Federated Learning for Cyber Physical Systems: A Comprehensive Survey

TL;DR

This survey analyzes how federated learning can be effectively integrated with cyber-physical systems to address privacy, safety, and heterogeneity challenges while enabling real-time AI-enabled CPS applications. It introduces a comprehensive FL-CPS taxonomy and integration framework, surveys FL and CPS foundations, and catalogs domain-specific applications in healthcare, smart cities, and vehicular systems, plus cybersecurity use cases. Key contributions include a structured architecture- and data-centric taxonomy, an end-to-end FL-CPS integration workflow, and a detailed appraisal of challenges (security, resource management, standardization, heterogeneity) with proposed future directions. The work highlights the practical significance of privacy-preserving, distributed learning for CPS across critical sectors, guiding researchers and practitioners toward scalable, secure, and interoperable FL-CPS deployments. The implications span improved privacy, reduced data movement, and enhanced collaborative intelligence in safety-critical settings, with concrete considerations for communication, computation, and verification in real-world CPS ecosystems.

Abstract

The integration of machine learning (ML) in cyber physical systems (CPS) is a complex task due to the challenges that arise in terms of real-time decision making, safety, reliability, device heterogeneity, and data privacy. There are also open research questions that must be addressed in order to fully realize the potential of ML in CPS. Federated learning (FL), a distributed approach to ML, has become increasingly popular in recent years. It allows models to be trained using data from decentralized sources. This approach has been gaining popularity in the CPS field, as it integrates computer, communication, and physical processes. Therefore, the purpose of this work is to provide a comprehensive analysis of the most recent developments of FL-CPS, including the numerous application areas, system topologies, and algorithms developed in recent years. The paper starts by discussing recent advances in both FL and CPS, followed by their integration. Then, the paper compares the application of FL in CPS with its applications in the internet of things (IoT) in further depth to show their connections and distinctions. Furthermore, the article scrutinizes how FL is utilized in critical CPS applications, e.g., intelligent transportation systems, cybersecurity services, smart cities, and smart healthcare solutions. The study also includes critical insights and lessons learned from various FL-CPS implementations. The paper's concluding section delves into significant concerns and suggests avenues for further research in this fast-paced and dynamic era.

Figures (9)

• Figure 1: The structure of the survey.
• Figure 2: The overall architecture of a typical CPS.
• Figure 3: The communication process of FL in CPS.
• Figure 4: The overall architecture of a typical FL model in CPS.
• Figure 8: FL-CPS taxonomy framework structure.