Leveraging Time Series Foundation Models Embeddings for Remaining Useful Life Prediction
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
Recent advances in Remaining Useful Life (RUL) prediction rely heavily on task-specific deep learning architectures, such as CNNs, LSTMs, and Transformers. While effective, these data-intensive models frequently struggle to generalize across varying operating conditions. Time Series Foundation Models (TSFMs) offer a promising zero-shot alternative to this training-from-scratch approach, yet directly applying their forecasting-optimized representations to prognostics often fails to capture the physical constraints of equipment degradation. To resolve this task-objective mismatch, we propose a domain-agnostic adapter architecture that applies the Wide & Deep learning paradigm to repurpose the frozen Chronos-2 foundation model for continuous RUL regression. Our methodology explicitly bridges the domain gap by extracting and flattening the model’s abstract multivariate embeddings (Deep), and fusing them with raw, normalized physical measurements (Wide). Experiments on the full C-MAPSS benchmark demonstrate that this approach achieves state-of-the-art performance, reaching an average RMSE of 10.32. By outperforming recent specialized architectures without requiring backbone fine-tuning, this work proves that lightweight adaptation of generalized temporal representations offers a scalable, robust alternative to traditional prognostic modeling.
How to Cite
##plugins.themes.bootstrap3.article.details##
Remaining Useful Life, Time-Series Foundation Models, Deep Learning, Prognostics & Health Management
Ayyat, M., Osman, M., & Nadeem, T. (2025). Opportunities and challenges of foundation models in industrial manufacturing. IEEE Access.
Benzine, F. K., Beldjoudi, S., Benzine, A., Zouaghi, I., Haouari, F., & Seridi, H. (2025). From LSTM to Chronos: Evaluating deep learning models for predictive maintenance. In 2025 Intelligent Methods, Systems, and Applications (IMSA) (pp. 12–17).
Chaoub, A., et al. (2022). An end-to-end MLP-LSTM-MLP model for remaining useful life estimation. Prognostics and Health Management Conference.
Dintén, R., & Zorrilla, M. (2025). Using time-series foundation models for few-shot remaining useful life prediction of aircraft engines. Computer Modeling in Engineering & Sciences (CMES), 144(1).
Feng, Y., Hu, G., & Zhang, Z. (2024). GPT4Battery: An LLM-driven framework for adaptive state-of-health estimation of raw Li-ion batteries. arXiv preprint arXiv:2402.00068.
Garza, A., Challu, C., & Mergenthaler-Canseco, M. (2023). TimeGPT-1. arXiv preprint arXiv:2310.03589.
Goswami, M., Szafer, K., Choudhry, A., Cai, Y., Li, S., & Dubrawski, A. (2024). MOMENT: A family of open time-series foundation models. arXiv preprint arXiv:2402.03885.
Gu, A., Goel, K., & Ré, C. (2022). Efficiently modeling long sequences with structured state spaces. In International Conference on Learning Representations (ICLR).
Han, J., Kwon, Y., & Yoon, H. (2025). Mamba-attention: A self-supervised framework for efficient remaining useful life prediction. Reliability Engineering & System Safety, 111492.
Heimes, F. O. (2008). Recurrent neural networks for remaining useful life estimation. In 2008 International Conference on Prognostics and Health Management (pp. 1–6).
Kim, J., Kim, H., Kim, H., Lee, D., & Yoon, S. (2025). A comprehensive survey of deep learning for time-series forecasting: Architectural diversity and open challenges. Artificial Intelligence Review, 58(7), 1–95.
Kottapalli, S. R. K., Hubli, K., Chandrashekhara, S., Jain, G., Hubli, S., Botla, G., & Doddaiah, R. (2025). Foundation models for time series: A survey. arXiv preprint arXiv:2504.04011.
Li, H., Zhao, W., Zhang, Y., & Zio, E. (2020). Remaining useful life prediction using multi-scale deep convolutional neural network. Applied Soft Computing, 89, 106113.
Li, X., Ding, Q., & Sun, J.-Q. (2018). Remaining useful life estimation in prognostics using deep convolution neural networks. Reliability Engineering & System Safety, 172, 1–11.
Li, Y., Hu, C., Zhou, Z., Sun, C., Peng, J., & Yan, R. (2025). Learning globally ordered and locally consistent degradation representations for remaining useful life prediction. Advanced Engineering Informatics, 68, 103692.
Liu, J., et al. (2024). Large language models as general pattern machines for industrial prognostics. Reliability Engineering & System Safety.
Liu, L., Song, X., & Zhou, Z. (2022). Aircraft engine remaining useful life estimation via a double attention-based data-driven architecture. Reliability Engineering & System Safety, 221, 108330.
MacQueen, J. (1967). Some methods for classification and analysis of multivariate observations. In Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability (Vol. 1, pp. 281–297). University of California Press.
Miller, J. A., Aldosari, M., Saeed, F., Barna, N. H., Rana, S., Arpinar, I. B., & Liu, N. (2024). A survey of deep learning and foundation models for time-series forecasting. arXiv preprint arXiv:2401.13912.
Saxena, A., Goebel, K., Simon, D., & Eholzer, N. (2008). Damage propagation modeling for aircraft engine run-to-failure simulation. In International Conference on Prognostics and Health Management.
Shi, J., Zhong, J., Zhang, Y., Xiao, B., Xiao, L., & Zheng, Y. (2024). A dual attention LSTM lightweight model based on exponential smoothing for remaining useful life prediction. Reliability Engineering & System Safety, 243, 109821.
Wang, X., & Wang, Y. (2024). PULSE: Prompt-based latent space encoder for time series. arXiv preprint.
Wang, Y., Li, M., Liu, S., & Cong, H. (2026). Degradation manifold dynamic consistency network for probabilistic remaining useful life prediction. Nondestructive Testing and Evaluation, 1–27.
Wang, Y., Yu, Y., Sun, K., Lei, P., Zhang, Y., Zio, E., ... Li, Y. (2025). RMGPT: A foundation model with generative pre-trained transformer for fault diagnosis and prognosis in rotating machinery. IEEE Internet of Things Journal.
Ye, J., Yu, Y., Zhang, W., Wang, L., Li, J., & Tsung, F. (2025). Empowering time-series analysis with foundation models: A comprehensive survey. Retrieved from https://arxiv.org/abs/2405.02358
Yu, Y., Zhu, F., Wang, D., & Tsung, F. (2025). Empowering PHM applications with time-series foundation models: A unified multi-task learning approach. In 2025 IEEE 21st International Conference on Automation Science and Engineering (CASE) (pp. 3300–3305).
Zhang, Z., Song, W., Wu, Q., Sun, W., Li, Q., & Jia, L. (2025). A novel local enhanced channel self-attention based on transformer for industrial remaining useful life prediction. Engineering Applications of Artificial Intelligence, 141, 109815.
Zheng, S., Ristovski, K., Farahat, A., & Gupta, C. (2017). Long short-term memory network for remaining useful life estimation. In 2017 IEEE International Conference on Prognostics and Health Management (ICPHM) (pp. 88–95).
Zheng, S., Ruelens, K., Valkenborg, D., & Lemmens, B. (2017). Long short-term memory network for remaining useful life estimation. In 2017 IEEE International Conference on Prognostics and Health Management (ICPHM) (pp. 88–95).
Zhuang, F., Qi, Z., Duan, K., Xi, D., Zhu, Y., Zhu, H., ... He, Q. (2020). A comprehensive survey on transfer learning. Proceedings of the IEEE, 109(1), 43–76.
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.