Industry 4.0 Science 39, 2023, 88-94
Predictive Manufacturing – An Intelligent Monitoring System to Detect Anomalies in 3D Printing

Bibtex

Cite as text

						@Article{Uhrich+Lange+Carnot+Schäfer,
							Cite-key = "Uhrich2023Pre", 
							Year= "2023", 
							Number= "1", 
							 Volume= "Industry 4.0 Science 39", 
							Pages= "88-94", 
							Journal   = "Industry 4.0 Science",
							 Title= "Predictive Manufacturing – An Intelligent Monitoring System to Detect Anomalies in 3D Printing", 
							Author= "Benjamin Uhrich, Shirin Lange and Miriam Louise Carnot, University of Leipzig, Martin Schäfer, SIEMENS AG Berlin", 
							Doi= "https://doi.org/10.30844/I4SE.23.1.88", 
							 Abstract= "In selective laser melting, metal powder is melted layer by layer and fused with the already manufactured part. Within this process, defective layers are created, which can be avoided. Such defects can only be detected by various compression and tensile strength experiments after printing is complete. This procedure is costly and inefficient. Therefore, a demonstrator is presented that uses machine learning methods to identify defective layers during the manufacturing process. In addition, the machine operator
is supported with decision recommendations.", 
							 Keywords= "intelligent systems, autoencoders, 3D printing, machine learning, decision support, selective laser melting, SLM, Siemens, LAB color model, Contrast Limited Adaptive Histogram Equalization, CLAHE", 
							}
					
Benjamin Uhrich, Shirin Lange and Miriam Louise Carnot, University of Leipzig, Martin Schäfer, SIEMENS AG Berlin(2023): Predictive Manufacturing – An Intelligent Monitoring System to Detect Anomalies in 3D Printing. Industry 4.0 Science 391(2023), S. 88-94. Online: https://doi.org/10.30844/I4SE.23.1.88 (Abgerufen 28.02.24)

Abstract

Abstract

In selective laser melting, metal powder is melted layer by layer and fused with the already manufactured part. Within this process, defective layers are created, which can be avoided. Such defects can only be detected by various compression and tensile strength experiments after printing is complete. This procedure is costly and inefficient. Therefore, a demonstrator is presented that uses machine learning methods to identify defective layers during the manufacturing process. In addition, the machine operator is supported with decision recommendations.

Keywords

Schlüsselwörter

intelligent systems, autoencoders, 3D printing, machine learning, decision support, selective laser melting, SLM, Siemens, LAB color model, Contrast Limited Adaptive Histogram Equalization, CLAHE

References

Referenzen

[1] Bremen, S.; Meiner, W.; Diatlov, A.: Selective Laser Metling- A manufacturing technology for the future? In: Laser Technik Journal, (2012), S. 33-38
[2] Chang, Y.; Jung, C.; Ke, P.; Song, H.; Hwang, J.: Automatic contrast-limited adaptive histogram equalization with dual gamma correction. IEEE Access (2018), 6: S.11782-11792
[3] Li, Z.; Liu, F.; Yang, W.; Peng, S.; Zhou, J.: A Survey of Convolutional Neural Networks: Analysis, Applications and Prospects, IEEE Transactions on Neural Networks and Learning Systems (2021), Volume 33: S. 6999 -7019
[4] Yao, Y.; Rosasco, L.; Caponnetto, A.: On Early Stopping in Gradient Descent Learning, Constructive Approximation, (2007): S. 289 – 315
[5] Zimmerer, D.; Kohl, S.; Petersen, J; Meier-Hein, F.; Isensee, K.: Context encoding variational autoencoder for unsupervised anomaly detection. CoRR (2018): S. 1-13
[6] An, J.; Cho, S.; Variational autoencoder based anomaly detection using reconstruction probability. Special Lecture on IE (2015), S. 1-18
[7] Wang, Y.; Yao, H.; Zhao, S.: Auto-encoder based dimensionality reduction. Neurocomputing (2016), S. 232-242
[8] Dette, H.; Härdle, W.; Principal component analysis, Multivariate Analysemethoden von Andreas Handl, Springer Verlag (2010), S. 126-147
[9] Uhrich, Benjamin; Schäfer, Martin; Theile, Oliver; Rahm, Erhard (2023): Using Physics-Informed Machine Learning to Optimize 3D Printing Processes. In: Joel Oliveira Correia Vasco, Henrique de Amorim Almeida, Anabela Gonçalves Rodrigues Marto, Carlos Alexandre Bento Capela, Flávio Gabriel Da Silva Craveiro, Helena Maria Da Coelho Rocha Terreiro Galha Bárt et al. (Hg.): Progress in Digital and Physical Manufacturing. Cham: Springer International Publishing (Springer Tracts in Additive Manufacturing), S. 206–221.
[10] Bauer, M; Uhrich, B; Schäfer, M.; Theile, O.; Augenstein, C; Rahm, E.: Mulit-Modal Artificial Intelligence in Additive Manufacturing: Combining Thermal and Camera Images for 3D-Print Quality Monitoring. In: Proceedings of the 25th International Conference on Enterprise Information Systems – SCITEPRESS – Science and Technology Publications; 2023.p. 539–546
[11] Uhrich, Benjamin; Hlubek, Nikolai; Häntschel, Tim; Rahm, Erhard (2023): Using differential equation inspired machine learning for valve faults prediction. In: 2023 IEEE 21st International Conference on Industrial Informatics (INDIN). 2023 IEEE 21st International Conference on Industrial Informatics (INDIN). Lemgo, Germany, 18.07.2023 – 20.07.2023: IEEE, S. 1–8.

 

Most viewed articles

Meist angesehene Beiträge

GITO events | library.gito