A Conceptual Framework for Advanced Predictive Maintenance and Fault Detection Solutions in Building Management

Predictive maintenance (PdM) and fault detection (FD) in buildings have been extensively studied, but their integration with Building Information Modeling (BIM) and Industry Foundation Classes (IFC), especially considering the Construction Operations Building Information Exchange (COBie) standards and Building Management Systems (BMS) remains a key challenge. This paper presents a conceptual framework outlining key factors that can guide the future development of AI-driven PdM and FD tools, leveraging BIM, IFC, and COBie, with data sourced from the BMS. The paper also examines the influence of AI technologies and how they have been implemented in PdM and FD. A conceptual framework consisting of seven blocks is proposed, including building systems, infrastructure technologies, input parameters, equipment sensors, models and algorithms, output parameters and visualization tools. Building systems and input parameters are also mapped to facilitate their integration with PdM and FD tools. Moreover, benefits and applications of PdM and FD in buildings have also been described. The conceptual framework is validated by building industry experts and practitioners to ensure its practical applicability across different building types and management systems. Finally, the paper discusses the potential of developing AI-driven PdM and FD tools to create intelligent and sustainable building environments.