DIGITAL TWIN FOR BUILDINGS: A MODEL FOR PREDICTING BUILDING PERFORMANCE AND POST-CONSTRUCTION MANAGEMEN

The Architecture, Engineering, and Construction (AEC) sector is undergoing rapid digital transformation, with the Digital Twin (DT) emerging as a key paradigm for bridging design and operation.Unlike conventional BIM or standalone simulations, a DT provides a continuously updated virtual replica of a building, integrating IoT-based sensing, semantic interoperability, and AI-driven predictive models.This research develops and tests a comprehensive DT framework that combines BIM (IFC standard), Brick Schema and Project Haystack ontologies, IoT data streams, and hybrid predictive analytics (EnergyPlus + LSTM, XGBoost).A virtual medium-sized office building equipped with HVAC and lighting systems is used as a case study.Results show that the DT improves energy prediction accuracy to 5% MAPE, increases ASHRAE 55 thermal comfort compliance by 13%, and reduces unplanned HVAC failures by 35%.Annual energy use is lowered by 7.2%, equivalent to 18 tonnes of avoided CO emissions.Beyond these technical outcomes, the study highlights the importance of integrating cost-benefit considerations, occupant behaviour variability, and cybersecurity safeguards (aligned with NIST and IEC 62443) to ensure real-world applicability.The findings confirm the transformative potential of DTs for sustainable, resilient, and economically viable building management, while also providing a replicable framework for future large-scale adoption.