依據中華民國經濟部國貿局2021年出口資料，面板液晶裝置對外出口產值約8,592佰萬美元，佔比為1.925%，排名第7，在財政部統計處2020年近期經貿與稅收情勢的資料顯示，液晶裝置及其零件出口排名始終維持前10位，可見其對國內產業的重要性，近年在大陸政府強力財政支持下大陸面板業者的崛起也已影響台廠的生存空間。從工廠經營管理的角度來探討如何提升台廠的競爭力，早期有豐田生產方式注重生產的七大浪費，後有限制理論關注生產系統的瓶頸點，以及運用六標準差的手法來改善控制生產流程中對產品品質會產生變異的因素，近期在進入工業4.0的時代下，大數據、物聯網、電腦算力以及人工智慧演算法不斷的蓬勃發展，如何利用各種不同的人工智慧程度其來提升工廠的競爭力，成為一個值得探討的議題。
本研究考量目前面板製造業在台灣的重要性，選取了此類製程中，有一重要任務塗膠製程的精密噴嘴內部影像可以將其自動化，再利用人工輔助智慧協助將精密噴嘴的良窳進行分類，此研究透過類神經網路模型對其良窳進行判斷，取代原先人工的判定方法。在方法的選擇上，研究中選擇ResNet-50及Vanilla兩模型測試的F1結果分別為85.3%及90.9%。將Vanilla模型所判定良好的精密噴嘴上機實測塗膠的品質結果，其塗膠線寬製程能力標準差為51.2 um，優於人工判定方法的標準差55.1 um，此一方法的應用在人力的節省上，單機每月可減少人力調整參數工時792分鐘，在產能方面單機每月可增加720分鐘的生產時間，結果顯示Vanilla模型判定精密噴嘴的良窳相對於人工判定方法的均一性較優，期能經由此論文對製造管理上提出一有效的貢獻。

Based on the recent economic, trade and tax situation data of Finance in 2020, the export ranking of LCD display has always remained in the top 10, which shows its importance to the industry. In recent years, the rise of Chinese display manufacturers has affected the living space of Taiwanese companies. Discuss how to enhance the competitiveness of Taiwanese companies, recently, in the era of Industry 4.0, big data, artificial intelligence algorithms continue to flourish. How to use various levels of artificial intelligence to enhance the competitiveness of factories has become a topic worth discussing.
This research selects a bottleneck process, the internal image of the precision nozzle of the sealing process can be image synthesis automated, and then uses assisted intelligence to assist in classifying the quality of the nozzle , this study uses a neural network model to judge its goodness, replacing the manual judgment method.
ResNet-50 and Vanilla models were selected for testing, and their F1 results were 85.3% and 90.9%. The standard deviation of the seal width of Vanilla is 51.2 μm, which is better than 55.1 μm of the manual judgment. This application of this method can save manpower,. as we calculate a single machine can reduce adjustment time 792 minutes per month. In production capacity, a single machine can increase 720 minutes run time per month. It is hoped that this paper can make an effective contribution to manufacturing management.
Key words: neural network, sealing process, precision nozzle, industry 4.0, assisted intelligence

A. Krizhevsky, I. Sutskever, & G. Hinton. (2017). ImageNet Classification with Deep Convolutional Neural Networks. Communications of the Acm, 60(6), 84-90. https://doi.org/10.1145/3065386
A. Mass. (2013). Rectifier Nonlinearities Improve Neural Network Acoustic Models. Proceedings of the 30 th International Conference on Machine Learning, Atlanta, Georgia, USA, 2013, 28.
Arini, A., Bahaweres, R., & Haq, J. (2022, 29-30 Jan. 2022). Quick Classification of Xception And Resnet-50 Models on Deepfake Video Using Local Binary Pattern. 2021 International Seminar on Machine Learning, Optimization, and Data Science (ISMODE), Jakarta, Indonesia.
D. Rumelhart, & J. McClelland. (1988). A General Framework for Parallel Distributed Processing. IEEE New York.
F. Rosenblatt. (1958). The Perceptron - A Probabilistic Model for Information-Storage and Organization in the Brain. Psychological review, 65(6), 386-408. https://doi.org/10.1037/h0042519
Haifeng, J., Song, Q., & Hu, X. (2019). Auto-keras: An Efficient Neural Architecture Search System. Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, Anchorage AK USA.
He, K., Zhang, X., Ren, S., & Sun, J. (2016, 27-30 June 2016). Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.
Huang, G., Liu, Z., Maaten, L. V. D., & Weinberger, K. Q. (2017, 21-26 July 2017). Densely Connected Convolutional Networks. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA.
Jaju, S., & Chandak, M. (2022, 9-11 May 2022). A Transfer Learning Model Based on ResNet-50 for Flower Detection. 2022 International Conference on Applied Artificial Intelligence and Computing (ICAAIC), Salem, India.
K. Wan, C. Wong, H. Ip, D. Fan, P. Yuen, H. Fong, & M. Ying. (2021). Evaluation of the Performance of Traditional Machine lLearning Algorithms, Convolutional Neural Network and AutoML Vision in Ultrasound Breast Lesions Classification: A Comparative Study. Quantitive Imaging in Medicine and Surgery, 11(4), 1381-1393. https://doi.org/10.21037/qims-20-922
L. Sobrecueva. (2021). AutoML 自動化機器學習：用 AutoKeras 超輕鬆打造高效能 AI 模型 (許佩瑩, Trans.). 旗標科技股份有限公司.
Nyarko, B. N. E., Bin, W., Zhou, J., Agordzo, G. K., Odoom, J., & Koukoyi, E. (2022, 6-9 June 2022). Comparative Analysis of AlexNet, Resnet-50, and Inception-V3 Models on Masked Face Recognition. 2022 IEEE World AI IoT Congress (AIIoT), Seattle, WA, USA.
PricewaterhouseCoopers. (2018). 抓住AI商機. https://www.pwc.tw/zh/news/press-release/press-20180201.html
Taiwan01., C. A. (2019). CNN入門-Overfitting. https://cinnamonaitaiwan.medium.com/cnn%E5%85%A5%E9%96%80-overfitting-d10acd15ec21
Taiwan02., C. A. (2019). 深度學習：CNN原理. https://cinnamonaitaiwan.medium.com/%E6%B7%B1%E5%BA%A6%E5%AD%B8%E7%BF%92-cnn%E5%8E%9F%E7%90%86-keras%E5%AF%A6%E7%8F%BE-432fd9ea4935
Zahisham, Z., Lee, C. P., & Lim, K. M. (2020, 26-27 Sept. 2020). Food Recognition with ResNet-50. 2020 IEEE 2nd International Conference on Artificial Intelligence in Engineering and Technology (IICAIET), Kota Kinabalu, Malaysia.
三津村直貴. (2018). 圖解 AI 人工智慧大未來 (陳子安, Trans.). 旗標出版社.
吳堃齊. (2021). 改良ResNet-50的糖尿病視網膜病變檢測 明志科技大學]. 新北市. https://hdl.handle.net/11296/gny3wz
邱強. (2021a). 零錯誤 : 全球頂尖企業都採用的科技策略. 遠見天下文化出版有限公司.
邱強. (2021b). 零錯誤決策 : 快速提升與個人競爭力 (張玄竺, Trans.). 遠見天下文化出版股份有限公司.
門田安宏. (2007). 豐田現場管理方式-豐田巨額利潤的秘密 (李偉 & 李晴, Trans.; 吳玉萍 & 劉嘉怡, Eds.). 東方出版社
東方音像電子出版社.
徐潔磐. (2021). 人工智慧導論 (陳俊成 & 何敏煌, Trans.). 台科大圖書股份有限公司.
財政部統計處. (109). 近期經貿與稅收情勢(109年7月24日) - 專題：我國出口領先廠商分析. https://www.mof.gov.tw/singlehtml/384fb3077bb349ea973e7fc6f13b6974?cntId=0d37921f11944c7e8f68b08b38a8bfe1
高德拉特. (2002). 關鍵鏈 (羅嘉穎, Trans.). 天下遠見出版股份有限公司.
高德拉特. (2010). 絕不是靠運氣 (周怜利, Trans.). 天下遠見出版股份有限公司.
高德拉特. (2011). 目標 (齊若蘭, Trans.). 天下遠見出版股份有限公司.
麥可.喬治, 大衛.羅蘭, 馬克.普萊斯, & 約翰.馬西. (2009). 精實六標準差工具手冊 (丁惠民, Trans.). 美商麥格羅.希爾國際股份有限公司 台灣分公司.
傑佛瑞.萊克. (2006). 豐田模式. 美商麥格羅.希爾國際股份有限公司 台灣分公司.
黃文, 成波, & 柴寶亭. (2007). 豐田管理模式全集. 武漢大學出版社.
A. Krizhevsky, I. Sutskever, & G. Hinton. (2017). ImageNet Classification with Deep Convolutional Neural Networks. Communications of the Acm, 60(6), 84-90. https://doi.org/10.1145/3065386
A. Mass. (2013). Rectifier Nonlinearities Improve Neural Network Acoustic Models. Proceedings of the 30 th International Conference on Machine Learning, Atlanta, Georgia, USA, 2013, 28.
Arini, A., Bahaweres, R., & Haq, J. (2022, 29-30 Jan. 2022). Quick Classification of Xception And Resnet-50 Models on Deepfake Video Using Local Binary Pattern. 2021 International Seminar on Machine Learning, Optimization, and Data Science (ISMODE), Jakarta, Indonesia.
D. Rumelhart, & J. McClelland. (1988). A General Framework for Parallel Distributed Processing. IEEE New York.
F. Rosenblatt. (1958). The Perceptron - A Probabilistic Model for Information-Storage and Organization in the Brain. Psychological review, 65(6), 386-408. https://doi.org/10.1037/h0042519
Haifeng, J., Song, Q., & Hu, X. (2019). Auto-keras: An Efficient Neural Architecture Search System. Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, Anchorage AK USA.
He, K., Zhang, X., Ren, S., & Sun, J. (2016, 27-30 June 2016). Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.
Huang, G., Liu, Z., Maaten, L. V. D., & Weinberger, K. Q. (2017, 21-26 July 2017). Densely Connected Convolutional Networks. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA.
Jaju, S., & Chandak, M. (2022, 9-11 May 2022). A Transfer Learning Model Based on ResNet-50 for Flower Detection. 2022 International Conference on Applied Artificial Intelligence and Computing (ICAAIC), Salem, India.
K. Wan, C. Wong, H. Ip, D. Fan, P. Yuen, H. Fong, & M. Ying. (2021). Evaluation of the Performance of Traditional Machine lLearning Algorithms, Convolutional Neural Network and AutoML Vision in Ultrasound Breast Lesions Classification: A Comparative Study. Quantitive Imaging in Medicine and Surgery, 11(4), 1381-1393. https://doi.org/10.21037/qims-20-922
L. Sobrecueva. (2021). AutoML 自動化機器學習：用 AutoKeras 超輕鬆打造高效能 AI 模型 (許佩瑩, Trans.). 旗標科技股份有限公司.
Nyarko, B. N. E., Bin, W., Zhou, J., Agordzo, G. K., Odoom, J., & Koukoyi, E. (2022, 6-9 June 2022). Comparative Analysis of AlexNet, Resnet-50, and Inception-V3 Models on Masked Face Recognition. 2022 IEEE World AI IoT Congress (AIIoT), Seattle, WA, USA.
PricewaterhouseCoopers. (2018). 抓住AI商機. https://www.pwc.tw/zh/news/press-release/press-20180201.html
Taiwan01., C. A. (2019). CNN入門-Overfitting. https://cinnamonaitaiwan.medium.com/cnn%E5%85%A5%E9%96%80-overfitting-d10acd15ec21
Taiwan02., C. A. (2019). 深度學習：CNN原理. https://cinnamonaitaiwan.medium.com/%E6%B7%B1%E5%BA%A6%E5%AD%B8%E7%BF%92-cnn%E5%8E%9F%E7%90%86-keras%E5%AF%A6%E7%8F%BE-432fd9ea4935
Zahisham, Z., Lee, C. P., & Lim, K. M. (2020, 26-27 Sept. 2020). Food Recognition with ResNet-50. 2020 IEEE 2nd International Conference on Artificial Intelligence in Engineering and Technology (IICAIET), Kota Kinabalu, Malaysia.
三津村直貴. (2018). 圖解 AI 人工智慧大未來 (陳子安, Trans.). 旗標出版社.
吳堃齊. (2021). 改良ResNet-50的糖尿病視網膜病變檢測 明志科技大學]. 新北市. https://hdl.handle.net/11296/gny3wz
邱強. (2021a). 零錯誤 : 全球頂尖企業都採用的科技策略. 遠見天下文化出版有限公司.
邱強. (2021b). 零錯誤決策 : 快速提升與個人競爭力 (張玄竺, Trans.). 遠見天下文化出版股份有限公司.
門田安宏. (2007). 豐田現場管理方式-豐田巨額利潤的秘密 (李偉 & 李晴, Trans.; 吳玉萍 & 劉嘉怡, Eds.). 東方出版社
東方音像電子出版社.
徐潔磐. (2021). 人工智慧導論 (陳俊成 & 何敏煌, Trans.). 台科大圖書股份有限公司.
財政部統計處. (109). 近期經貿與稅收情勢(109年7月24日) - 專題：我國出口領先廠商分析. https://www.mof.gov.tw/singlehtml/384fb3077bb349ea973e7fc6f13b6974?cntId=0d37921f11944c7e8f68b08b38a8bfe1
高德拉特. (2002). 關鍵鏈 (羅嘉穎, Trans.). 天下遠見出版股份有限公司.
高德拉特. (2010). 絕不是靠運氣 (周怜利, Trans.). 天下遠見出版股份有限公司.
高德拉特. (2011). 目標 (齊若蘭, Trans.). 天下遠見出版股份有限公司.
麥可.喬治, 大衛.羅蘭, 馬克.普萊斯, & 約翰.馬西. (2009). 精實六標準差工具手冊 (丁惠民, Trans.). 美商麥格羅.希爾國際股份有限公司 台灣分公司.
傑佛瑞.萊克. (2006). 豐田模式. 美商麥格羅.希爾國際股份有限公司 台灣分公司.
黃文, 成波, & 柴寶亭. (2007). 豐田管理模式全集. 武漢大學出版社.
經濟部國際貿易局. (2021). 中華民國進出口貿易統計. https://cuswebo.trade.gov.tw/FSC6030I/FSC6030I?menuURL=FSC6030I
維基百科. (2022a). 人工神經網路. https://zh.wikipedia.org/zh-tw/%E4%BA%BA%E5%B7%A5%E7%A5%9E%E7%BB%8F%E7%BD%91%E7%BB%9C
維基百科. (2022b). 工業4.0. https://zh.wikipedia.org/zh-tw/%E5%B7%A5%E6%A5%AD4.0