Scalable Knowledge Representation for Fault Diagnosis of Cyber Physical Systems: a Systematic Literature Review

Fault diagnosis in Cyber-Physical Systems (CPS) is essential for minimizing downtime, ensuring operational safety, and improving system resilience. As CPSs become increasingly interconnected and complex, traditional diagnostic methods struggle to capture their dynamic interactions and dependencies. Semantic technologies, including knowledge graphs and ontologies, offer a powerful solution by enabling structured representation, integration, and reasoning over diverse sources of diagnostic knowledge. This paper provides a comprehensive review of semantic approaches for fault diagnosis in CPS through a Systematic Literature Review (SLR). It covers key stages such as knowledge acquisition from domain experts, knowledge extraction from documents, semantic modeling of domain knowledge and data, and model enhancement. To the best of our knowledge, no prior systematic literature review has covered all these critical aspects. Unlike previous reviews, we systematically analyze and categorize the findings related to each stage. Additionally, we explore the role of available manufacturing data sources and their integration with semantic models. By bridging the gap between fault diagnosis and semantic technologies, this work highlights the potential of semantic representations to enhance interpretability, interoperability, and automation in CPS fault detection. We further discuss open challenges and outline future research directions, emphasizing the role of semantic frameworks in advancing intelligent fault diagnosis. Our findings aim to guide researchers and practitioners in leveraging semantic web technologies for more robust and explainable fault diagnosis in CPS.