New Ontology for HVAC Fault Detection and Diagnostics

This paper introduces FDD-ON, a new ontology specifically developed for Fault Detection and Diagnosis (FDD) in Variable Air Volume (VAV) HVAC systems. The core problem FDD-ON addresses is the fragmented nature of data and knowledge within the FDD domain, which hinders the effective deployment of FDD solutions, digital twins, and AI-driven maintenance systems. By providing a structured, machine-interpretable representation of HVAC components, fault types, symptoms, and their impacts, FDD-ON aims to improve data interpretability and interoperability. FDD-ON is designed to be modular and extensible, integrating comprehensive libraries for faults, symptoms, and impacts, along with explicit cause-fault-symptom-impact relationships. This semantic framework allows for querying diagnostic knowledge, mapping diverse FDD outputs, and developing scalable, transparent, and interoperable FDD applications. The ontology's effectiveness was evaluated using publicly available VAV HVAC system datasets, demonstrating its potential as a foundational tool for advanced FDD solutions.