Through-Life Monitoring of Resource-constrained Systems and Fleets
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
A Digital Twin (DT) is a representation of a physical system that provides information to make decisions that add economic, social or commercial value. DTs are widely used for prognostics and anomaly detection by continuously comparing measured system behaviour with the DT predictions to identify deviations and estimate degradation. The behaviour of a physical system changes over time; a DT must therefore be continually updated with data from the physical system to reflect its changing behaviour. In this paper, we consider a DT of a complex, non-linear and dynamic system subject to slow nominal degradation, disturbances and the risk of anomalies. The DT runs on a resource-constrained system, making up dating non-trivial due to limitations in computation, storage,
and data transfer bandwidth. Consequently, only a subset of the generated data can be retained or transmitted, making data prioritisation essential. Data must be evaluated online in order to select the most relevant subset with which to perform the update. DT updating must address the continual learning
challenge of adapting to new system behaviours, such as the response to previously unseen operating conditions, while retaining knowledge of previously observed behaviours. This paper presents a framework for updating a data-driven DT of a resource-constrained system. The proposed solution consists of: (1) an on-board, lightweight DT that enables the prioritisation and parsimonious transfer of data generated by the physical system; and (2) an off-board system for robust DT updating that enables the reliable detection of anomalous be haviours across the asset’s lifetime. The framework allows the DT to accurately replicate the behaviour of the system throughout its life, improve sensitivity to anomaly detection, and reduce the risk of forgetting previous system behaviours after updating the DT. An in-service gas turbine engine case study is used for demonstration.
How to Cite
##plugins.themes.bootstrap3.article.details##
Digital Twin, Anomaly detection, Model updating
Brinker, K. (2003). Incorporating diversity in active learning with support vector machines. In Proceedings of the 20th International Conference on Machine Learning (pp. 59–66).
Cai, W., Zhang, Y., & Zhou, J. (2013). Maximizing expected model change for active learning in regression. In Proceedings of ICDM (pp. 51–60).
Dasgupta, S. (2011). Two faces of active learning. Theoretical Computer Science, 412(19), 1767–1781.
Daumé III, H. (2009). Frustratingly easy domain adaptation. arXiv preprint arXiv:0907.1815.
The DTHub. (2026). The Digital Twin Hub. Retrieved from https://digitaltwinhub.co.uk/. Accessed: February 20, 2026.
Ducoffe, M., & Precioso, F. (2018). Adversarial active learning for deep networks: A margin-based approach. arXiv preprint arXiv:1802.09841.
Freytag, A., Rodner, E., & Denzler, J. (2014). Selecting influential examples: Active learning with expected model output changes. In European Conference on Computer Vision (pp. 562–577).
Gal, Y., Islam, R., & Ghahramani, Z. (2017). Deep Bayesian active learning with image data. In International Conference on Machine Learning (pp. 1183–1192).
Gemaque, R. N., Costa, A. F. J., Giusti, R., & Dos Santos, E. M. (2020). An overview of unsupervised drift detection methods. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 10(6), e1381.
Gissin, D., & Shalev-Shwartz, S. (2019). Discriminative active learning. arXiv preprint arXiv:1907.06347.
Glaessgen, E., & Stargel, D. (2012). The digital twin paradigm for future NASA and U.S. Air Force vehicles. In Proceedings of the AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference (p. 118).
Guo, T., Xu, Z., Yao, X., Chen, H., Aberer, K., & Funaya, K. (2016). Robust online time-series prediction with recurrent neural networks. In Proceedings of DSAA (pp. 816–825).
Habib, C., Makhoul, A., Darazi, R., & Salim, C. (2016). Self-adaptive data collection and fusion for health monitoring based on body sensor networks. IEEE Transactions on Industrial Informatics, 12(6), 2342–2352.
Harb, H., & Makhoul, A. (2017). Energy-efficient sensor data collection approach for industrial process monitoring. IEEE Transactions on Industrial Informatics, 14(2), 661–672.
Hartwell, A., Montana, F., Jacobs, W., Kadirkamanathan, V., Ameri, N., & Mills, A. (2024). Distributed digital twins for health monitoring: Resource constrained aero-engine fleet management. The Aeronautical Journal, 128(1325), 1556–1575.
He, T., Jin, X., Ding, G., Yi, L., & Yan, C. (2019). Towards better uncertainty sampling: Active learning with multiple views for deep convolutional neural network. In 2019 IEEE International Conference on Multimedia and Expo (pp. 1360–1365).
Kirsch, A., Van Amersfoort, J., & Gal, Y. (2019). BatchBALD: Efficient and diverse batch acquisition for deep Bayesian active learning. Advances in Neural Information Processing Systems, 32.
Kong, L., Zhang, D., He, Z., Xiang, Q., Wan, J., & Tao, M. (2016). Embracing big data with compressive sensing: A green approach in industrial wireless networks. IEEE Communications Magazine, 54(10), 53–59.
Li, C., Mahadevan, S., Ling, Y., Choze, S., & Wang, L. (2017). Dynamic Bayesian network for aircraft wing health monitoring digital twin. AIAA Journal, 55(3), 930–941.
Li, X., Grandvalet, Y., & Davoine, F. (2018). Explicit inductive bias for transfer learning with convolutional networks. arXiv preprint arXiv:1802.01483.
Lu, J., Liu, A., Dong, F., Gu, F., Gama, J., & Zhang, G. (2018). Learning under concept drift: A review. IEEE Transactions on Knowledge and Data Engineering, 31(12), 2346–2363.
Lyu, Y., Li, H., Jiang, Z. M., & Hassan, A. E. (2024). On the model update strategies for supervised learning in AIOps solutions. ACM Transactions on Software Engineering and Methodology, 33(7), 1–38.
Magargle, R., Johnson, L., Mandloi, P., Davoudabadi, P., Kesarkar, O., Krishnaswamy, S., ... Pitchaikani, A. (2017). A simulation-based digital twin for model-driven health monitoring and predictive maintenance of an automotive braking system. In Proceedings of the International Modelica Conference (pp. 35–46).
McCloskey, M., & Cohen, N. J. (1989). Catastrophic interference in connectionist networks: The sequential learning problem. In Psychology of Learning and Motivation (Vol. 24, pp. 109–165). Elsevier.
Pop, R., & Fulop, P. (2018). Deep ensemble Bayesian active learning: Addressing the mode collapse issue in Monte Carlo dropout via ensembles. arXiv preprint arXiv:1811.03897.
Qi, Q., & Tao, F. (2018). Digital twin and big data towards smart manufacturing and Industry 4.0: 360-degree comparison. IEEE Access, 6, 3585–3593.
Ranganathan, H., Venkateswara, H., Chakraborty, S., & Panchanathan, S. (2017). Deep active learning for image classification. In 2017 IEEE International Conference on Image Processing (pp. 3934–3938).
Ren, P., Xiao, Y., Chang, X., Huang, P.-Y., Li, Z., Gupta, B. B., ... Wang, X. (2021). A survey of deep active learning. ACM Computing Surveys, 54(9), 1–40.
Richard, R., & Belacel, N. (2024). A conditioned unsupervised regression framework attuned to the dynamic nature of data streams. In Proceedings of the Future Technologies Conference (pp. 1–23).
Saurav, S., Malhotra, P., TV, V., Gugulothu, N., Vig, L., Agarwal, P., & Shroff, G. (2018). Online anomaly detection with concept drift adaptation using recurrent neural networks. In Proceedings of the ACM India Joint International Conference on Data Science and Management of Data (pp. 78–87).
Tao, F., Zhang, H., Liu, A., & Nee, A. Y. (2018). Digital twin in industry: State of the art. IEEE Transactions on Industrial Informatics, 15(4), 2405–2415.
Tuegel, E. J., Ingraffea, A. R., Eason, T. G., & Spottswood, S. M. (2011). Reengineering aircraft structural life prediction using a digital twin. International Journal of Aerospace Engineering, 2011.
Van de Ven, G. M., Soures, N., & Kudithipudi, D. (2024). Continual learning and catastrophic forgetting. arXiv preprint arXiv:2403.05175.
Wang, Z., Du, B., Zhang, L., & Zhang, L. (2016). A batch-mode active learning framework by querying discriminative and representative samples for hyperspectral image classification. Neurocomputing, 179, 88–100.
Wu, D., Lin, C.-T., & Huang, J. (2019). Active learning for regression using greedy sampling. Information Sciences, 474, 90–105.
Xu, Q., & Zheng, R. (2017). When data acquisition meets data analytics: A distributed active learning framework for optimal budgeted mobile crowdsensing. In IEEE INFOCOM 2017 — IEEE Conference on Computer Communications (pp. 1–9).
Yang, K., & Shahabi, C. (2005). A PCA-based kernel for kernel PCA on multivariate time series. In Proceedings of ICDM.
Yin, C., Qian, B., Cao, S., Li, X., Wei, J., Zheng, Q., & Davidson, I. (2017). Deep similarity-based batch mode active learning with exploration-exploitation. In 2017 IEEE International Conference on Data Mining (pp. 575–584).
Yoo, D., & Kweon, I. S. (2019). Learning loss for active learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 93–102).
Yosinski, J., Clune, J., Bengio, Y., & Lipson, H. (2014). How transferable are features in deep neural networks? In Advances in Neural Information Processing Systems (pp. 3320–3328).
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
The Prognostic and Health Management Society advocates open-access to scientific data and uses a Creative Commons license for publishing and distributing any papers. A Creative Commons license does not relinquish the author’s copyright; rather it allows them to share some of their rights with any member of the public under certain conditions whilst enjoying full legal protection. By submitting an article to the International Conference of the Prognostics and Health Management Society, the authors agree to be bound by the associated terms and conditions including the following:
As the author, you retain the copyright to your Work. By submitting your Work, you are granting anybody the right to copy, distribute and transmit your Work and to adapt your Work with proper attribution under the terms of the Creative Commons Attribution 3.0 United States license. You assign rights to the Prognostics and Health Management Society to publish and disseminate your Work through electronic and print media if it is accepted for publication. A license note citing the Creative Commons Attribution 3.0 United States License as shown below needs to be placed in the footnote on the first page of the article.
First Author et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.