Implementing BIM-based digital twins for preventive maintenance in facility managementBaig, Ahtisham Ali (Avtor) Košir, Mitja (Mentor) Vetršek, Jure (Komentor) master thesisBuilding Information Modeling (BIM)COBie ReportModel Context Protocol (MCP)Industry Foundation Classes (IFC)natural language queriesAI-BIM Integrationbidirectional communicationconversational interfacesGraphical User Interface (GUI)pluginThis research investigates the implementation of AI for facility management (FM), enabling communication with IFC models. Additionally, it explores open-source BIM-based digital twin solutions for improving data handover from construction to facility management, with specific focus on Construction Operations Building Information Exchange (COBie) standardization and intelligent documentation linking. The difficulties in extracting and transferring info to FM Software present significant challenges in the construction industry. Currently, no such software exports all formats simultaneously: Excel (xlsx), CSV Files, JSON, XML, Enhanced IFC, RDF/TTL, SQL, HTML, YAML, Existing solutions typically extract only Excel formats, and such software is expensive. A five solutions open-source architecture was developed to address ACE specific challenges with IFC compatibility, and documentation handover. Results demonstrate significant improvements in construction-to-operations data transfer including complete equipment documentation coverage during facility handovers, dramatic time reduction for COBie generation. The system successfully processes complex building models and generates industry-standard facility management reports through multiple export formats. The research contributes practical implementation guidance for organizations considering open-source BIM-to-FM solutions, demonstrating that collaborative development can produce superior alternatives to expensive commercial software for facility management data handover challenges.[A. A. Baig]20252025-09-23 08:45:36Magistrsko delo/naloga173780sl