Abstract

The accumulation of dust, bird droppings, rain residues, and physical damage on the surface of solar panels reduces energy production efficiency, leading to significant long-term losses. Detecting and cleaning these contaminants and structural damages in a timely manner is critical for reducing maintenance costs and extending the lifespan of the panels. Traditional cleaning systems typically focus on cleaning the entire surface, which results in unnecessary water and energy consumption, leading to inefficiencies. In this study, an AI-assisted segmentation system has been developed to detect only dirty or damaged areas, enabling targeted cleaning and maintenance. The proposed system is integrated with a YOLO-based deep learning model to perform real-time object detection and segmentation. A large-scale dataset containing over 35,000 images was created, and data augmentation techniques were applied to enhance the model’s robustness against various environmental conditions. YOLOv9, YOLOv10, and YOLOv11 models were trained and compared in terms of object detection and segmentation accuracy. Experimental results demonstrated that the YOLOv11 model achieved the highest performance with a 96.9% mAP accuracy. The model exhibited superior performance in detecting details such as cracks, dust layers, and stains with high precision. This study optimizes the maintenance processes of solar panels by preventing unnecessary energy and water consumption, reducing maintenance costs, and enhancing sustainability in clean energy production. Operating in real time, this system supports intelligent maintenance processes in solar power plants, offering an innovative solution for the industry.

References

[1] Yolcan, O. O., & Köse, R. (2020). Türkiye’nin güneş enerjisi durumu ve güneş enerjisi santrali kurulumunda önemli parametreler. Kırklareli Üniversitesi Mühendislik ve Fen Bilimleri Dergisi, 6(2), 196-215.

[2] Nsengiyumva, W., Chen, S. G., Hu, L., & Chen, X. (2018). Recent advancements and challenges in Solar Tracking Systems (STS): A review. Renewable and Sustainable Energy Reviews, 81, 250-279.

[3] İçel, Y. (2024). Kirliliğin PV Panel Verimine Etkisi ve Panel Temizleme Yöntemleri. Interdisciplinary Studies on Contemporary Research Practices in Engineering in the 21st Century-VI, 201.

[4] Derakhshandeh, J. F., AlLuqman, R., Mohammad, S., AlHussain, H., AlHendi, G., AlEid, D., & Ahmad, Z. (2021). A comprehensive review of automatic cleaning systems of solar panels. Sustainable Energy Technologies and Assessments, 47, 101518.

[5] El-Shobokshy, M. S., & Hussein, F. M. (1993). Degradation of photovoltaic cell performance due to dust deposition on to its surface. Renewable energy, 3(6-7), 585-590.

[6] Ghazi, S., Sayigh, A., & Ip, K. (2014). Dust effect on flat surfaces–A review paper. Renewable and Sustainable Energy Reviews, 33, 742-751.

[7] Sayyah, A., Horenstein, M. N., & Mazumder, M. K. (2014). Energy yield loss caused by dust deposition on photovoltaic panels. Solar Energy, 107, 576-604.

[8] Akinyele, D. O., & Rayudu, R. K. (2014). Sustainable energy technologies and assessments. Review of energy storage technologies for sustainable power networks, 8, 2213-1388.

[9] Jaradat, M. A., Tauseef, M., Altaf, Y., Saab, R., Adel, H., Yousuf, N., & Zurigat, Y. H. (2015, December). A fully portable robot system for cleaning solar panels. In 2015 10th International Symposium on Mechatronics and its Applications (ISMA) (pp. 1-6). IEEE.

[10] Huang, J., Zeng, K., Zhang, Z., & Zhong, W. (2023). Solar panel defect detection design based on YOLO v5 algorithm. Heliyon, 9(8).

[11] Olorunfemi, B. O., Ogbolumani, O. A., & Nwulu, N. (2022). Solar panels dirt monitoring and cleaning for performance improvement: a systematic review on smart systems. Sustainability, 14(17), 10920.

[12] Ronnaronglit, N., & Maneerat, N. (2019, July). A cleaning robot for solar panels. In 2019 5th International Conference on Engineering, Applied Sciences and Technology (ICEAST) (pp. 1-4). IEEE.

[13] He, Z., Zhang, Y., & Li, H. (2019, March). Self-inspection cleaning device for photovoltaic power plant based on machine vision. In IOP Conference Series: Earth and Environmental Science (Vol. 242, p. 032020). IOP Publishing.

[14] al-Urdunīyah, J. A. Z. (2016). Universiti Sains Malaysia, and Institute of Electrical and Electronics Engineers. Internet of Things: Architectures, Protocols, and Applications.

[15] Abuqaaud, K. A., & Ferrah, A. (2020, February). A novel technique for detecting and monitoring dust and soil on solar photovoltaic panel. In 2020 Advances in Science and Engineering Technology International Conferences (ASET) (pp. 1-6). IEEE.

[16] He, K., Gkioxari, G., Dollár, P., & Girshick, R. (2017). Mask r-cnn. In Proceedings of the IEEE international conference on computer vision (pp. 2961-2969).

[17] Vuola, A. O., Akram, S. U., & Kannala, J. (2019, April). Mask-RCNN and U-net ensembled for nuclei segmentation. In 2019 IEEE 16th international symposium on biomedical imaging (ISBI 2019) (pp. 208-212). IEEE.

[18 ]Widyaningrum, R., Candradewi, I., Aji, N. R. A. S., & Aulianisa, R. (2022). Comparison of Multi-Label U-Net and Mask R-CNN for panoramic radiograph segmentation to detect periodontitis. Imaging science in dentistry, 52(4), 383.

[19 ] Chou, A., Li, W., & Roman, E. (2022). GI tract image segmentation with U-Net and mask R-CNN. Image Segmentation with U-Net and Mask R-CNN, 1-9.

[20] Chen, L. C., Papandreou, G., Kokkinos, I., Murphy, K., & Yuille, A. L. (2017). Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs. IEEE transactions on pattern analysis and machine intelligence, 40(4), 834-848.

[21] Wang, C. Y., Yeh, I. H., & Mark Liao, H. Y. (2025). Yolov9: Learning what you want to learn using programmable gradient information. In European Conference on Computer Vision (pp. 1-21). Springer, Cham.

[22] Sapkota, R., Meng, Z., Ahmed, D., Churuvija, M., Du, X., Ma, Z., & Karkee, M. (2024). Comprehensive performance evaluation of yolov10, yolov9 and yolov8 on detecting and counting fruitlet in complex orchard environments. Authorea Preprints.

[23] Imran, A., Hulikal, M. S., & Gardi, H. A. (2024). Real Time American Sign Language Detection Using Yolo-v9. arXiv preprint arXiv:2407.17950.