Physics-Informed Machine Learning for Robust Industrial Diagnostics: A Systematic Investigation Using Heat Pump Systems
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
Industrial prognostics systems must operate reliably under real-world constraints: limited labeled data, shifting operational conditions, and deployment across heterogeneous units. While Industrial Internet of Things (IIoT) -enabled systems generate vast sensor data, purely data-driven approaches lack the robustness to exploit it effectively, as they tend to overfit to training distributions and fail when conditions change. Physics-Informed Machine Learning (PIML) offers a principled solution by grounding learned models in physical laws, making them more transferable and interpretable. This thesis investigates whether physical knowledge can provide a robust foundation for industrial diagnostics, using heat pump systems to study generalization across operating conditions and units.
How to Cite
##plugins.themes.bootstrap3.article.details##
Physics Informed Machine Learning, Domain Adaptation, OOD, Heat Pump Systems
Fink, O., Nejjar, I., Sharma, V., Niresi, K. F., Sun, H., & Dong, . . K. Y., H. (2026). From physics to machine learning and back: Part ii-learning with inductive biases in prognostics and health management. Reliability Engineering System Safety, 112376.
Fink, O., Sharma, V., Nejjar, I., Von Krannichfeldt, L., Garmaev, S., & Zhang, . . S. K., Z. (2026). From physics to machine learning and back: Part i-learning with inductive biases in prognostics and health management. Reliability Engineering System Safety, 112213.
Guo, Y., Wang, N., Shao, S., Huang, C., Zhang, Z., Li, X., & Wang, Y. (2024). A review on hybrid physics and data-driven modeling methods applied in air source heat pump systems for energy efficiency improvement. Renewable and Sustainable Energy Reviews, 204, 114804.
Karniadakis, G. E., Kevrekidis, I. G., Lu, L., Perdikaris, P., Wang, S., & Yang, L. (2021). Physics-informed machine
learning. Nature Reviews Physics, 3, 422-440.
Li, G., Hu, Y., Liu, J., Fang, X., & Kang, J. (2021). Review on fault detection and diagnosis feature engineering in building heating, ventilation, air conditioning and refrigeration systems. IEEE Access, 9, 2153-2187.
Niresi, K. F., Bissig, H., Baumann, H., & Fink, O. (2024). Physics-enhanced graph neural networks for soft sensing in industrial internet of things. IEEE Internet of Things Journal, 11(21), 34978-34990.
Taghiyarrenani, Z., Nowaczyk, S., Pashami, S., & Bouguelia, M. R. (2023). Multi-domain adaptation for regression under conditional distribution shift. Expert systems with applications, 224, 119907.
Wang, Z., Chen, J., Guo, K., Xu, B., & Chen, Z. (2025). Study on data augmentation with physics-informed generative adversarial networks and the extrapolation performance of cop prediction for chillers. Energy Conversion and Management, 346, 120418.
Wang, Z., Xia, P., Guo, J., Zhou, S., Wang, L., Wang, Y., & Zhang, C. (2025). Efficient feature selection for enhanced chiller fault diagnosis: A multi-source ranking information-driven ensemble approach. Building Simulation, 18, 141-159.
Zhao, C., Zio, E., & Shen, W. (2024). Domain generalization for cross-domain fault diagnosis: an application oriented perspective and a benchmark study. Reliability Engineering System Safety, 245, 109964. 3
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.