Abstract
The integration of Extended Reality (XR) and Artificial Intelligence (AI) is transforming the way professionals interact with digital building models. This study explores how the integration of XR and AI can enhance interaction with Building Information Modeling (BIM) data in the Architecture, Engineering, and Construction (AEC) industry. We propose an AI-driven approach that enables users to query and visualize 3D building models, inspect specific building elements, and interact with floor plans through natural spoken language. By leveraging OpenAI’s GPT Large Language Model as a real-time AI assistant, the system facilitates interactions, significantly improving data retrieval, model comprehension, and decision-making in construction workflows. To ensure interoperability and seamless data exchange, the approach supports open data standards, specifically the Industry Foundation Classes (IFC) format. Based on this methodology, we developed a web-based application that runs directly in the browser, providing cross-platform accessibility beyond device-specific XR solutions. The implementation utilizes open-source web technologies ensuring flexibility and scalability. The application has been tested on the Meta Quest Pro, demonstrating its potential to bridge digital twin interaction and real-world construction management. Results indicate that this AI-assisted XR approach reduces complexity in BIM interaction, enhances collaboration, and addresses key industry challenges such as BIM data accessibility, model interpretation, and communication barriers.