EUVIMEDEUVIMED entdecken
Crossref

FedCOT: Personalized Federated Transfer Learning With Conditional Optimal Transport for Manufacturing Predictive Modeling

Anyi Li, Jia Liu

IEEE Transactions on Automation Science and Engineering · 2026 · 23 · 6363-6382

DOI: 10.1109/tase.2026.3673201 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Automation Science and Engineering, 2026. Verfasst von Anyi Li und 1 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 86 Literaturangaben. Gefördert von: National Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Effective predictive modeling in large-scale manufacturing is hampered by the isolated and limited data from individual organizations, collected from costly experiments and various inspections. Collaboration across organizations can handle these limitations, but it faces two main challenges: privacy concerns over organizations and heterogeneous features from varied sensing and inspection capabilities. Federated learning (FL) offers a solution by allowing organizations to collaboratively train a predictive model without sharing raw data, but standard FL struggles with the problem of feature heterogeneity. To address these challenges, we propose a personalized federated transfer learning framework with conditional optimal transport (FedCOT). FedCOT enables “target” organizations with limited features to benefit from “source” organizations with sufficient features in prediction performance while keeping data privacy through a central server. Each organization learns a personalized encoder-regressor structure by alternating optimization: the encoder maps heterogeneous inputs into a shared latent space, and the regressor predicts responses from the latent representations. Target organizations align their latent representations’ structure and corresponding responses with the source organizations’ information via COT. We evaluate FedCOT through simulations and a manufacturing case study on fatigue life prediction of additive-manufactured parts. Our case study demonstrates that FedCOT achieves the latent space where target organizations are highly aligned with source organizations, leading to a significant 34.99% improvement in prediction performance over the baseline method. Additionally, we provide theoretical guarantees on OT gradients and predictive consistency. Note to Practitioners—This work was motivated by the critical challenge of effective predictive modeling in large-scale manufacturing, where data is often limited, siloed, and hard to share due to costly experiments and privacy concerns. Existing approaches, such as federated learning, allow organizations to collaboratively build a shared predictive model without exchanging raw data. However, they usually assume that all organizations collect the same measurements, which is rarely the case in practice. Some organizations may adopt high-resolution inspection tools, while others use lower-cost methods, resulting in inconsistent data. In this work, we propose a personalized federated transfer learning framework that enables “target” organizations with limited measurement capability to benefit from the knowledge of “source” organizations with sufficient data, without exposing sensitive information. It builds a personalized predictive model for each organization while leveraging shared knowledge. Our experiments showed that our framework improved prediction performance by 34.99% compared with the baseline method. This framework is most practical when one organization has advanced sensing or inspection capabilities, and others do not, such as in additive manufacturing labs. It is generalizable to various applications, including healthcare, finance, and intelligent transportation. A limitation is that all organizations still need to measure the same response (e.g., part quality). Future work will address cases where responses differ across organizations or some lack responses entirely.

Förderung

  • National Science Foundation

Zitieren

APA

Li, A., & Liu, J. (2026). FedCOT: Personalized Federated Transfer Learning With Conditional Optimal Transport for Manufacturing Predictive Modeling. IEEE Transactions on Automation Science and Engineering, 23, 6363-6382. https://doi.org/10.1109/tase.2026.3673201

Vancouver

Li A, Liu J. FedCOT: Personalized Federated Transfer Learning With Conditional Optimal Transport for Manufacturing Predictive Modeling. IEEE Transactions on Automation Science and Engineering. 2026;23:6363-6382. doi:10.1109/tase.2026.3673201

BibTeX
@article{Li2026,
  author = {Li, Anyi and Liu, Jia},
  title = {FedCOT: Personalized Federated Transfer Learning With Conditional Optimal Transport for Manufacturing Predictive Modeling},
  journal = {IEEE Transactions on Automation Science and Engineering},
  year = {2026},
  volume = {23},
  pages = {6363--6382},
  doi = {10.1109/tase.2026.3673201}
}
Kostenlos starten →