在智慧製造與預測性維護發展下，設備健康監測已不僅是判斷異常，更需將模型結果轉換為可追溯、可檢查且具維護意義的診斷資訊。軸承振動訊號具有高頻、非平穩與局部衝擊等特性，傳統閾值與人工經驗難以穩定處理長期退化；深度學習模型雖能提升故障辨識與RUL預測能力，但若僅輸出故障標籤或數值，仍難以支援實際維護決策。
本研究建構一套結合可解釋深度學習與診斷代理人之軸承設備健康監測方法，以CWRU與XJTU-SY軸承資料集為基礎，透過統一前處理、故障分類、RUL預測、事件融合、XAI證據、FMEA概念圖譜、維修文件檢索與歷史案例檢索，將視窗層級模型輸出彙整為具時間脈絡與來源依據的監測事件。大型語言模型代理人在本系統中不重新判斷故障類別或RUL，而是在固定模型事實與引用來源限制下，生成診斷摘要、檢查建議與維護方向。
實驗結果顯示，TCN-Attention於故障分類任務中取得最佳表現，Macro-F1達0.9614。RUL預測方面，採用工程特徵序列的Causal TCN將minute MAE由236.90分鐘降低至147.72分鐘，改善幅度約37.64%。事件融合於11顆XJTU-SY軸承、共6,013個資料擷取批次中皆成功形成監測事件，事件召回率為1.000，篩選後故障正確率亦為1.000。診斷代理人經契約式後處理後，JSON物件輸出率、固定事實保留率與引用來源白名單有效率皆達0.95。
本研究將深度學習模型輸出由單一判斷提升為具事件脈絡、模型證據、知識來源與受控回應的診斷資訊流程。透過事件融合、XAI與檢索式知識支援，系統可協助維護人員理解異常形成原因、檢視模型依據，並取得具來源限制的維護建議，為後續即時監控與人機協同PHM系統奠定基礎。

This study develops an equipment health monitoring method for rolling bearing diagnosis by integrating explainable deep learning, event-level evidence fusion, maintenance knowledge retrieval, and a controlled diagnostic agent. Conventional condition monitoring methods often rely on fixed thresholds or isolated model outputs, which are insufficient for supporting maintenance decisions when vibration signals are non-stationary, noisy, and degradation-dependent. To address this issue, this study uses CWRU and XJTU-SY bearing datasets to construct a workflow covering unified preprocessing, fault classification, remaining useful life prediction, anomaly event fusion, explainable artificial intelligence evidence, FMEA-based knowledge representation, retrieval-augmented maintenance information, and diagnostic report generation. Experimental results show that TCN-Attention achieves the best fault classification performance, with a Macro-F1 score of 0.9614. For RUL prediction, the Causal TCN using engineered feature sequences reduces minute MAE from 236.90 to 147.72 minutes. The event fusion mechanism was evaluated using 6,013 acquisition batches from 11 XJTU-SY bearings. Monitoring events were successfully formed for all 11 bearings, yielding an event recall of 1.000 and a filtered fault accuracy of 1.000. After deterministic contract postprocessing, the valid JSON-object output rate, immutable fact preservation rate, and citation allowlist validity all reached 0.95. These results indicate that the proposed method can transform model outputs into traceable, explainable, and reviewable diagnostic information for maintenance decision support.

Bahdanau, D., Cho, K., & Bengio, Y. (2014). Neural machine translation by jointly learning to align and translate [Preprint]. arXiv. https://arxiv.org/abs/1409.0473
Bai, S., Kolter, J. Z., & Koltun, V. (2018). An empirical evaluation of generic convolutional and recurrent networks for sequence modeling [Preprint]. arXiv. https://arxiv.org/abs/1803.01271
Borré, A., Seman, L. O., Camponogara, E., Stefenon, S. F., Mariani, V. C., & Coelho, L. S. (2023). Machine fault detection using a hybrid CNN-LSTM attention-based model. Sensors, 23(9), Article 4512. https://doi.org/10.3390/s23094512
Brehme, L., Dornauer, B., Ströhle, T., Ehrhart, M., & Breu, R. (2025). Retrieval-augmented generation in industry: An interview study on use cases, requirements, challenges, and evaluation. In Proceedings of the 17th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (pp. 110–122). SCITEPRESS. https://doi.org/10.5220/0013739500004000
Cação, J., Santos, J., & Antunes, M. (2025). Explainable AI for industrial fault diagnosis: A systematic review. Journal of Industrial Information Integration, 47, Article 100905. https://doi.org/10.1016/j.jii.2025.100905
Case Western Reserve University Bearing Data Center. (n.d.). Welcome to the Case Western Reserve University Bearing Data Center website. Retrieved August 2, 2026, from https://engineering.case.edu/bearingdatacenter/welcome
Cha, M., Yoon, S.-I., Kim, S., Kang, D., Nam, K., Lee, T., & Kim, J.-Y. (2025). Large language model-based autonomous agent for prognostics and health management. Machines, 13(9), Article 831. https://doi.org/10.3390/machines13090831
Chen, L.-C., Pardeshi, M. S., Liao, Y.-X., & Pai, K.-C. (2025). Application of retrieval-augmented generation for interactive industrial knowledge management via a large language model. Computer Standards & Interfaces, 94, Article 103995. https://doi.org/10.1016/j.csi.2025.103995
Cho, K., van Merriënboer, B., Bahdanau, D., & Bengio, Y. (2014). On the properties of neural machine translation: Encoder-decoder approaches. In Proceedings of SSST-8, Eighth Workshop on Syntax, Semantics and Structure in Statistical Translation (pp. 103–111). Association for Computational Linguistics. https://doi.org/10.3115/v1/W14-4012
Deng, H., Namoano, B., Zheng, B., Khan, S., & Erkoyuncu, J. A. (2024). From prediction to prescription: Large language model agent for context-aware maintenance decision support. PHM Society European Conference, 8(1), Article 10. https://doi.org/10.36001/phme.2024.v8i1.4114
Es, S., James, J., Espinosa-Anke, L., & Schockaert, S. (2024). RAGAs: Automated evaluation of retrieval augmented generation. In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations (pp. 150–158). Association for Computational Linguistics. https://doi.org/10.18653/v1/2024.eacl-demo.16
International Organization for Standardization. (2003). Condition monitoring and diagnostics of machines—Data processing, communication and presentation—Part 1: General guidelines (ISO Standard No. 13374-1:2003). https://www.iso.org/standard/21832.html
International Organization for Standardization. (2015). Condition monitoring and diagnostics of machine systems—Data processing, communication and presentation—Part 4: Presentation (ISO Standard No. 13374-4:2015). https://www.iso.org/standard/54933.html
Jardine, A. K. S., Lin, D., & Banjevic, D. (2006). A review on machinery diagnostics and prognostics implementing condition-based maintenance. Mechanical Systems and Signal Processing, 20(7), 1483–1510. https://doi.org/10.1016/j.ymssp.2005.09.012
Jiang, L., Chen, J., Cao, H., Li, P., & Wang, K. (2025). A novel bearing fault diagnosis method using a hybrid TCN-transformer architecture: A deep learning approach. Journal of Mechanical Science and Technology, 39, 3821–3834. https://doi.org/10.1007/s12206-025-0607-5
Kamat, P., Kumar, S., Patil, S., & Kotecha, K. (2024). Anomaly-informed remaining useful life estimation (AIRULE) of bearing machinery using deep learning framework. MethodsX, 12, Article 102555. https://doi.org/10.1016/j.mex.2024.102555
Khan, U., Cheng, D. S., Setti, F., Fummi, F., Cristani, M., & Capogrosso, L. (2026). A comprehensive survey on deep learning-based predictive maintenance. ACM Transactions on Embedded Computing Systems, 25(2), 1–43. https://doi.org/10.1145/3732287
Lee, J., Wu, F., Zhao, W., Ghaffari, M., Liao, L., & Siegel, D. (2014). Prognostics and health management design for rotary machinery systems—Reviews, methodology and applications. Mechanical Systems and Signal Processing, 42(1–2), 314–334. https://doi.org/10.1016/j.ymssp.2013.06.004
Lei, Y., Li, N., Guo, L., Li, N., Yan, T., & Lin, J. (2018). Machinery health prognostics: A systematic review from data acquisition to RUL prediction. Mechanical Systems and Signal Processing, 104, 799–834. https://doi.org/10.1016/j.ymssp.2017.11.016
Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W.-T., Rocktäschel, T., Riedel, S., & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive NLP tasks. Advances in Neural Information Processing Systems, 33, 9459–9474. https://proceedings.neurips.cc/paper_files/paper/2020/hash/6b493230205f780e1bc26945df7481e5-Abstract.html
Li, R., Verhagen, W. J. C., & Curran, R. (2020). A systematic methodology for prognostic and health management system architecture definition. Reliability Engineering & System Safety, 193, Article 106598. https://doi.org/10.1016/j.ress.2019.106598
Li, Y., Zhao, H., Jiang, H., Pan, Y., Liu, Z., Wu, Z., Shu, P., Tian, J., Yang, T., Xu, S., Lyu, Y., Blenk, P., Pence, J., Rupram, J., Banu, E., Song, K., Zhu, D., Wang, X., & Liu, T. (2026). Large language models for manufacturing. Journal of Manufacturing Systems, 86, 516–545. https://doi.org/10.1016/j.jmsy.2026.02.014
Li, Y.-F., Wang, H., & Sun, M. (2024). ChatGPT-like large-scale foundation models for prognostics and health management: A survey and roadmaps. Reliability Engineering & System Safety, 243, Article 109850. https://doi.org/10.1016/j.ress.2023.109850
Li, Z., He, Q., & Li, J. (2024). A survey of deep learning-driven architecture for predictive maintenance. Engineering Applications of Artificial Intelligence, 133, Article 108285. https://doi.org/10.1016/j.engappai.2024.108285
Liang, Y., Mao, W., & Wu, C. (2025). Unsupervised incremental transfer learning with knowledge distillation for online remaining useful life prediction of rotating machinery. Proceedings of the Institution of Mechanical Engineers, Part O: Journal of Risk and Reliability, 239(1), 15–30. https://doi.org/10.1177/1748006X231223777
Lin, L., Zhang, S., Fu, S., & Liu, Y. (2025). FD-LLM: Large language model for fault diagnosis of complex equipment. Advanced Engineering Informatics, 65(Part A), Article 103208. https://doi.org/10.1016/j.aei.2025.103208
Liu, H., Zhang, F., Tan, Y., Huang, L., Li, Y., Huang, G., Luo, S., & Zeng, A. (2024). Multi-scale quaternion CNN and BiGRU with cross self-attention feature fusion for fault diagnosis of bearing. Measurement Science and Technology, 35(8), Article 086138. https://doi.org/10.1088/1361-6501/ad4c8e
Liu, Y., Zhang, W., Bao, Z., Chai, X., Gu, M., Jiang, W., Zhang, Z., Tian, Y., & Wang, F.-Y. (2025). Brain-like cognition-driven model factory for IIoT fault diagnosis by combining LLMs with small models. IEEE Internet of Things Journal, 12(16), 32296–32309. https://doi.org/10.1109/JIOT.2024.3503274
Lukens, S., & Ali, A. (2023). Evaluating the performance of ChatGPT in the automation of maintenance recommendations for prognostics and health management. Annual Conference of the PHM Society, 15(1). https://doi.org/10.36001/phmconf.2023.v15i1.3487
Lukens, S., McCabe, L. H., Gen, J., & Ali, A. (2024). Large language model agents as prognostics and health management copilots. Annual Conference of the PHM Society, 16(1). https://doi.org/10.36001/phmconf.2024.v16i1.3906
Lundberg, S. M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30, 4765–4774. https://proceedings.neurips.cc/paper_files/paper/2017/hash/8a20a8621978632d76c43dfd28b67767-Abstract.html
Moenks, N., Penava, P., & Buettner, R. (2025). A systematic literature review of large language model applications in industry. IEEE Access, 13, 160010–160033. https://doi.org/10.1109/ACCESS.2025.3608650
Mohd Ghazali, M. H., & Rahiman, W. (2021). Vibration analysis for machine monitoring and diagnosis: A systematic review. Shock and Vibration, 2021, Article 9469318. https://doi.org/10.1155/2021/9469318
Mustapha, K. B. (2025). A survey of emerging applications of large language models for problems in mechanics, product design, and manufacturing. Advanced Engineering Informatics, 64, Article 103066. https://doi.org/10.1016/j.aei.2024.103066
Nikiforidis, K., Kyrtsoglou, A., Vafeiadis, T., Kotsiopoulos, T., Nizamis, A., Ioannidis, D., Votis, K., Tzovaras, D., & Sarigiannidis, P. (2025). Enhancing transparency and trust in AI-powered manufacturing: A survey of explainable AI (XAI) applications in smart manufacturing in the era of Industry 4.0/5.0. ICT Express, 11(1), 135–148. https://doi.org/10.1016/j.icte.2024.12.001
Noot, J.-P., Martin, M., & Birmele, E. (2025). LSTM and transformers based methods for remaining useful life prediction considering censored data. International Journal of Prognostics and Health Management, 16(2). https://doi.org/10.36001/IJPHM.2025.v16i2.4260
Nunes, P., Santos, J., & Rocha, E. (2023). Challenges in predictive maintenance: A review. CIRP Journal of Manufacturing Science and Technology, 40, 53–67. https://doi.org/10.1016/j.cirpj.2022.11.004
Pan, Y., Kang, S., Kong, L., Wu, J., Yang, Y., & Zuo, H. (2025). Remaining useful life prediction methods of equipment components based on deep learning for sustainable manufacturing: A literature review. Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 39, Article e4. https://doi.org/10.1017/S0890060424000271
Randall, R. B., & Antoni, J. (2011). Rolling element bearing diagnostics—A tutorial. Mechanical Systems and Signal Processing, 25(2), 485–520. https://doi.org/10.1016/j.ymssp.2010.07.017
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016). “Why should I trust you?”: Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 1135–1144). Association for Computing Machinery. https://doi.org/10.1145/2939672.2939778
Shi, D., Li, J., Meyer, O., & Bauernhansl, T. (2025). Enhancing retrieval-augmented generation for interoperable industrial knowledge representation and inference toward cognitive digital twins. Computers in Industry, 171, Article 104330. https://doi.org/10.1016/j.compind.2025.104330
Smith, W. A., & Randall, R. B. (2015). Rolling element bearing diagnostics using the Case Western Reserve University data: A benchmark study. Mechanical Systems and Signal Processing, 64–65, 100–131. https://doi.org/10.1016/j.ymssp.2015.04.021
Sundararajan, M., Taly, A., & Yan, Q. (2017). Axiomatic attribution for deep networks. In D. Precup & Y. W. Teh (Eds.), Proceedings of the 34th International Conference on Machine Learning (Vol. 70, pp. 3319–3328). PMLR. https://proceedings.mlr.press/v70/sundararajan17a.html
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., & Polosukhin, I. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 30, 5998–6008. https://proceedings.neurips.cc/paper_files/paper/2017/hash/3f5ee243547dee91fbd053c1c4a845aa-Abstract.html
Vogl, G. W., Weiss, B. A., & Helu, M. (2019). A review of diagnostic and prognostic capabilities and best practices for manufacturing. Journal of Intelligent Manufacturing, 30(1), 79–95. https://doi.org/10.1007/s10845-016-1228-8
Wang, B., Lei, Y., Li, N., & Li, N. (2020a). A hybrid prognostics approach for estimating remaining useful life of rolling element bearings. IEEE Transactions on Reliability, 69(1), 401–412. https://doi.org/10.1109/TR.2018.2882682
Wang, B., Lei, Y., Yan, T., Li, N., & Guo, L. (2020b). Recurrent convolutional neural network: A new framework for remaining useful life prediction of machinery. Neurocomputing, 379, 117–129. https://doi.org/10.1016/j.neucom.2019.10.064
Wang, L., Ma, C., Feng, X., Zhang, Z., Yang, H., Zhang, J., Chen, Z., Tang, J., Chen, X., Lin, Y., Zhao, W. X., Wei, Z., & Wen, J.-R. (2024). A survey on large language model-based autonomous agents. Frontiers of Computer Science, 18(6), Article 186345. https://doi.org/10.1007/s11704-024-40231-1
Wang, T., Zhang, B., Jiang, D., & Li, D. (2025). A multimodal large language model framework for intelligent perception and decision-making in smart manufacturing. Sensors, 25(10), Article 3072. https://doi.org/10.3390/s25103072
Wen, Y., Rahman, M. F., Xu, H., & Tseng, T.-L. B. (2022). Recent advances and trends of predictive maintenance from data-driven machine prognostics perspective. Measurement, 187, Article 110276. https://doi.org/10.1016/j.measurement.2021.110276
Zhang, W., Peng, G., Li, C., Chen, Y., & Zhang, Z. (2017). A new deep learning model for fault diagnosis with good anti-noise and domain adaptation ability on raw vibration signals. Sensors, 17(2), Article 425. https://doi.org/10.3390/s17020425
Zhao, R., Yan, R., Chen, Z., Mao, K., Wang, P., & Gao, R. X. (2019). Deep learning and its applications to machine health monitoring. Mechanical Systems and Signal Processing, 115, 213–237. https://doi.org/10.1016/j.ymssp.2018.05.050
Zhou, R., Gebraeel, N., & Serban, N. (2012). Degradation modeling and monitoring of truncated degradation signals. IIE Transactions, 44(9), 793–803. https://doi.org/10.1080/0740817X.2011.618175
Zio, E. (2022). Prognostics and health management (PHM): Where are we and where do we (need to) go in theory and practice. Reliability Engineering & System Safety, 218, Article 108119. https://doi.org/10.1016/j.ress.2021.108119