在智慧製造中，公司透過機器感測器獲取大量製造數據。透過整合數據提取資訊，實現故障預測和健康管理（PHM）。PHM採用數據挖掘、大數據分析和人工智慧，包括狀態監控、故障診斷、預測和壽命追蹤，為機器配置和維護決策提供指南以提高系統的可靠性、效率和安全性，同時降低維護成本和減少系統停機時間。
半導體在台灣被視為領先的高科技產業並具有重要的地位，然而它也是本國能源消耗最大的行業之一，因此半導體產業設施系統中的預測性和健康管理議題至關重要，因為它可以幫助降低整體能源消耗成本，提高機器效率。在半導體產業中，生產線的主要驅動力是由空壓機提供的氣流，因此空壓機的生產效率對生產線的運作有顯著影響。
由於工業上存在大量未標記的數據，本研究演示了如何利用人工智慧中的非監督學習技術將原始數據轉換為機器健康分數，並進一步識別參數偏差的權重，並開發了先進設備監控系統以使用者介面的視覺化方式呈現，並可透過與其互動的方式使現場人員能迅速識別異常參數，判斷是否需要預測性維護，而不是傳統的預防性維護方法。本研究執行來自台灣半導體公司的實證研究成功地識別了異常的時間並計算了參數偏差的權重，為決策者提供了關鍵資訊以促進決策。此方法降低過度維護成本、能源消耗成本，並提高機器效率。

In smart manufacturing, companies acquire extensive manufacturing data from machine sensors. Extracting value requires integrating data, achieving fault prediction, and health management, which is the　essence of Prognostic and Health Management (PHM). PHM employs data mining, big data analysis, and AI. It encompasses condition monitoring, fault diagnosis, prediction, and lifespan tracking, offering guidance for machine configuration and maintenance decisions.
The semiconductor industry is regarded as a leading high-tech sector in Taiwan and holds significant importance. However, it is also one of the largest energy-consuming industries in the country. Therefore, the issue of Prognostic and Health Management (PHM) in the facility systems of the semiconductor industry is crucial as it can help reduce overall energy consumption costs and increase machine efficiency. In the semiconductor industry, the main driving force of the production line is the airflow supplied by the air compressor, so the production efficiency of the air compressor has a significant impact on the operation of the production line.
Due to the predominance of unlabeled data in the industrial sector, this research demonstrates how unsupervised learning techniques in artificial intelligence can be utilized to convert raw data into machine health scores. It further identifies the weight of parameter deviations and presents them visually, enabling on-site personnel to quickly identify the abnormal parameter and determine whether predictive maintenance is necessary, rather than the conventional preventive maintenance approach.
An empirical case study from a semiconductor company in Taiwan successfully identifies the timing of anomalies and calculates the weight of parameter deviations, providing decision-makers with crucial information to facilitate decision-making. This approach aims to reduce excessive maintenance costs, energy consumption costs, and enhance machine efficiency.

Ahmad, S., Styp-Rekowski, K., Nedelkoski, S., & Kao, O. (2020). Autoencoder-based condition monitoring and anomaly detection method for rotating machines. In 2020 IEEE International Conference on Big Data (Big Data) (pp. 4093-4102). IEEE.
Bae, Y. M., Kim, Y. G., Seo, J. W., Kim, H. A., Shin, C. H., Son, J. H., ... & Kim, K. J. (2023). Detecting abnormal behavior of automatic test equipment using autoencoder with event log data. Computers & Industrial Engineering, 183, 109547.
Batzel, T. D., & Swanson, D. C. (2009). Prognostic health management of aircraft power generators. IEEE Transactions on Aerospace and Electronic Systems, 45(2), 473-482.
Biggio, L., & Kastanis, I. (2020). Prognostics and health management of industrial assets: Current progress and road ahead. Frontiers in Artificial Intelligence, 3, 578613.
Che, C., Wang, H., Fu, Q., & Ni, X. (2019). Combining multiple deep learning algorithms for prognostic and health management of aircraft. Aerospace Science and Technology, 94, 105423.
Colombo, A. W., Schleuter, D., & Kircher, M. (2015). An approach to qualify human resources supporting the migration of SMEs into an Industrie4. 0-compliant company infrastructure. In IECON 2015-41st Annual Conference of the IEEE Industrial Electronics Society (pp. 003761-003766). IEEE.
Dong, Y., Xia, T., Fang, X., Zhang, Z., & Xi, L. (2019). Prognostic and health management for adaptive manufacturing systems with online sensors and flexible structures. Computers & Industrial Engineering, 133, 57-68.
Duan, C., Makis, V., & Deng, C. (2020). A two-level Bayesian early fault detection for mechanical equipment subject to dependent failure modes. Reliability Engineering & System Safety, 193, 106676.
Gürdür, D., El-Khoury, J., Seceleanu, T., & Lednicki, L. (2016). Making interoperability visible: Data visualization of cyber-physical systems development tool chains. Journal of Industrial Information Integration, 4, 26-34.
Jeong, H., Park, B., Park, S., Min, H., & Lee, S. (2019). Fault detection and identification method using observer-based residuals. Reliability Engineering & System Safety, 184, 27-40.
Li, Z., Sun, Y., Yang, L., Zhao, Z., & Chen, X. (2022). Unsupervised machine anomaly detection using autoencoder and temporal convolutional network. IEEE Transactions on Instrumentation and Measurement, 71, 1-13.
Melakhsou, A. A., Batton-Hubert, M., & Casoetto, N. (2023). Welding fault detection and diagnosis using one-class SVM with distance substitution kernels and random convolutional kernel transform. The International Journal of Advanced Manufacturing Technology, 128(1-2), 459-477.
Meng, H., & Li, Y. F. (2019). A review on prognostics and health management (PHM) methods of lithium-ion batteries. Renewable and Sustainable Energy Reviews, 116, 109405.
Oesterreich, T. D., & Teuteberg, F. (2016). Understanding the implications of digitisation and automation in the context of Industry 4.0: A triangulation approach and elements of a research agenda for the construction industry. Computers in Industry, 83, 121-139.
Oztemel, E., & Gursev, S. (2020). Literature review of Industry 4.0 and related technologies. Journal of Intelligent Manufacturing, 31, 127-182.
Pang, G., Shen, C., Cao, L., & Hengel, A. V. D. (2021). Deep learning for anomaly detection: A review. ACM Computing Surveys (CSUR), 54(2), 1-38.
Peruzzini, M., Gregori, F., Luzi, A., Mengarelli, M., & Germani, M. (2017). A social life cycle assessment methodology for smart manufacturing: The case of study of a kitchen sink. Journal of Industrial Information Integration, 7, 24-32.
Sahli, A., Evans, R., & Manohar, A. (2021). Predictive maintenance in industry 4.0: Current themes. Procedia CIRP, 104, 1948-1953.
Tolo, S., Tian, X., Bausch, N., Becerra, V., Santhosh, T. V., Vinod, G., & Patelli, E. (2019). Robust on-line diagnosis tool for the early accident detection in nuclear power plants. Reliability Engineering & System Safety, 186, 110-119.
Torabi, H., Mirtaheri, S. L., & Greco, S. (2023). Practical autoencoder based anomaly detection by using vector reconstruction error. Cybersecurity, 6(1), 1.
Varghese, A., & Tandur, D. (2014). Wireless requirements and challenges in Industry 4.0. In 2014 International Conference on Contemporary Computing and Informatics (IC3I) (pp. 634-638). IEEE.
Vrignat, P., Kratz, F., & Avila, M. (2022). Sustainable manufacturing, maintenance policies, prognostics and health management: A literature review. Reliability Engineering & System Safety, 218, 108140.
Xia, T., Dong, Y., Xiao, L., Du, S., Pan, E., & Xi, L. (2018). Recent advances in prognostics and health management for advanced manufacturing paradigms. Reliability Engineering & System Safety, 178, 255-268.
Yan, S., Shao, H., Xiao, Y., Liu, B., & Wan, J. (2023). Hybrid robust convolutional autoencoder for unsupervised anomaly detection of machine tools under noises. Robotics and Computer-Integrated Manufacturing, 79, 102441.
Yucesan, Y. A., Dourado, A., & Viana, F. A. (2021). A survey of modeling for prognosis and health management of industrial equipment. Advanced Engineering Informatics, 50, 101404.
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, 108119.