在深度學習快速發展的現在，以數據驅動的方法應用在資本密集型半導體製造中已有極大成功。在文獻中已有很多關於Wafer Bin Map（WBM）上辨識研究，其主要用於識別WBM上的缺陷模式和尋找造成錯誤的根本原因，藉以降低生產時不良率所造成的成本損失。過去WBM辨識的研究中僅會給出分類的結果，並沒有提供預警機制。因此，本研究建立了在封裝廠中，覆晶封裝製程下的Strip Bin Map (SBM) 預測系統。通過系統中建立的SBM預測模組和判別模組，在給定前站已經出現過的SBM的情況下，此系統可以預測後續站點可能會出現的SBM圖像以及其所屬的缺陷模式。此時系統會提醒工程師處理異常，以減少製造資源的浪費。在實務上，並非所有製程站點都有進行功能測試，因此，本研究透過貝式定理的概念去模擬出缺少的站點。在衡量的部分，我們透過相似性指標以及所提出的成長軌跡指標，建構了一個可以衡量時空效應的時空正確性指標，用以衡量系統的輸出。最後，本研究對台灣半導體封裝製造商進行了一項實證研究。結果表明，所提出的系統在預測出的SBM上，時空正確度的分數比隨機生成高了12倍。也就是說，此系統能夠有效預測未來站點中出現的SBM圖像以及其所屬的缺陷模式。而所提出的時空正確性指標也能夠同時衡量系統預測的SBM在時空效應下的正確性。

With the rapid development of deep learning, the application of data-driven methods in semiconductor manufacturing has achieved great success. There are many studies on Wafer Bin Map (WBM) recognition in the literature, which are mainly used to identify the failure modes of WBM and find the root cause to reduce yield loss during production. However, past WBM recognition studies only gave classification results. Therefore, this study develops a Strip Bin Map (SBM) prognostic system under the flip-chip bonding process. The SBM prediction and recognition module established in this system can predict the future SBM image and the defect type that appears in the subsequent stations. At this point, the system will remind engineers to handle defects to reduce the waste of manufacturing resources. In practice, not all process stations have functional tests. Therefore, we use the concept of the Bayesian theorem to simulate the missing station data. In the measurement part, we construct the spatio-temporal accuracy metrics through the similarity metrics and the proposed growth trajectory metrics, which can measure the spatio-temporal effect of the system output. Finally, this study conducts an empirical study of Taiwanese semiconductor packaging manufacturers. The results show that the SBM predicted by the proposed system is 12 times higher than the random generation on the score of spatio-temporal accuracy. That is, the system can effectively predict the SBM images that will appear in future stations and the defect patterns to which they belong. The proposed spatio-temporal accuracy metrics can also measure the correctness of the SBM predicted by the system under the spatio-temporal effect.

[1] Abbasimehr, H., et al. (2020). "An optimized model using LSTM network for demand forecasting." Computers & Industrial Engineering 143: 106435.
[2] Azzalini, D., et al. (2021). "A minimally supervised approach based on variational
autoencoders for anomaly detection in autonomous robots." IEEE Robotics and
Automation Letters 6(2): 2985-2992.
[3] Bai, S., et al. (2018). "An empirical evaluation of generic convolutional and
recurrent networks for sequence modeling." arXiv preprint arXiv:1803.01271.
[4] Batool, U., et al. (2021). "A Systematic Review of Deep Learning for Silicon
Wafer Defect Recognition." Ieee Access.
[5] Chen, F.-L. and S.-F. Liu (2000). "A neural-network approach to recognize defect
spatial pattern in semiconductor fabrication." IEEE Transactions on Semiconductor
Manufacturing 13(3): 366-373.
[6] Donahue, J., et al. (2015). Long-term recurrent convolutional networks for visual
recognition and description. Proceedings of the IEEE conference on computer vision and pattern recognition.
[7] Ezzat, A. A., et al. (2021). "A graph-theoretic approach for spatial filtering and its
impact on mixed-type spatial pattern recognition in wafer bin maps." IEEE Transactions on Semiconductor Manufacturing 34(2): 194-206.
[8] Fan, M., et al. (2016). Wafer defect patterns recognition based on OPTICS and
multi-label classification. 2016 IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC), IEEE.
[9] Fan, S. K. S., et al. (2020). "Data-Driven Approach for Fault Detection and Diagnostic in Semiconductor Manufacturing." IEEE Transactions on Automation Science and Engineering 17(4): 1925-1936.
[10] Goodfellow, I., et al. (2014). "Generative adversarial nets." Advances in neural information processing systems 27.
[11] Hansen, M. H., et al. (1997). "Monitoring wafer map data from integrated circuit fabrication processes for spatially clustered defects." Technometrics 39(3): 241-253.
[12] He, K., et al. (2015). "Deep Residual Learning for Image Recognition." arXiv
pre-print server.
[13] Hochreiter, S. and J. Schmidhuber (1997). "Long short-term memory." Neural
computation 9(8): 1735-1780.
[14] Hsieh, Y.-M., et al. (2021). "Convolutional Neural Networks for Automatic
Virtual Metrology." IEEE Robotics and Automation Letters 6(3): 5720-5727.
[15] Hung, S.-Y., et al. (2019). "Data science for delamination prognosis and online
batch learning in semiconductor assembly process." IEEE Transactions on
Components, Packaging and Manufacturing Technology 10(2): 314-324.
[16] Islam, M. Z., et al. (2020). "A combined deep CNN-LSTM network for the
detection of novel coronavirus (COVID-19) using X-ray images." Informatics in
Medicine Unlocked 20: 100412.
[17] Jin, C. H., et al. (2019). "A novel DBSCAN-based defect pattern detection and
classification framework for wafer bin map." IEEE Transactions on Semiconductor
Manufacturing 32(3): 286-292.
[18] Kim, E., et al. (2019). "Fault Detection and Diagnosis Using Self-Attentive
Convolutional Neural Networks for Variable-Length Sensor Data in Semiconductor
Manufacturing." IEEE Transactions on Semiconductor Manufacturing 32(3): 302-309.
[19] Kim, T.-Y. and S.-B. Cho (2019). "Predicting residential energy consumption
using CNN-LSTM neural networks." Energy 182: 72-81.
[20] Kim, T. S. and A. Reiter (2017). Interpretable 3d human action analysis with
temporal convolutional networks. 2017 IEEE conference on computer vision and
pattern recognition workshops (CVPRW), IEEE.
[21] Koshoubu, N., et al. (2000). "Advanced flip chip bonding techniques using
transferred microsolder bumps." IEEE transactions on components and packaging
technologies 23(2): 399-404.
[22] Lea, C., et al. (2017). Temporal convolutional networks for action segmentation
and detection. proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.
[23] Lea, C., et al. (2016). "Temporal Convolutional Networks for Action Segmentation and Detection." arXiv pre-print server.
[24] Lee, C.-Y., et al. (2020). "In-line predictive monitoring framework." IEEE
Transactions on Automation Science and Engineering 18(4): 1668-1678.
[25] Li, B. and L. Han (2013). Distance Weighted Cosine Similarity Measure for Text
Classification, Springer Berlin Heidelberg: 611-618.
[26] Li, B., et al. (2021). "Self-Attention ConvLSTM and Its Application in RUL
Prediction of Rolling Bearings." IEEE Transactions on Instrumentation and
Measurement 70: 1-11.
[27] Lim, S., et al. (2015). Flux challenges in flip-chip die-attach. 2015 IEEE 17th
Electronics Packaging and Technology Conference (EPTC), IEEE.
[28] Lin, Z., et al. (2020). "Self-Attention ConvLSTM for Spatiotemporal Prediction."
Proceedings of the AAAI Conference on Artificial Intelligence 34(07): 11531-11538.
[29] Liu, C.-W. and C.-F. Chien (2013). "An intelligent system for wafer bin map
defect diagnosis: An empirical study for semiconductor manufacturing." Engineering Applications of Artificial Intelligence 26(5-6): 1479-1486.
[30] Lu, G. Y. and D. W. Wong (2008). "An adaptive inverse-distance weighting
spatial interpolation technique." Computers & geosciences 34(9): 1044-1055.
[31] Masci, J., et al. (2011). Stacked Convolutional Auto-Encoders for Hierarchical
Feature Extraction, Springer Berlin Heidelberg: 52-59.
[32] Mikolajczyk, A. and M. Grochowski Data augmentation for improving deep
learning in image classification problem, IEEE.
[33] Nelson, D. M. Q., et al. Stock market's price movement prediction with LSTM
neural networks, IEEE.
[34] Niwattanakul, S., et al. (2013). Using of Jaccard coefficient for keywords
similarity. Proceedings of the international multiconference of engineers and computer scientists.
[35] Piao, M., et al. (2018). "Decision tree ensemble-based wafer map failure pattern
recognition based on radon transform-based features." IEEE Transactions on Semiconductor Manufacturing 31(2): 250-257.
[36] Pratikakis, I., et al. (2012). ICFHR 2012 Competition on Handwritten Document
Image Binarization (H-DIBCO 2012). 2012 International Conference on Frontiers in
Handwriting Recognition: 817-822.
[37] Rostami, H., et al. (2018). "Automatic equipment fault fingerprint extraction for
the fault diagnostic on the batch process data." Applied Soft Computing 68: 972-989.
[38] Shen, P.-C. and C.-Y. Lee (2022). "Wafer Bin Map Recognition with Autoencoder-based Data Augmentation in Semiconductor Assembly Process." IEEE
Transactions on Semiconductor Manufacturing.
[39] Shi, X., et al. (2015). "Convolutional LSTM network: A machine learning
approach for precipitation nowcasting." Advances in neural information processing
systems 28.
[40] Szegedy, C., et al. (2014). "Going Deeper with Convolutions." arXiv pre-print
server.
[41] Wang, C.-H., et al. (2006). "Detection and classification of defect patterns on
semiconductor wafers." IIE transactions 38(12): 1059-1068.
[42] Wang, D., et al. DeepSTCL: A Deep Spatio-temporal ConvLSTM for Travel
Demand Prediction, IEEE.
[43] Zhao, G., et al. Research advances in fault diagnosis and prognostic based on
deep learning, IEEE.