Development of FDD-ON: an Ontology for VAV HVAC System Fault Detection and Diagnostics

Analysis

Why This Paper Matters

Fault detection and diagnostics (FDD) in HVAC systems is critical for energy efficiency and operational reliability, yet its deployment is hindered by fragmented data and lack of interoperability. This paper tackles a fundamental challenge: the absence of a structured, shared semantic framework to bridge heterogeneous data sources, equipment types, and diagnostic outputs. By introducing FDD-ON, an ontology specifically for VAV HVAC systems, the authors provide a machine-interpretable foundation that can unify FDD knowledge, enabling more scalable and transparent solutions.

The significance extends beyond traditional FDD. As buildings increasingly adopt digital twins and AI-driven maintenance, the need for interoperable data models becomes paramount. FDD-ON directly addresses this by formalizing fault-symptom-impact relations, which are essential for reasoning and querying across different systems. This work is a step toward breaking down information silos that have long impeded the integration of FDD with broader building management and analytics platforms.

Technical Contributions

• Modular and extensible ontology design: FDD-ON is structured to allow extension to other HVAC system types or fault categories, promoting reuse and adaptability.
• Comprehensive libraries: The ontology includes libraries for faults, symptoms, and impacts, capturing a wide range of operational abnormalities and their consequences, which is crucial for practical deployment.
• Controlled vocabulary: By defining a well-structured vocabulary, FDD-ON ensures consistent terminology across different data sources and applications, enhancing data interpretability.
• Explicit causal relations: The contributing cause-fault-symptom-impact relations enable sophisticated querying and reasoning, supporting diagnostic processes and mapping of heterogeneous FDD outputs.
• Evaluation with public datasets: The ontology is validated using publicly available VAV HVAC datasets, demonstrating its applicability to real-world data and FDD application development.

Results

The abstract reports that FDD-ON was evaluated using publicly available VAV HVAC system datasets and demonstrated through FDD development applications. The results indicate that FDD-ON provides a foundational semantic framework for advancing scalable, transparent, and interoperable FDD solutions. However, no specific quantitative metrics (e.g., accuracy, precision, or performance improvements) are provided in the abstract. This lack of quantitative evaluation makes it difficult to assess the ontology's practical impact in numerical terms, but the qualitative demonstration suggests its utility in real-world scenarios.

Significance

FDD-ON addresses a critical gap in the HVAC FDD domain by providing a shared semantic model that can facilitate data integration and interoperability. This is particularly relevant as the industry moves toward digital twin-enabled FDD and AI-driven maintenance, where consistent data representation is essential. The ontology's modular design and comprehensive libraries make it a valuable resource for researchers and practitioners, potentially accelerating the development of interoperable FDD applications. By enabling machine-interpretable querying and mapping of diagnostic knowledge, FDD-ON could serve as a cornerstone for future building analytics platforms, contributing to more efficient and reliable building operations. Its focus on VAV systems, while a limitation, provides a solid foundation that can be extended to other HVAC configurations.