Enhancing data interoperability and connectivity in digital twin applications for building operations and maintenance

As buildings contribute significantly to global energy consumption and operational costs throughout their lifecycle, digital twins have emerged as a transformative solution for enabling real-time monitoring, quick fault detection, predictive maintenance, and intelligent facility management. However, their implementation remains limited in practice due to fragmented data systems, poor semantic alignment, and unreliable connectivity. These limitations hinder seamless data exchange between systems and compromise the ability of digital twins to support timely and accurate decision-making in real-world operations. To address these challenges, this thesis develops and validates a semantic digital twin prototype that applies semantic modelling and real-time data integration techniques to improve interoperability and real-time connectivity in building operations and maintenance. The study adopts a design science research methodology, with fault detection in air handling units chosen as the application domain due to its operational complexity and high data demands. The research begins by developing structured process models using the IDEF0 method to map key workflows and information exchanges. These are followed by semantic data models based on the IFC 4.3 schema. The data models are visualised using Unified Modelling Language to describe system structure and dynamic behaviour. They inform the design of a multi-layered prototype that integrates data from building information models, building management systems and Internet of Things systems using lightweight messaging protocols, specifically MQTT. The prototype is implemented in a live university building and evaluated through technical testing, stakeholder feedback, and real-time system monitoring. The study makes several contributions to the field of digital twin development for building operations. It introduces a set of hierarchical IDEF0 process models that clarify operational and diagnostic workflows for fault detection. It also extends the IFC 4.3 schema through custom property sets to enable semantic representation of dynamic operational data. Additionally, it proposes a topic structuring method for MQTT that aligns with the IFC data hierarchy, facilitating seamless integration of real-time sensor streams with BIM-based models. The prototype demonstrates how semantic consistency and real-time synchronisation can be achieved using open data standards and lightweight communication protocols. However, the study also identifies key limitations. These include the manual IFC data mapping, which is time-consuming and susceptible to human error, and dependency on existing IT infrastructure for real-time connectivity. Future research should investigate automated methods for semantic annotation and scalable deployment across diverse building systems.

Grantor dc:publisher

Oxford Brookes University

Author dc:creator

• Tuhaise, Valerian Vanessa

Contributors dc:contributor

• Tah, Joseph H.M
• Abanda, Henry

Statement dc:rights

• All rights reserved

Language dc:language

en

DOI dc:identifier

https://doi.org/10.24384/m1ke-ek79

OAI identifier oai:identifier

tle:6ee75d2b-1b1d-4cd3-9c61-86eadda3f3b8:d6bd9758-527a-46cd-bfe2-c433766e8fca:1