Hostname: page-component-f554764f5-8cg97 Total loading time: 0 Render date: 2025-04-20T09:30:53.576Z Has data issue: false hasContentIssue false
  • English
  • Français

A semi-supervised anomaly detection approach for detecting mechanical failures

Published online by Cambridge University Press:  13 November 2024

Colin Soete Open the ORCID record for Colin Soete [Opens in a new window] ,
Michaël Rademaker  and
Sofie Van Hoecke
Colin Soete*
Affiliation:
IDLab, Ghent University, IMEC, Technologiepark-Zwijnaarde 126, Ghent, 9052, Belgium
Michaël Rademaker
Affiliation:
IDLab, Ghent University, IMEC, Technologiepark-Zwijnaarde 126, Ghent, 9052, Belgium
Sofie Van Hoecke
Affiliation:
IDLab, Ghent University, IMEC, Technologiepark-Zwijnaarde 126, Ghent, 9052, Belgium
*
Corresponding author: Colin Soete; Email: [email protected]
Get access Rights & Permissions [Opens in a new window]

Abstract

Predictive maintenance attempts to prevent unscheduled downtime by scheduling maintenance before expected failures and/or breakdowns while maximally optimizing uptime. However, this is a non-trivial problem, which requires sufficient data analytics knowledge and labeled data, either to design supervised fault detection models or to evaluate the performance of unsupervised models. While today most companies collect data by adding sensors to their machinery, the majority of this data is unfortunately not labeled. Moreover, labeling requires expert knowledge and is very cumbersome. To solve this mismatch, we present an architecture that guides experts, only requiring them to label a very small subset of the data compared to today’s standard labeling campaigns that are used when designing predictive maintenance solutions. We use auto-encoders to highlight potential anomalies and clustering approaches to group these anomalies into (potential) failure types. The accompanied dashboard then presents the anomalies to domain experts for labeling. In this way, we enable domain experts to enrich routinely collected machine data with business intelligence via a user-friendly hybrid model, combining auto-encoder models with labeling steps and supervised models. Ultimately, the labeled failure data allows for creating better failure prediction models, which in turn enables more effective predictive maintenance. More specifically, our architecture gets rid of cumbersome labeling tasks, allowing companies to make maximum use of their data and expert knowledge to ultimately increase their profit. Using our methodology, we achieve a labeling gain of 90% at best compared to standard labeling tasks.

Keywords

labelingpredictive maintenancesemi-supervised learninganomaly detection
Type
Research Article
Information
AI EDAM , Volume 38 , 2024 , e16
Check for updates
Copyright
© The Author(s), 2024. Published by Cambridge University Press

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Article purchase

Temporarily unavailable

References

Bonte, Pieter, Vanden Hautte, Sander, Lejon, Annelies, Ledoux, Veerle, De Turck, Filip, Van Hoecke, Sofie, and Ongenae, Femke. 2020. Unsupervised anomaly detection for communication networks: an autoencoder approach. In Iot streams for data-driven predictive maintenance and iot, edge, and mobile for embedded machine learning, edited by Gama, Joao, 1325:160–172. Ghent, Belgium: Springer. ISBN: 9783030667696. https://doi.org/10.1007/978-3-030-66770-2_12.CrossRefGoogle Scholar
Chakraborty, Supriyo, Tomsett, Richard, Raghavendra, Ramya, Harborne, Daniel, Alzantot, Moustafa, Cerutti, Federico, Srivastava, Mani, et al. 2017. Interpretability of deep learning models: a survey of results. In 2017 IEEE Smartworld, ubiquitous intelligence computing, advanced trusted computed, scalable computing communications, cloud big data computing, internet of people and smart city innovation (smartworld/scalcom/uic/atc/cbdcom/iop/sci), 1–6. https://doi.org/10.1109/UIC-ATC.2017.8397411.CrossRefGoogle Scholar
Christ, Maximilian, Braun, Nils, Neuffer, Julius, and Kempa-Liehr, Andreas W.. 2018. Time series feature extraction on basis of scalable hypothesis tests (tsfresh–a python package). Neurocomputing 307:72–77. ISSN: 0925-2312. https://doi.org/10.1016/j.neucom.2018.03.067. https://www.sciencedirect.com/science/article/pii/S0925231218304843.CrossRefGoogle Scholar
De Paepe, Dieter, Vanden Hautte, Sander, Steenwinckel, Bram, De Turck, Filip, Ongenae, Femke, Janssens, Olivier, and Van Hoecke, Sofie. 2020. A generalized matrix profile framework with support for contextual series analysis. Engineering Applications of Artificial Intelligence 90:103487. issn: 0952-1976. https://doi.org/10.1016/j.engappai.2020.103487. https://www.sciencedirect.com/science/article/pii/S0952197620300087.CrossRefGoogle Scholar
Fährmann, Daniel, Damer, Naser, Kirchbuchner, Florian, and Kuijper, Arjan. 2022. Lightweight long short-term memory variational auto-encoder for multivariate time series anomaly detection in industrial control systems. Sensors 22 (8): 2886.CrossRefGoogle ScholarPubMed
Gittler, Thomas, Glasder, Magnus, Öztürk, Elif, Lüthi, Michel, Weiss, Lukas, and Wegener, Konrad. 2021. International conference on advanced and competitive manufacturing technologies milling tool wear prediction using unsupervised machine learning. The International Journal of Advanced Manufacturing Technology 117, no. 7 (December): 2213–2226. ISSN: 1433-3015. https://doi.org/10.1007/s00170-021-07281-2.CrossRefGoogle Scholar
Hindy, Hanan, Atkinson, Robert, Tachtatzis, Christos, Colin, Jean-Noël, Bayne, Ethan, and Bellekens, Xavier. 2020. Towards an effective zero-day attack detection using outlier-based deep learning techniques. https://doi.org/10.48550/arXiv.2006.15344.CrossRefGoogle Scholar
Kotsiantis, Sotiris B, Zaharakis, Ioannis D, and Pintelas, Panayiotis E. 2006. Machine learning: a review of classification and combining techniques. Artificial Intelligence Review 26 (3): 159–190.CrossRefGoogle Scholar
Li, Jinbo, Izakian, Hesam, Pedrycz, Witold, and Jamal, Iqbal. 2021. Clustering-based anomaly detection in multivariate time series data. Applied Soft Computing 100:106919. ISSN: 1568-4946. https://doi.org/10.1016/j.asoc.2020.106919. https://www.sciencedirect.com/science/article/pii/S1568494620308577.CrossRefGoogle Scholar
Li, Xiang, Li, Xu, and Ma, Hui. 2020. Deep representation clustering-based fault diagnosis method with unsupervised data applied to rotating machinery. Mechanical Systems and Signal Processing 143:106825. ISSN: 0888-3270. https://doi.org/10.1016/j.ymssp.2020.106825. https://www.sciencedirect.com/science/article/pii/S0888327020302119.CrossRefGoogle Scholar
Li, Xuhong, Xiong, Haoyi, Li, Xingjian, Wu, Xuanyu, Zhang, Xiao, Liu, Ji, Bian, Jiang, and Dou, Dejing. 2022. Interpretable deep learning: interpretation, interpretability, trustworthiness, and beyond. Knowledge and Information Systems 64 (12): 3197–3234.CrossRefGoogle Scholar
Li, Zhe, Li, Jingyue, Wang, Yi, and Wang, Kesheng. 2019. A deep learning approach for anomaly detection based on sae and lstm in mechanical equipment. International Journal of Advanced Manufacturing Technology 103, no. 1 (July): 499–510. ISSN: 1433-3015. https://doi.org/10.1007/s00170-019-03557-w.CrossRefGoogle Scholar
Lu, Yue, Srinivas, Thirumalai Vinjamoor Akhil, Nakamura, Takaaki, Imamura, Makoto, and Keogh, Eamonn. 2023. Matrix profile xxx: madrid: a hyper-anytime and parameter-free algorithm to find time series anomalies of all lengths. In 2023 ieee international conference on data mining (icdm), 1199–1204. IEEE. https://doi.org/10.1109/ICDM58522.2023.00148.CrossRefGoogle Scholar
McInnes, Leland, Healy, John, and Melville, James. 2018. Umap: uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426.CrossRefGoogle Scholar
Mulders, Michel, and Haarman, Mark. 2017. Predictive maintenance 4.0: predict the unpredictable. Edited by PwC. Accessed March 6, 2021. https://www.pwc.be/en/documents/20171016-predictive-maintenance-4-0.pdf.Google Scholar
Nasir, Vahid, and Sassani, Farrokh. 2021. A review on deep learning in machining and tool monitoring: methods, opportunities, and challenges. The International Journal of Advanced Manufacturing Technology 115, no. 9 (August): 2683–2709. ISSN: 1433-3015. https://doi.org/10.1007/s00170-021-07325-7.CrossRefGoogle Scholar
Ochella, Sunday, Shafiee, Mahmood, and Sansom, Chris. 2021. Adopting machine learning and condition monitoring pf curves in determining and prioritizing high-value assets for life extension. Expert Systems with Applications 176:114897. ISSN: 0957-4174. https://doi.org/10.1016/j.eswa.2021.114897. https://www.sciencedirect.com/science/article/pii/S0957417421003389.CrossRefGoogle Scholar
Pang, Guansong, Shen, Chunhua, Cao, Longbing, and Van Den Hengel, Anton. 2021. Deep learning for anomaly detection: a review. ACM computing surveys (CSUR) 54 (2): 1–38.CrossRefGoogle Scholar
Pu, Guo, Wang, Lijuan, Shen, Jun, and Dong, Fang. 2020. A hybrid unsupervised clustering-based anomaly detection method. Tsinghua Science and Technology 26 (2): 146–153.CrossRefGoogle Scholar
Schubert, Erich, Sander, Jörg, Ester, Martin, Kriegel, Hans Peter, and Xu, Xiaowei. 2017. Dbscan revisited, revisited: why and how you should (still) use dbscan. ACM Transactions on Database Systems (TODS) 42 (3): 1–21.CrossRefGoogle Scholar
Serradilla, Oscar, Zugasti, Ekhi, de Okariz, Julian Ramirez, Rodriguez, Jon, and Zurutuza, Urko. 2021. Adaptable and explainable predictive maintenance: semi-supervised deep learning for anomaly detection and diagnosis in press machine data. Applied Sciences 11 (16). issn: 2076-3417. https://doi.org/10.3390/app11167376. https://www.mdpi.com/2076-3417/11/16/7376.CrossRefGoogle Scholar
Van der Maaten, Laurens, and Hinton, Geoffrey. 2008. Visualizing data using t-sne. Journal of machine learning research 9 (86): 2579–2605. http://jmlr.org/papers/v9/vandermaaten08a.html.Google Scholar
Weichert, Dorina, Link, Patrick, Stoll, Anke, Rüping, Stefan, Ihlenfeldt, Steffen, and Wrobel, Stefan. 2019. A review of machine learning for the optimization of production processes. The International Journal of Advanced Manufacturing Technology 104 (5): 1889–1902.CrossRefGoogle Scholar
Yan, Jihong, Meng, Yue, Lu, Lei, and Li, Lin. 2017. Industrial big data in an industry 4.0 environment: challenges, schemes, and applications for predictive maintenance. IEEE Access 5:23484–23491. https://doi.org/10.1109/ACCESS.2017.2765544.CrossRefGoogle Scholar