Edge-Server Collaborative System for Real-Time and In-Depth Damage Detection of Wind Turbine Blades using Acoustic Signals
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
Efficient health monitoring is indispensable for the reliable operation of wind turbines. Damage to wind turbine blades, such as cracks and holes, typically generates whistle-like sounds during rotation. This study proposes a two-stage edge-server collaborative system for detecting blade damage using acoustic signals captured by arrays built from commodity microphones. The first stage employs a lightweight attention-based convolutional neural network to run on edge devices for the real-time binary classification to determine whether anomalous sounds are present. Suspicious time segments are stored for further analysis. The second stage uses a time-frequency sound event detection model that employs a detection transformer with an audio spectrogram transformer backbone to identify the time and frequency ranges of sound events via bounding boxes in the spectrograms. Owing to its high computational demand, this in-depth analysis is performed on a server. To validate the proposed system, acoustic signals were recorded intermittently for more than a year using micro-electromechanical system (MEMS) microphones externally attached to wind turbine towers. The models were trained and evaluated on a manually annotated dataset comprising 4,210 audio clips (15 s each) containing 14,420 sound events. The experimental results demonstrated that the binary classification model achieved an area under the receiver operating characteristic curve (AUC) of 0.920, whereas the sound event detection model attained an average precision at a 50% intersection-over-union threshold (AP50) of 0.510. Furthermore, evaluations on test data under unseen conditions, comprising 496 clips with 135 sound events recorded by handheld recorders at different locations, yielded an AUC of 0.867 and an AP50 of 0.440. The results highlight the robustness of the proposed system to variations in microphone types, recording locations, and environmental noise, demonstrating its strong potential for practical continuous automatic damage detection in wind power infrastructure.
How to Cite
##plugins.themes.bootstrap3.article.details##
Wind Turbine, Damage Detection, Deep Learning, Sound Event Detection, Audio Classification
ATR-Promotions. (2005). ATR ambient noise sound database 2. https://www.atr-p.com/products/esd.html.
Baade, A., Peng, P., & Harwath, D. (2022). MAE-AST: Masked autoencoding audio spectrogram transformer. In Interspeech (pp. 2438–2442).
Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., & Zagoruyko, S. (2020). End-to-end object detection with transformers. In Proceedings of the European Conference on Computer Vision. doi: 10.1007/978-3-030-58452-8_13
Chen, W., Liang, Y., Ma, Z., Zheng, Z., & Chen, X. (2024). EAT: Self-supervised pre-training with efficient audio transformer. In Proceedings of the International Joint Conference on Artificial Intelligence (pp. 3807–3815). doi: 10.24963/ijcai.2024/421
Chong, D., Wang, H., Zhou, P., & Zeng, Q. (2023). Masked spectrogram prediction for self-supervised audio pre-training. In ICASSP.
Cuesta, J., Leturiondo, U., Vidal, Y., & Pozo, F. (2024). A review of prognostics and health management in wind turbine components. In Proceedings of the European Conference of the Prognostics and Health Management Society. doi: 10.36001/phme.2024.v8i1.4093
Ding, S., Yang, C., & Zhang, S. (2023). Acoustic-signal-based damage detection of wind turbine blades: A review. Sensors, 23(11), 4987. doi: 10.3390/s23114987
Fonseca, E., Favory, X., Pons, J., Font, F., & Serra, X. (2022). FSD50K: An open dataset of human-labeled sound events. IEEE/ACM Transactions on Audio, Speech, and Language Processing, 30, 829–852. doi: 10.1109/TASLP.2021.3133208
Fremmelev, M. A., Ladpli, P., Orlowitz, E., Bernhammer, L. O., McGugan, M., & Branner, K. (2022). Structural health monitoring of 52-meter wind turbine blade: Detection of damage propagation during fatigue testing. Data-Centric Engineering, 3, e22. doi: 10.1017/dce.2022.20
Gemmeke, J. F., Ellis, D. P. W., Freedman, D., Jansen, A., Lawrence, W., & Moore, R. C. (2017). Audio set: An ontology and human-labeled dataset for audio events. In Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing. doi: 10.1109/ICASSP.2017.7952261
Gong, Y., Chung, Y.-A., & Glass, J. (2021). AST: Audio spectrogram transformer. In Proceedings of Interspeech (pp. 571–575). doi: 10.21437/Interspeech.2021-698
Gong, Y., Lai, C.-I., Chung, Y.-A., & Glass, J. (2022). SSAST: Self-supervised audio spectrogram transformer. In AAAI (Vol. 36, pp. 10699–10709).
Huang, P.-Y., Xu, H., Li, J., Baevski, A., Auli, M., Galuba, W., ... Feichtenhofer, C. (2022). Masked autoencoders that listen. In NeurIPS.
Kuo, S.-F., Cheng, S., Lo, F.-C., & Tu, T.-H. (2023). Wind turbine blade damage detection and classification based on sound feature signal using machine learning. In Proceedings of the Asia Pacific Conference of Sound and Vibration. doi: 10.3397/IN_2023_0637
Lamdjad, B., & Chaiter, A. (2026, March). AI-powered predictive maintenance and prognostic health management using edge-based predictive algorithms for industrial operations. Preprints. doi: 10.20944/preprints202603.0010.v1
Li, F., Li, L., & Peng, Y. (2021). Research on digital twin and collaborative cloud and edge computing applied in operations and maintenance in wind turbines of wind power farm. Advances in Transdisciplinary Engineering, 17, 80–92. doi: 10.3233/ATDE210263
Li, F., Zhang, H., Liu, S., Guo, J., Ni, L. M., & Zhang, L. (2022). DN-DETR: Accelerate DETR training by introducing query denoising. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 13619–13627). doi: 10.1109/TPAMI.2023.3335410
Li, M., Xu, Z., Li, S., Kikuchi, Y., Dong, Y., Gryllias, K. C., ... Carroll, J. (2026). Health prognostics and maintenance decision-making for wind energy: A comprehensive overview. Renewable and Sustainable Energy Reviews, 226, 116269. doi: 10.1016/j.rser.2025.116269
Li, Y., Mao, H., Girshick, R., & He, K. (2022). Exploring plain vision transformer backbones for object detection. In Proceedings of the European Conference on Computer Vision. doi: 10.1007/978-3-031-20077-9_17
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., ... Zitnick, C. L. (2014). Microsoft COCO: Common objects in context. In Proceedings of the European Conference on Computer Vision. doi: 10.1007/978-3-319-10602-1_48
Liu, S., Li, F., Zhang, H., Yang, X., Qi, X., Su, H., ... Zhang, L. (2022). DAB-DETR: Dynamic anchor boxes are better queries for DETR. In Proceedings of the International Conference on Learning Representations.
Niizumi, D., Takeuchi, D., Ohishi, Y., Harada, N., & Kashino, K. (2022). Masked spectrogram modeling using masked autoencoders for learning general-purpose audio representation. In PMLR (Vol. 166, pp. 1–24).
Solimine, J., Niezrecki, C., & Inalpolat, M. (2020). An experimental investigation into passive acoustic damage detection for structural health monitoring of wind turbine blades. Structural Health Monitoring, 19(6), 1711–1725. doi: 10.1177/1475921719895
Sun, B., Ooi, K. T., & Su, M. (2026). Wind turbine blade damage: A systematic review of detection, diagnosis, performance impact, and lifecycle health management. Renewable and Sustainable Energy Reviews, 230, 116668. doi: 10.1016/j.rser.2025.116668
Van Dam, J., & Bond, L. J. (2015). Acoustic emission monitoring of wind turbine blades. In Proc. SPIE 9439, Smart Materials and Nondestructive Evaluation for Energy Systems (p. 94390C). doi: 10.1117/12.2084527
von Däniken, E., Mikhaylov, D., Moallemi, A., Polonelli, T., & Magno, M. (2024). Tiny on-device structural health monitoring for wind turbines using MEMS pressure sensors. In 2024 IEEE Sensors Applications Symposium (SAS) (pp. 1–6). doi: 10.1109/SAS60918.2024.10636460
Wang, C., Guo, R., Yu, H., Hu, Y., Liu, C., & Deng, C. (2023). Task offloading in cloud-edge collaboration-based cyber physical machine tool. Robotics and Computer-Integrated Manufacturing, 79, 102439. doi: 10.1016/j.rcim.2022.102439
Wang, W., Xue, Y., He, C., & Zhao, Y. (2022). Review of the typical damage and damage-detection methods of large wind turbine blades. Energies, 15(15), 5672. doi: 10.3390/en15155672
Wei, W., Zhu, H., Benetos, E., & Wang, Y. (2021). Environmental sound classification using temporal-frequency attention based convolutional neural network. Scientific Reports, 11, 21557. doi: 10.1038/s41598-021-01045-4
Yang, C., Ding, S., & Zhou, G. (2025). Wind turbine blade damage detection based on acoustic signals. Scientific Reports, 15(1), 3930. doi: 10.1038/s41598-025-88276-x
Zhang, H., Li, F., Liu, S., Zhang, L., Su, H., Zhu, J., ... Shum, H.-Y. (2023). DINO: DETR with improved denoising anchor boxes for end-to-end object detection. In Proceedings of the International Conference on Learning Representations.
Zhang, Y., Avallone, F., & Watson, S. (2023). Leading edge erosion detection for a wind turbine blade using far-field aerodynamic noise. Applied Acoustics, 207, 109365. doi: 10.1016/j.apacoust.2023.109365
Zhang, Z., He, T., Zhang, H., Zhang, Z., Xie, J., & Li, M. (2019). Bag of freebies for training object detection neural networks.
Zhang, Z., Xu, S., Zhang, S., & Cao, S. (2021). Attention-based convolutional recurrent neural network for environmental sound classification. Neurocomputing, 453. doi: 10.1016/j.neucom.2020.08.069
Zhou, X., Kang, Z., Canady, R., Bao, S., Balasubramanian, D. A., & Gokhale, A. (2021). Exploring cloud-assisted tiny machine learning application patterns for PHM scenarios. In Annual Conference of the PHM Society (Vol. 13). doi: 10.36001/phmconf.2021.v13i1.3054
Zhu, X., Su, W., Lu, L., Li, B., Wang, X., & Dai, J. (2021). Deformable DETR: Deformable transformers for end-to-end object detection. In Proceedings of the International Conference on Learning Representations.
Zhu, Y., & Liu, X. (2023). A lightweight CNN for wind turbine blade defect detection based on spectrograms. Machines, 11(1), 99. doi: 10.3390/machines11010099
Zhu, Y., Liu, X., Li, S., Wan, Y., & Cai, Q. (2022). Wind turbine blade defect detection based on acoustic features and small sample size. Machines, 10(12), 1184. doi: 10.3390/machines10121184
Zhu, Z., & Sato, Y. (2025). Sound event detection using time-frequency bounding boxes with a self-supervised audio spectrogram transformer. In Proceedings of the Workshop on Detection and Classification of Acoustic Scenes and Events (DCASE) (pp. 150–154). doi: 10.5281/zenodo.17251589
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
The Prognostic and Health Management Society advocates open-access to scientific data and uses a Creative Commons license for publishing and distributing any papers. A Creative Commons license does not relinquish the author’s copyright; rather it allows them to share some of their rights with any member of the public under certain conditions whilst enjoying full legal protection. By submitting an article to the International Conference of the Prognostics and Health Management Society, the authors agree to be bound by the associated terms and conditions including the following:
As the author, you retain the copyright to your Work. By submitting your Work, you are granting anybody the right to copy, distribute and transmit your Work and to adapt your Work with proper attribution under the terms of the Creative Commons Attribution 3.0 United States license. You assign rights to the Prognostics and Health Management Society to publish and disseminate your Work through electronic and print media if it is accepted for publication. A license note citing the Creative Commons Attribution 3.0 United States License as shown below needs to be placed in the footnote on the first page of the article.
First Author et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.