Ontology-driven energy management in smart buildings: A comprehensive review of methodologies, tools, and challenges

• Comprehensive survey of ontology-driven methods in building energy management. • 81 studies evaluated across various building energy analysis applications. • Compares Brick, SAREF, SSN, and Open Energy Ontology in building data standardisation. • Identifies open challenges in digital twins, multi-agent AI, and data harmonisation. • Guides future research on occupant-aware analytics, fault detection, and DR strategies. While building energy systems (BES) increasingly rely on advanced automation, occupant-centric control, and digital twin technologies, a lack of robust semantic interoperability continues to hinder integrated, adaptive, and efficient energy management. Ontology-based approaches have emerged as a powerful solution, offering structured knowledge representation, consistent data exchange, and improved interpretability for AI-driven analytics. However, existing reviews tend to focus on subfields - such as occupant behaviour, IoT sensor data, or demand response - without providing a holistic synthesis of ontology-driven frameworks across diverse BES. In response, this study systematically reviewed 997 articles using a series of academic databases from 2011 to 2024, identifying 81 primary studies relevant to building energy analysis for comprehensive review. The results reveal that although ontologies facilitate standardised data flows and multi-agent coordination, significant gaps remain in domain coverage, alignment with emerging industry standards, and the integration of time-series data for predictive modelling. Furthermore, the interdependence of occupant comfort, multi-energy systems, and real-time decision-making has not been sufficiently addressed. This review compares commonly utilised ontological models (Brick, SAREF, SSN, and Open Energy Ontology) and proposes future directions emphasising interoperability with domain-specific data sources, bridging occupant modelling with data-driven optimisation, and leveraging digital twins for real-time operations. These findings aim to guide both academic and industry stakeholders in harnessing ontology-based methods for more intelligent, adaptive, and sustainable building environments.