A review of AIoT-enabled cyber-physical systems in building energy management: towards intelligent operation

Traditional building energy management (BEM) systems suffer from fragmented data, limited interoperability, and a lack of real-time optimization, hindering energy efficiency and carbon reduction efforts. While emerging technologies such as the Artificial Intelligence of Things (AIoT) and Cyber-Physical Systems (CPS) offer promising intelligent pathways to address these challenges, their integration into BEM remains limited and fragmented. This study conducts a systematic literature review following PRISMA guidelines, drawing on 109 peer-reviewed articles from Scopus and Web of Science, to synthesize the state of the art in AIoT-enabled CPS for BEM. The review categorizes advancements into four layers: perception, computation, system, and control. It details current practices, key technologies, interaction mechanisms, and limitations. The findings reveal an evolving trend from isolated device-level optimization toward integrated, closed-loop, and autonomous BEM. Despite notable progress, challenges persist in data interoperability, cybersecurity, scalability, and human-centered comfort modeling. To address these gaps, the study proposes future development pathways including self-powered and privacy-preserving sensing networks, federated edge intelligence, open and secure integration platforms, and multi-objective adaptive control frameworks. By establishing an end-to-end reference model from sensing to decision-making, this review provides a holistic understanding of AIoT-enabled CPS in BEM, offering theoretical and practical insights to accelerate the transition toward intelligent, low-carbon, and resilient built environments.