台灣傳統製造業目前面臨因勞動人口高齡化、勞動力短缺等問題進而可能會造成技術失傳與經驗斷層的情況。因此，該如何有效的保留老師傅的經驗或是能夠善用智慧技術與工具找到以其他方式來輔助或是進一步取代須高度依賴經驗的決策模式是本研究欲探討的主題。
本研究以一間板金加工廠（Sheet Metal Fabrication）作為研究對象，然而板金加工是個在製造或決策時皆需高度依賴經驗的工藝，加上為了滿足顧客不同的需求皆採用接近100%高度客製化（Mass Customization）的生產方式導致各工件間的規格差異大，而且在現實狀況中常有各機台因機台結構設計不同會具有不同的工作能力的情形，將會大幅提升了排程（Scheduling）的困難度。以本研究之案例公司而言，多數的生產項目皆會經過折床製程（Bending process），加上廠內四台折床機台的機台差異性較大，每台皆有自己特殊的加工條件限制，綜合上述原因目前的排程作業需高度依賴具豐富經驗之員工來執行，所以本研究欲分別以專家系統（Expert System）與類神經網路（Artificial Neural Network）模型兩種方法來建構基於考量不同機台合適度限制（Machine Eligibility Constraint）的折床機台排程決策系統來輔助目前需依資深員工依過往經驗來執行決策的情況。
首先藉由與廠內資深員工進行訪談，並藉由訪談的過程將其過去的排程經驗以「IF-THEN」的形式建立一套規則專家系統，此外也透過智慧機上盒（Smart Machine Box, SMB）蒐集到的歷史資料來進行資料前處理，再藉由網格搜索法（Grid Search）與交叉驗證（Cross Validation）來找出最佳的參數組合來建構類神經網路模型並加以訓練，最後再將相同的測試資料分別放入此兩個模型中，並以資深員工判斷之結果作為決策標準對其結果進行比較與分析。
以本實驗之實驗結果可知，專家系統的表現較類神經網路模型優越，其準確率分別為0.9080與0.8613，所以以本研究而言利用專家系統來建構折床機台排程決策系統較為合適，但是以接近100%之高度客製化的板金產業來說，兩者皆有準確、可靠的表現，足以提供一個貼近於目前排程邏輯的結果輔助員工進行決策。

The traditional manufacturing industry in Taiwan is currently facing the problems of an aging workforce and labor shortage, which may lead to the losing of skills and experience gaps. Therefore, this research investigates how to effectively retain the experience from master staffs or to find alternative ways to support or to replace the highly experience-dependent decision-making model by using intelligent technologies and tools.
This study is based on a sheet metal fabrication factory. Sheet metal fabrication; however, is a crafting process that heavily relies on experience in both manufacturing and decision making. In addition, to meet the different needs of customers, the production method is nearly 100% customized causing large differences in specifications among work pieces, and in reality, each machine has different working capabilities due to different machine structure designs which will greatly increase the difficulty in scheduling. In this case study, most of the production items will go through the bending process. In addition, the four bending machines in the factory have large variability and have their special processing conditions. From those reasons, the current scheduling operation highly relies on experienced employees. Therefore, this research wants to construct a bending process scheduling decision system based on different machine eligibility constraints by using two methods: an expert system and an artificial neural network model, to assist the current situation in which decisions have to be made by experienced staff based on their experience.
First, by interviewing experienced staff in the factory and using their past scheduling experience in the form of "IF-THEN" to build a rule-based expert system. In addition, data pre-processing is conducted by using historical data collected from Smart Machine Box, and then the optimal combination of parameters is found by grid search and cross-validation to construct an artificial neural network model and train the model. Finally, the same tested data were put into these two models, and the results were compared and analyzed by using the results of experienced staff’ decisions as to the criteria.
The results of this experiment showed that the expert system performed better than the artificial neural network model, with the accuracy of 0.9080 and 0.8613 respectively, so it is more appropriate to use the expert system to construct the decision system for the bending process. However, with the highly customized sheet metal industry, which is close to 100%, both are accurate and reliable enough to provide a result that is close to the current scheduling logic to support staff in their decision making.

行政院主計處，2016，中華民國行業標準分類，available at https://tie.tier.org.tw/db/industry_definition/class_2.aspx?code=IND18-13 (取得日期：2021.03.01)
經濟部工業局，2019，108年度智慧機上盒(SMB)輔導計畫，available at http://www.tami.org.tw/wisdom_machine/wisdom_machine-735-1.pdf (取得日期：2021.03.15)
經濟部統計處，2020，工業產銷存價值統計調查，available at https://dmz26.moea.gov.tw/GMWeb/investigate/InvestigateDA.aspx (取得日期：2021.02.28)
楊大和、李兆華，2021，精實管理4.0之自働生產管理環，工業工程雜誌，2021年5月，available at https://sites.google.com/ciie.org.tw/column501/精實管理4-0之自働生產管理環?authuser=2 (取得日期：2021.06.12)

方曉嵐，1996，排程方法研究，技術與訓練期刊，第 21 卷第一期，149-250頁。
林我聰，1994，現場排程專家系統-應用個體導向技術建立之研究，資訊與電腦公司出版，臺北市。
林純行，2002，多階平行機器零工式多目標排程之模式化與系統之研究，東海大學工業工程與經營資訊研究所，碩士，臺中市。
張國鎮、宋裕祺、洪曉慧、劉光晏、陳俊仲、賴明俊、許家銓，2013，橋梁檢測專家系統之建置研究，國家地震工程研究中心，臺北市。
張勝富，2017，板金業自動化之成本效益評估 － 以b公司雷射加工作業為例，崑山科技大學企業管理研究所，碩士，臺南市。
陳允傑，2019，Tensorflow 與 Keras：Python 深度學習應用實務，旗標科技股份有限公司，臺北市。
葉怡成，2006，類神經網路模式應用與實作，儒林圖書公司，臺北市。
葉怡成、郭耀煌，1991，專家系統:方法應用與實作，全欣，臺北市。
詹文男、李亦晴、李建勳、朱師右、張俐婷，2020，智慧製造與機器人應用發展趨勢，財團法人資訊工業策進會產業情報研究所(MIC)，臺北市。
簡禎富、許嘉裕，2014，資料挖礦與大數據分析，前程文化事業有限公司，新北市。
蘇芳慶，2018，科學發展-工業 4.0 時代來臨，科技部，No.544，P22-25。

Akyol, D. E., 2004, Application of neural networks to heuristic scheduling algorithms, Computers & Industrial Engineering, 46(4), 679-696.
Al Shalabi, L., Shaaban, Z., & Kasasbeh, B., 2006, Data mining: a preprocessing engine, Journal of Computer Science, 2(9), 735-739.
Anderson, D., & McNeill, G., 1992, Artificial neural networks technology, Kaman Sciences Corporation, 258(6), 1-83.
Ansari, S., 2020, Building Computer Vision Applications Using Artificial Neural Networks, Apress, New York.
Anthimopoulos, M., Christodoulidis, S., Ebner, L., Christe, A., & Mougiakakou, S., 2013, Lung pattern classification for interstitial lung diseases using a deep convolutional Neural Network, IEEE Transactions on Medical Imaging, 35(5), 1207-1216.
Bergstra, J., & Bengio, Y., 2012, Random search for hyper-parameter optimization, Journal of Machine Learning Research, 13(2) , 15-35.
Blazewicz, J., Domschke, W., & Pesch, E., 1996, The job shop scheduling problem: conventional and new solution techniques, European Journal of Operational Research, 93(1), 1-33.
Bock, S., & Weiß, M., 2019. A proof of local convergence for the adam optimizer. Paper presented at the 2019 International Joint Conference on Neural Networks (IJCNN) (pp. 1-8).
Calis, B., & Bulkan, S., 2015, A research survey: review of AI solution strategies of job shop scheduling problem, Journal of Intelligent Manufacturing, 26(5), 961-973.
Clevert, D. A., Unterthiner, T., & Hochreiter, S., 2015, Fast and accurate deep network learning by exponential linear units (Elus), arXiv preprint arXiv:1511.07289.
Conway, R. W., Miller, L. W., & Maxwell, W. L., 2003, Theory of Scheduling, Addison-Wesley Pub. Co, United States.
Dongare, A., Kharde, R., & Kachare, A. D., 2012, Introduction to artificial neural network, International Journal of Engineering and Innovative Technology (IJEIT), 2(1), 189-194.
Gupta, N., 2013, Artificial neural network, Network and Complex Systems, 3(1), 24-28.
Hagan, M. T., Demuth, H. B., & Beale, M., 1997, Neural Network Design, PWS Publishing Co, Boston.
Heaton, J., 2005, Introduction to Neural Networks with Java, Heaton Research, Washington.
Hopp, W. J., & Spearman, M. L., 2011, Factory Physics, 3rd Edition, Waveland Press, Long Grove, Illion.
Huang, S. H., & Zhang, H. C., 1995, Neural-expert hybrid approach for intelligent manufacturing: a survey, Computers in Industry, 26(2), 107-126.
Iqbal, A., He, N., Li, L., & Dar, N. U., 2007, A fuzzy expert system for optimizing parameters and predicting performance measures in hard-milling process, Expert Systems with Applications, 32(4), 1020-1027.
Jabbar, H., & Khan, R. Z., 2015, Methods to avoid overfitting and underfitting in supervised machine learning (Comparative study), Computer Science, Communication and Instrumentation Devices (pp. 163-172). Research Publishing Services, India.
Jackson, P., 1986, Introduction to Expert Systems, Addison-Wesley Pub. Co, United States.
Kagermann, H., Wahlster, W. & Helbig, J., 2013, Recommendations for Implementing the Strategic Initiative Industrie 4.0 -- Securing the Future of German Manufacturing Industry ; final report of the Industrie 4.0 Working Group, Forschungsunion.
Kals, R. T., 1999, Fundamentals on the Miniaturization of Sheet Metal Working Processes, Meisenbach, Bamberg.
Kang, H. S., Lee, J. Y., Choi, S., Kim, H., Park, J. H., Son, J. Y., Kim, B. H., & Do Noh, S., 2016, Smart manufacturing: past research, present findings, and future directions, International Journal of Precision Engineering and Manufacturing-green Technology, 3(1), 111-128.
Karsoliya, S., 2012, Approximating number of hidden layer neurons in multiple hidden layer BPNN architecture, International Journal of Engineering Trends and Technology, 3(6), 714-717.
Kashid, S., & Kumar, S., 2012, Applications of artificial neural network to sheet metal work- a review, American Journal of Intelligent Systems, 2(7), 168-176.
Koehrsen, W., 2018, Overfitting v.s. underfitting: a complete example, Towards Data Science, 6(1), 800-818.
Kohonen, T., 1988, An introduction to neural computing, Neural networks, 1(1), 3-16.
Kotsiantis, S. B., Kanellopoulos, D., & Pintelas, P. E., 2006, Data preprocessing for supervised leaning, International Journal of Computer Science, 3(5), 111-117.
Lawrence, S., & Giles, C. L., 2000. Overfitting and neural networks: Conjugate gradient and backpropagation. Paper presented at the Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks. IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium (pp. 114-119).
Li, J., Cheng, J. H., Shi, J. Y., & Huang, F., 2012, Brief introduction of back propagation (BP) neural network algorithm and its improvement. In Advances in Computer Science and Information Engineering (pp. 553-558). Springer, Berlin, Heidelberg.
Liao, L. W., & Sheen, G. J., 2008, Parallel machine scheduling with machine availability and eligibility constraints, European Journal of Operational Research, 184(2), 458-467.
Mijwel, M. M., 2018, Artificial neural networks advantages and disadvantages, Network and Complex Systems, 10(5), 38-47.
Nagori, V., & Trivedi, B., 2014, Types of expert Ssystem: Comparative study, Asian Journal of Computer and Information Systems, 2(2), 555-561.
Negnevitsky, M., 2005, Artificial Intelligence: A Guide to Intelligent Systems, 3rd Edition, Prentice Hall, London, England.
Nilsson, N. J., 2014, Principles of Artificial Intelligence, Morgan Kaufmann, Burlington, Massachusetts.
Pandit, M., 2013, Expert system–a review article, International Journal of Engineering Sciences & Research Technology, 2(6), 1583-1585.
Phansalkar, V., & Sastry, P., 1994, Analysis of the back-propagation algorithm with momentum, IEEE Transactions on Neural Networks, 5(3), 505-506.
Pinedo, M. L., 1995, Scheduling: Theory, Algorithms, and Systems, Springer, New York.
Pontes, F. J., Amorim, G., Balestrassi, P. P., Paiva, A., & Ferreira, J. R., 2016, Design of experiments and focused grid search for neural network parameter optimization, Neurocomputing, 186(3), 22-34.
Prechelt, L., 1998, Early stopping-but when? . In Neural Networks: Tricks of the trade (pp. 55-69). Springer, Berlin, Heidelberg.
Qiumei, Z., Dan, T., & Fenghua, W., 2019, Improved convolutional neural network based on fast exponentially linear unit activation function, IEEE Access, 7, 151359-151367.
Rabelo, L. C., & Alptekin, S., 1989, Integrating scheduling and control functions in computer integrated manufacturing using artificial intelligence, Computers & Industrial Engineering, 17(3), 101-106.
Sazli, M. H., 2006, A brief review of feed-forward neural networks, IEEE Transactions on Neural Networks, 8(2), 220-240.
Scherer, D., Yang, Z., & Hoffmann, H., 2010. Driving–a flexible manufacturing method for individualized sheet metal products. In Key Engineering Materials, 447(8), 795-800.
Sharma, S., 2017, Activation functions in neural networks, Towards Data Science, 6(12), 310-316.
Sokolova, M., & Lapalme, G., 2009, A systematic analysis of performance measures for classification Tasks, Information Processing & Management, 45(4), 427-437.
Tao, F., Qi, Q., Liu, A., & Kusiak, A., 2018, Data-driven smart manufacturing, Journal of Manufacturing Systems, 48(6), 157-169.
Tharwat, A., 2020, Classification assessment methods, Applied Computing and Informatics, 17(1), 168-192.
Tsai, C. W., Huang, W. C., Chiang, M. H., Chiang, M. C., & Yang, C. S., 2014, A hyper-heuristic scheduling algorithm for cloud, IEEE Transactions on Cloud Computing, 2(2), 236-250.
Udo, G. J., 1992, Neural networks applications in manufacturing processes, Computers & Industrial Engineering, 23(14), 97-100.
Walczak, S., & Cerpa, N., 1999, Heuristic principles for the design of artificial neural networks, Information and Software Technology, 41(2), 107-117.
Wang, S. C., 2003, Artificial neural network. In Interdisciplinary Computing in Java Programming (pp. 81-100). Springer, Boston, MA.
Wortman, J., Euwe, M., TAAL, M., & Wiers, V., 1996, A review of capacity planning techniques within standard software packages, Production Planning & Control, 7(2), 117-128.
Yahia, M., Mahmod, R., Sulaiman, N., & Ahmad, F., 2000, Rough neural expert systems, Expert Systems with Applications, 18(2), 87-99.
Yilbas, B., 2008, Laser cutting of thick sheet metals: effects of cutting parameters on kerf size variations, Journal of Materials Processing Technology, 201(2), 285-290.