mance. Innovative methods for large-scale solar panel inspection are made possible by drones equipped with thermal imaging, YOLOv8 object detection, Convolutional Neural Networks (CNNs), Single Shot Multibox Detectors (SSD), and mapping tools like Pix4D. Every method has proven effective in locating flaws, including hotspots, dust buildup, and cracks, enhancing system performance. This study thoroughly analyses these techniques, evaluating their feasibility, accuracy, and efficiency for small and large-scale solar installations. We hope to provide insights into the best drone-based solutions for solar panel inspection and maintenance by contrasting the advantages and disadvantages of each strategy, thereby assisting in developing a more dependable and sustainable renewable energy infrastructure.