Chemical-Informed Deep Learning for Early-Stage Photovoltaic Degradation: Decoding Material Failure Signatures From UAV- Based Thermal Imaging
Abstract
The long-term performance and economic viability of photovoltaic (PV) systems, primary technologies in green chemistry initiatives, are critically limited by the chemical and microstructural degradation of their semiconductor materials. Early detection of such degradation manifested as hotspots in thermal imagery due to inefficient energy conversion is essential for predictive maintenance. This study presents a deep learning framework designed to interpret thermal images as signatures of underlying material failure. We develop a hybrid model combining Inception modules for spatial feature extraction, Bidirectional LSTM layers for temporal analysis of degradation progression, and Multi-Head Attention mechanisms to focus on the most diagnostically relevant image regions. Trained on a dataset of UAV-captured thermal images, the model classifies solar cells into "defected" and "non-defected" states with an accuracy of 89.65%, a precision of 93.67%, and an Area Under the ROC Curve of 0.9613. The high performance demonstrates the model's capability to identify the thermal footprints of early-stage chemical and electronic defects. This work bridges materials chemistry and artificial intelligence, providing a robust, automated tool for large-scale health assessment of PV materials. It underscores the potential of chemistry-aware deep learning to monitor and preserve the integrity of energy-conversion materials in the field.
Keywords
Photovoltaic DegradationSemiconductor Materials ChemistryThermal ImagingDeep LearningInception ModuleBidirectional LSTMMulti-Head AttentionPredictive MaintenanceGreen Chemistry. 1
References
Akay, S. S., Özcan, O., Özcan, O., & Yetemen, Ö. (2024). Efficiency analysis of solar farms by UAV-based thermal monitoring. Engineering Science and Technology, an International Journal, 53, 101688. https://doi.org/10.1016/j.jestch.2024.101688 Binomairah, A., Al-Amri, A., Al-Ghamdi, A., & Al-Goorani, M. (2022). Detection of microcracks and dark spots in monocrystalline PERC cells using photoluminescence imaging and YOLO-based CNN with spatial pyramid pooling. EPJ Photovoltaics, 13, 27. https://doi.org/10.1051/epjpv/2022025 Bommes, L., Pickel, T., Buerhop-Lutz, C., Hauch, J. A., Brabec, C. J., & Peters, I. M. (2021). Computer vision tool for detection, mapping, and fault classification of photovoltaics modules in aerial IR videos. Progress in Photovoltaics: Research and Applications, 29(12), 1236–1251. https://doi.org/10.1002/pip.3448 Bosman, L. B., Leon-Salas, W. D., Hutzel, W., & Soto, E. A. (2020). PV system predictive maintenance: Challenges, current approaches, and opportunities. Energies, 13(6), 1398. https://doi.org/10.3390/en13061398 He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep Residual Learning for Image Recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 770–778. Jamuna, V., Muniraj, C., & Periasamy, P. S. (2023). Fault detection for photovoltaic panels in solar power plants by using Linear Iterative Fault Diagnosis technique based on thermal imaging system. Journal of Electrical Engineering & Technology, 18(4), 3091– 3103. https://doi.org/10.1007/s42835-023-01381-7 Jumaboev, S., Jurakuziev, D., & Lee, M. (2022). Photovoltaics plant fault detection using deep learning techniques. Remote Sensing, 14(15), 3728. https://doi.org/10.3390/rs14153728 Kirubakaran, V., Jain, V., Nithiyananthan, K., & Ramachandran, K. (2022). Infrared thermal images of solar PV panels for fault identification using image processing technique. International Journal of Photoenergy, 2022, 1–10. https://doi.org/10.1155/2022/6427076 Krishna Veni, K. S., Senthil Kumar, N., & Gnanavadivel, J. (2024). Non-invasive health status diagnosis of solar PV panel using ensemble classifier. Electrical Engineering. Advance online publication. https://doi.org/10.1007/s00202-024-02760-7 Li, X., Wang, Y., Monday, H. N., & Nneji, G. U. (2025). A novel residual learning of multi- scale feature extraction model for the classification of rice grain varieties. Computers and Electronics in Agriculture, 237, 110491. Menéndez, O., Guamán, R., Pérez, M., & Cheein, F. A. (2018). Photovoltaic modules diagnosis using artificial vision techniques for artifact minimization. Energies, 11(7), 1688. https://doi.org/10.3390/en11071688 Monday, H. N., Li, J., Nneji, G. U., Hossin, M. A., Nahar, S., Jackson, J., & Chikwendu, I. A. (2022a). WMR-DepthwiseNet: A wavelet multi-resolution depthwise separable convolutional neural network for COVID-19 diagnosis. Diagnostics, 12(3), 765. https://doi.org/10.3390/diagnostics12030765 Monday, H. N., Li, J., Nneji, G. U., Hossin, M. A., Nahar, S., Jackson, J., & Ejiyi, C. J. (2022b). COVID-19 diagnosis from chest X-ray images using a robust multi-resolution analysis Siamese neural network with super-resolution convolutional neural network. Diagnostics, 12(3), 741. https://doi.org/10.3390/diagnostics12030741 Monday, H. N., Nneji, G. U., Hossin, M. A., Mark, K. D., Umana, E. S., Mgbejime, G. T., & Li, J. (2025). Enhancing ECG classification in cardiac diagnostics: A novel approach using adaptive focal cross-entropy loss function. IEEE Journal of Biomedical and Health Informatics. Nneji, G. U., Cai, J., Deng, J., Monday, H. N., James, E. C., & Ukwuoma, C. C. (2022a). Identification of diabetic retinopathy using weighted fusion deep learning based on dual- channel fundus scans. Diagnostics, 12(2), 540. https://doi.org/10.3390/diagnostics12020540 Nneji, G. U., Cai, J., Deng, J., Monday, H. N., James, E. C., & Ukwuoma, C. C. (2022b). Multi- channel based image processing scheme for pneumonia identification. Diagnostics, 12(2), 325. https://doi.org/10.3390/diagnostics12020325 Nneji, G. U., Cai, J., Monday, H. N., Hossin, M. A., Nahar, S., Jackson, J., & Deng, J. (2022c). Fine-tuned Siamese network with modified enhanced super-resolution GAN plus based on low quality chest X-ray images for COVID-19 identification. Diagnostics, 12(3), 717. https://doi.org/10.3390/diagnostics12030717 Nneji, G. U., Deng, J., Monday, H. N., Cai, J., Hossin, M. A., Nahar, S., & Jackson, J. (2022d). COVID-19 identification from low-quality computed tomography using a modified enhanced super-resolution generative adversarial network plus and Siamese capsule network. Healthcare, 10(2), 403. https://doi.org/10.3390/healthcare10020403 Nneji, G. U., Monday, H. N., Pathapati, V. S. R., Nahar, S., Mgbejime, G. T., Umana, E. S., & Hossin, M. A. (2025). FFS-IML: Fusion-based statistical feature selection for machine learning-driven interpretability of chronic kidney disease. International Journal of Machine Learning and Cybernetics, 1–34. Pa, M., Uddin, M. N., & Kazemi, A. (2022, December). A fault detection scheme utilizing convolutional neural network for PV solar panels with high accuracy. In 2022 IEEE 1st Industrial Electronics Society Annual On-Line Conference (pp. 1–5). IEEE. https://doi.org/10.1109/ONCON56984.2022.10126746 Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016, June). You only look once: Unified, real-time object detection. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (pp. 779–788). IEEE. https://doi.org/10.1109/CVPR.2016.91 Rudro, R. A. M., et al. (2024). SPF-Net: Solar panel fault detection using U-Net based deep learning image classification. Energy Reports, 12, 1580–1594. https://doi.org/10.1016/j.egyr.2024.07.044 Segovia Ramírez, I., García Márquez, F. P., & Parra Chaparro, J. (2024). Convolutional neural networks and Internet of Things for fault detection by aerial monitoring of photovoltaic solar plants. Measurement, 234, 114861. https://doi.org/10.1016/j.measurement.2024.114861 Tian, Y. (2020). Artificial intelligence image recognition method based on convolutional neural network algorithm. IEEE Access, 8, 125731–125744. https://doi.org/10.1109/ACCESS.2020.3006097 Wang, J., et al. (2022). Deep-learning-based automatic detection of photovoltaic cell defects in electroluminescence images. Sensors, 23(1), 297. https://doi.org/10.3390/s23010297 Wang, X., Yang, W., Qin, B., Wei, K., Ma, Y., & Zhang, D. (2022). Intelligent monitoring of photovoltaic panels based on infrared detection. Energy Reports, 8, 5005–5015. https://doi.org/10.1016/j.egyr.2022.03.173
More Articles from WORLD JOURNAL OF INNOVATION AND MODERN TECHNOLOGY
AI-Assisted Risk and Incident Analysis: Advances and Future Opportunities
Author: Uchechi Mary-Linda Unamma, Ifeanyichukwu Jeffrey Okwesa, Serif Oyindamola, Oyesiji, Abubakar Umar Abdulmalik
Influence of Artificial Intelligence Tools on Teachers’ Instructional Delivery in Public Senior Secondary Schools in Gokana and Khana Local Government Areas of Rivers State, Nigeria
Author: BIIBALOO Juliet Legborsi, EDO Barineka Lucky, PhD NWILE Charles Befii, PhD.
Community-Centered Digital Skills: Empowering Adults for Life- Long-Learning in Yola North Local Government Area
Author: Adamu Abdullahi Potiskum, Emeka Godwin Timothy, Lawan Ladan
Dielectric Performance of Natural Ester Insulating Fluids as a Sustainable Alternative to Mineral Transformer Oil: A Systematic Review
Author: Ude Kingsley Okechukwu, Ugwu Kelvin Ikechukwu, Mmamel Ngozi Juliana, Nnamani, Micheal Obiora
Artificial Intelligence and Inclusive Practices in Special Needs Education: Opportunities, Challenges, and Future Directions
Author: Rahima Ahmadu Ribadu, Bobboi Abubakar