Integrating Knowledge Graphs and Digital Twins for Heritage Building Conservation

This study presents a framework for integrating digital twins and knowledge graphs to enhance heritage building conservation, addressing challenges in environmental stress management, material degradation, and structural integrity while preserving historical authenticity. Using validated synthetic data, the framework enables real-time monitoring, predictive maintenance, and emergency response through a digital twin connected to a knowledge graph. Four scenarios were simulated to evaluate system performance: high humidity exceeding a 75% threshold triggered alerts for limestone maintenance; temperature fluctuations caused strain levels up to 0.009 units in load-bearing components at 35 °C, necessitating structural inspection; cumulative degradation monitoring projected re-plastering needs by month eight as the plaster degradation index approached 85%; and sudden impact events simulated emergency responses, with strain spikes over 0.004 units prompting real-time alerts within 2.5 s. Response times averaged 50 ms under normal conditions, peaking at 150 ms during high-frequency updates, showing robust Application Programming Interface (API) performance and data synchronization. SPARQL (SPARQL Protocol and RDF Query Language) queries within the knowledge graph facilitated proactive maintenance scheduling, reducing reactive interventions and supporting sustainable heritage conservation, especially suited to humid–temperate climates. This framework offers a novel, structured approach that bridges modern technology with heritage preservation needs, addressing both urgent conservation challenges and long-term sustainability to ensure the resilience of heritage buildings.