Abstract The substantial contribution of the building sector to global carbon emissions highlights the limitations of conventional static assessment approaches and motivates the exploration of real-time, operation-oriented carbon management strategies. This study proposes and evaluates an integrated Building Information Modeling (BIM)–Internet of Things (IoT)–Artificial Intelligence (AI) framework for dynamic carbon footprint monitoring and operational optimization in green buildings. The proposed framework focuses on system-level integration and closed-loop decision support, rather than on the development of novel AI algorithms. A four-layer architecture is designed to integrate BIM-based static building information with IoT-driven real-time operational data, enabling continuous carbon assessment and multi-objective operational optimization. The framework was deployed over a 12-month period in a 15,000 m 2 LEED Gold-certified office building. Under the specific climatic, operational, and occupancy conditions of the case study, a 26.5% reduction in operational carbon emissions was observed relative to a baseline of 10,000 tCO 2 e, while maintaining acceptable occupant comfort levels. The LSTM-based forecasting module achieved a 24-h prediction accuracy of approximately 92%, supporting short-term proactive operational adjustments. Continuous monitoring further indicated that approximately 18% of operational emissions occurred during unoccupied periods, a pattern that would not be identifiable through static assessment methods. A techno-economic evaluation suggests that the proposed framework is financially feasible within the examined context, with an estimated simple payback period of 1.7 years and an internal rate of return of approximately 42%.
Yu et al. (Mon,) studied this question.