Enhancing Solar Panel Efficiency: An IoT-Integrated Machine Learning Approach for Dust Detection

India faces significant energy challenges due to rising demand and environmental concerns from fossil fuel use. Solar energy, abundant in India, offers a sustainable solution, but its efficiency is compromised by dust accumulation, especially in arid regions. This research develops an image-based machine learning model using EfficientNet to detect dust on solar panels, integrated with an IoT system featuring ESP32-CAM and Flask for automated cleaning. Utilizing a dataset of 2,563 labeled images, the EfficientNet model outperformed ResNet50 in accuracy and efficiency. The IoT system effectively triggered automated cleaning or sent alerts for manual intervention, optimizing solar panel performance. This integration of machine learning and IoT demonstrates significant potential for enhancing solar energy efficiency, reducing maintenance costs, and supporting renewable energy adoption.