隨著製造業之轉型，智慧製造的發展也日益成熟，全廠自動化物料搬運系統也在現今科技廠房中扮演著關鍵的角色，負責各製程間物料的傳輸，猶如人體中的血液，扮演著養份輸送至各器官的角色，所以自動化物料搬運系統需長時間穩定運轉，往往非預期性當機將造成產線生線中斷而影響產能。藉由物聯網及無線傳輸的進步，將振動感測器佈置於自動化物料搬運系統的減速機上，即時監控及收集減速機運轉時所產生的振動特徵資料，並透過人工智慧分析方法將收集到的大數據轉換為具有價值之資訊，得以改善自動化物料搬運系統的穩定度。
本研究主要對象為自動化物料搬運系統中的減速機，減速機往往隨著運轉時間的增加而產生內部齒輪磨耗，最終齒輪因疲勞而斷裂，造成內部齒輪咬死無法運轉，透過振動感測器收集減速機運轉時所產生的特徵資料，將收集到的時間序列資料進行深度學習並建立其預測模型，預估未來減速機劣化之趨勢，提供設備工程師安排預期性停機更換計劃，減少非預期性當機的情況。
本研究的結果顯示，在特定參數調整和疊代次數下，透過雙向長短期記憶（Bi-directional Long Short-Term Memory, Bi-LSTM）模型的訓練，可以預測減速機的劣化趨勢。這對設備工程師來說具有重要意義，因為他們可以根據預測的趨勢來判斷最佳的更換時機。本研究的重點在於預測未來劣化的趨勢，並結合設備工程師的經驗，期望能透過時間序列預測模型來準確預測未來的趨勢，以提供最佳的更換時機。此外，這項研究也有助於有效降低設備備品庫存成本，並提高設備備品的利用率。

In traditional maintenance operations of automated material handling system(AMHS), the practice of relying on regular shutdowns for maintenance and replacing parts based on their lifecycle is no longer adequate. It often leads to unexpected failures of high-utilization reducer when they reach the end of their lifespan. Therefore, to anticipate the degradation trends of reducer in advance, this study utilizes vibration sensors combined with engineering expertise to identify the degradation characteristics of reducers. Then, it employs Bi-LSTM to predict the degradation trends. The study first collects vibration sensor data from reducer and, through spectrum analysis combined with equipment engineer experience, identifies the degradation features. Based on a time-series model, the time-series feature data are used to construct predictive models for forecasting degradation trends. Finally, the performance of predicted data and actual data is evaluated through evaluation metrics and visual analysis.
The results demonstrate that, with appropriate parameter adjustments, the predicted degradation trend by Bi-LSTM can serve as a basis for engineering judgment. In the analysis of reducer degradation trends, the dataset comprises "vibration sensor features", and the predictive model is "Bidirectional Long Short-Term Memory (Bi-LSTM)". The predicted degradation trend aligns closely with the actual trend, and when visualized, their trends are roughly similar.

[1] 林俊良, & 郭雅玲. (2005). 應用馬可夫決策過程探討當機機台最佳之修復策略. 科技管理學刊, 10(2), 31-56.
[2] Munirathinam, S., & Ramadoss, B. (2016). Predictive models for equipment fault detection in the semiconductor manufacturing process. IACSIT International Journal of Engineering and Technology, 8(4), 273-285.
[3] Yang, G. (2014). The wear mechanism of the reducer box and preventive measures. Journal of applied science and engineering innovation Vol, 1(4).
[4] G. Bontempi, S. Ben Taieb, and Y.-A. L. Borgne, "Machine learning strategies for time series forecasting," in European business intelligence summer school, 2012: Springer, pp. 62-77.
[5] 張定原, & 簡孝龍. (2018). 行星減速機精度測試設備創新研發設計研究. 管理學術研討會, 729-741.
[6] Chandshah, T. A., & Kamoji, M. A. (2017). Design and Strength Analysis of Load Roller Trolley Conveyor System. International Research Journal of Engineering and Technology (IRJET) e-ISSN, 2395-0056.
[7] El-Nashar, A. H., & El-Kilany, K. S. MODELLING AND SIMULATION OF INTRABAY AMHS IN 300MM WAFER FABRICAT ION FACILITIES: AN OVERVIEW.
[8] 林坤濱（2008）。內嵌式自動倉儲 (In-line Stocker) 於大尺寸面板廠應用之探討–以 TFT-LCD 公司為例。國立中央大學工業管理研究所碩士在職專班碩士論文，桃園市。取自https://hdl.handle.net/11296/mb76x9
[9] Jang, Y. J., & Choi, G. H. (2006, October). Introduction to automated material handling systems in LCD panel production lines. In 2006 IEEE International Conference on Automation Science and Engineering (pp. 223-229). IEEE.
[10] Wei, N. C., & Lin, H. K. (2012). Evaluation of automated guided vehicle systems in thin film transistor liquid crystal display (TFT-LCD) Array manufacturing process. Scientific Research and Essays, 7(14), 3542-3548.
[11] Helo, P., Suorsa, M., Hao, Y., & Anussornnitisarn, P. (2014). Toward a cloud-based manufacturing execution system for distributed manufacturing. Computers in industry, 65(4), 646-656.
[12] Govindaraju, R., & Putra, K. (2016, February). A methodology for Manufacturing Execution Systems (MES) implementation. In IOP Conference Series: Materials Science and Engineering (Vol. 114, No. 1, p. 012094). IOP Publishing.
[13] Shojaeinasab, A., Charter, T., Jalayer, M., Khadivi, M., Ogunfowora, O., Raiyani, N., ... & Najjaran, H. (2022). Intelligent manufacturing execution systems: A systematic review. Journal of Manufacturing Systems, 62, 503-522.
[14] Chen, X., & Voigt, T. (2020). Implementation of the Manufacturing Execution System in the food and beverage industry. Journal of Food Engineering, 278, 109932.
[15] Sonar, A., Shinde, S., & Teh, S. (2013, May). Automation: Key to cycle time improvement in semiconductor manufacturing. In ASMC 2013 SEMI advanced semiconductor manufacturing conference (pp. 93-98). IEEE.
[16] Li, B., Wu, J., Herring, M. P., & Mouli, C. (2005, September). A breakthrough in optimizing throughput and manufacturing efficiency via factory-wide dynamic WIP dispatching in 300mm semiconductor manufacturing. In ISSM 2005, IEEE International Symposium on Semiconductor Manufacturing, 2005. (pp. 39-42). IEEE.
[17] Sun, D. S., Park, N. S., Lee, Y. J., Jang, Y. C., Ahn, C. S., & Lee, T. E. (2005, April). Integration of lot dispatching and AMHS control in a 300mm wafer FAB. In IEEE/SEMI Conference and Workshop on Advanced Semiconductor Manufacturing 2005. (pp. 270-274). IEEE.
[18] Lin, C. Y., Yeh, M. T., & Kuo, C. C. (2010, October). ISSM 2010 a high availability multi-layered MCS system. In 2010 International Symposium on Semiconductor Manufacturing (ISSM) (pp. 1-4). IEEE.
[19] Chang, J., & Cheng, Y. (2004, September). Just-in-time AMHS delivery for 300 mm FAB. In 2004 Semiconductor Manufacturing Technology Workshop Proceedings (IEEE Cat. No. 04EX846) (pp. 25-28). IEEE.
[20] Hwang, S., Hong, S. P., & Jang, Y. J. (2018, August). Dynamic scheduling of the dual stocker system using reinforcement learning. In IFIP International Conference on Advances in Production Management Systems (pp. 482-489). Cham: Springer International Publishing.
[21] Pop, C., & Mailat, G. (2011, April). Automated Material Handling Systems (AMHS) in libraries and archives Automated Storage/retrieval and Return/sorting Systems. In Proceedings of the 12th WSEAS international conference on Neural networks, fuzzy systems, evolutionary computing & automation (pp. 189-194).
[22] Chung, J. (2015). Estimating arrival times of transportation jobs for automated material handling in LCD fabrication facilities. Journal of Manufacturing Systems, 35, 112-119.
[23] Saffar, S., Jamaludin, Z., & Jafar, F. A. (2018). Investigating the Influences of Automated Material Handling System (AMHS) and Effect of Layout Changing in Automotive Assembly Process. In Intelligent Manufacturing & Mechatronics: Proceedings of Symposium, 29 January 2018, Pekan, Pahang, Malaysia (pp. 117-129). Springer Singapore.
[24] Lynch, L., Newe, T., Clifford, J., Coleman, J., Walsh, J., & Toal, D. (2018, December). Automated ground vehicle (agv) and sensor technologies-a review. In 2018 12th International Conference on Sensing Technology (ICST) (pp. 347-352). IEEE.
[25] Ganesharajah, T., Hall, N. G., & Sriskandarajah, C. (1998). Design and operational issues in AGV-served manufacturing systems. Annals of operations Research, 76(0), 109-154.
[26] Ananthi, K. (2023, June). Efficient Warehouse Management System with Automated Storage and Retrieval. In 2023 8th International Conference on Communication and Electronics Systems (ICCES) (pp. 1755-1760). IEEE.
[27] Abi-Char, P. E., & Riman, C. (2021, July). A Collision-Free Path Planning Algorithm for Non-Complex ASRS Using Heuristic Functions. In 2021 44th International Conference on Telecommunications and Signal Processing (TSP) (pp. 52-57). IEEE.
[28] Nadoli, G., & Pillai, D. (1994, December). Simulation in automated material handling systems design for semiconductor manufacturing. In Proceedings of Winter Simulation Conference (pp. 892-899). IEEE.
[29] Kong, F., Ge, Y., Zhu, X., Qiao, L., & Quan, L. (2013, October). Optimizing design of magnetic planetary reducer for reduction of cogging torque. In 2013 IEEE Vehicle Power and Propulsion Conference (VPPC) (pp. 1-5). IEEE.
[30] Lei, Y., Lin, J., Zuo, M. J., & He, Z. (2014). Condition monitoring and fault diagnosis of planetary reduceres: A review. Measurement, 48, 292-305.
[31] Sedak, M., & Rosić, B. (2021). Multi-objective optimization of planetary reducer with adaptive hybrid particle swarm differential evolution algorithm. Applied Sciences, 11(3), 1107.
[32] Feng, Z., & Zuo, M. J. (2012). Vibration signal models for fault diagnosis of planetary reduceres. Journal of Sound and Vibration, 331(22), 4919-4939.
[33] Zhang, M., Wang, K., Wei, D., & Zuo, M. J. (2018). Amplitudes of characteristic frequencies for fault diagnosis of planetary reducer. Journal of Sound and Vibration, 432, 119-132.
[34] Hong, L., & Dhupia, J. S. (2013, July). A time-domain fault detection method based on an electrical machine stator current measurement for planetary gear-sets. In 2013 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (pp. 1631-1636). IEEE.
[35] Rauber, T. W., de Assis Boldt, F., & Varejao, F. M. (2014). Heterogeneous feature models and feature selection applied to bearing fault diagnosis. IEEE Transactions on Industrial Electronics, 62(1), 637-646.

[36] Khazaee, M., Ahmadi, H., Omid, M., & Moosavian, A. (2012). An appropriate approach for condition monitoring of planetary reducer based on fast Fourier transform and least-square support vector machine. International Journal of Multidisciplinary Sciences and Engineering, 3(5), 22-26.
[37] Helmi, H., & Forouzantabar, A. (2019). Rolling bearing fault detection of electric motor using time domain and frequency domain features extraction and ANFIS. IET Electric Power Applications, 13(5), 662-669.
[38] Kelkar, S. S., Grigsby, L. L., & Langsner, J. (1983). An extension of Parseval's theorem and its use in calculating transient energy in the frequency domain. IEEE Transactions on Industrial Electronics, (1), 42-45.
[39] 林庭毅（2008）。一般化傅立葉轉換之研究。國立臺灣科技大學電機工程所碩士論文，臺北市。取自https://hdl.handle.net/11296/begh5v。
[40] Oberst, U. (2007). The fast Fourier transform. SIAM journal on control and optimization, 46(2), 496-540.
[41] 黃柏諭（2021）。小波分析與卷積神經網路於風機齒輪箱振動故障診斷之應用。國立臺灣大學工程科學及海洋工程學研究所碩士論文，臺北市。取自https://hdl.handle.net/11296/3b4jrp。
[42] McFadden, P. D. (1986). Detecting fatigue cracks in gears by amplitude and phase demodulation of the meshing vibration. Journal of Vibration and Acoustics, 108(2), 165-170.
[43] Y. Xing et al., "A survey on smart agriculture: Development modes, technologies, and security and privacy challenges," IEEE/CAA Journal of Automatica Sinica, vol. 8, no. 2, pp. 273-302, 2021.
[44] S. A. Bhat and N.-F. Huang, "Big data and ai revolution in precision agriculture: Survey and challenges," IEEE Access, vol. 9, pp. 110209-110222, 2021.
[45] C. Ren, D.-K. Kim, and D. Jeong, "A survey of deep learning in agriculture: techniques and their applications," Journal of Information Processing Systems, vol. 16, no. 5, pp. 1015-1033, 2020.
[46] Abid, A., Khan, M. T., & Khan, M. S. (2017). Multidomain features-based GA optimized artificial immune system for bearing fault detection. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 50(1), 348-359.
[47] H. Chan, R. Chinipardaz, and T. Cox, "Discrimination of AR, MA and ARMA time series models," Communications in Statistics-Theory and Methods, vol. 25, no. 6, pp. 1247-1260, 1996.
[48] Chandra, R., Goyal, S., & Gupta, R. (2021). Evaluation of deep learning models for multi-step ahead time series prediction. IEEE Access, 9, 83105-83123.
[49] Ning, Y., Kazemi, H., & Tahmasebi, P. (2022). A comparative machine learning study for time series oil production forecasting: ARIMA, LSTM, and Prophet. Computers & Geosciences, 164, 105126.
[50] S. Hochreiter and J. Schmidhuber, "Long short-term memory," Neural computation, vol. 9, no. 8, pp. 1735-1780, 1997.
[51] Roodschild, M., Gotay Sardiñas, J., & Will, A. (2020). A new approach for the vanishing gradient problem on sigmoid activation. Progress in Artificial Intelligence, 9(4), 351-360.
[52] M. Phi. "Illustrated Guide to LSTM’s and GRU’s: A step by step explanation." https://towardsdatascience.com/illustrated-guide-to-lstms-and-gru-s-a-step-by-step-explanation-44e9eb85bf21 (accessed.
[53] Landi, F., Baraldi, L., Cornia, M., & Cucchiara, R. (2021). Working memory connections for LSTM. Neural Networks, 144, 334-341.
[54] Siami-Namini, S., Tavakoli, N., & Namin, A. S. (2019, December). The performance of LSTM and BiLSTM in forecasting time series. In 2019 IEEE International conference on big data (Big Data) (pp. 3285-3292). IEEE.
[55] Shen, S. L., Atangana Njock, P. G., Zhou, A., & Lyu, H. M. (2021). Dynamic prediction of jet grouted column diameter in soft soil using Bi-LSTM deep learning. Acta Geotechnica, 16(1), 303-315.
[56] Yang, C. H., Wu, C. H., Shao, J. C., Wang, Y. C., & Hsieh, C. M. (2022). AIS-based intelligent vessel trajectory prediction using bi-LSTM. IEEE Access, 10, 24302-24315.
[57] Jaseena, K. U., & Kovoor, B. C. (2021). Decomposition-based hybrid wind speed forecasting model using deep bidirectional LSTM networks. Energy Conversion and Management, 234, 113944.
[58] Taylor, S. J., & Letham, B. (2018). Forecasting at scale. The American Statistician, 72(1), 37-45.
[59] D. Chicco, M. J. Warrens, and G. Jurman, "The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation," PeerJ Computer Science, vol. 7, p. e623, 2021.
[60] Graves, A., Fernández, S., & Schmidhuber, J. (2005, September). Bidirectional LSTM networks for improved phoneme classification and recognition. In International conference on artificial neural networks (pp. 799-804). Berlin, Heidelberg: Springer Berlin Heidelberg.
[61] Siami-Namini, S., Ta vakoli, N., & Namin, A. S. (2019, December). The performance of LSTM and BiLSTM in forecasting time series. In 2019 IEEE International conference on big data (Big Data) (pp. 3285-3292). IEEE.
[62] Vartholomaios, A., Karlos, S., Kouloumpris, E., & Tsoumakas, G. (2021, September). Short-term renewable energy forecasting in greece using prophet decomposition and tree-based ensembles. In International Conference on Database and Expert Systems Applications (pp. 227-238). Cham: Springer International Publishing.
[63] Zhoul, L., Chenl, M., & Ni, Q. (2020, December). A hybrid prophet-LSTM model for prediction of air quality index. In 2020 IEEE Symposium Series on Computational Intelligence (SSCI) (pp. 595-601). IEEE.
[64] Arslan, S. (2022). A hybrid forecasting model using LSTM and Prophet for energy consumption with decomposition of time series data. PeerJ Computer Science, 8, e1001.