Journal of Engineering Research and Reports

Large healthcare campuses contain dissimilar clinical and non-clinical buildings, central thermal plants and increasingly connected building-control systems. Their energy demand is unusually difficult to optimise because continuity of care, infection prevention, thermal conditions and equipment reliability constrain operational flexibility. This critical narrative review evaluates whether artificial intelligence and intelligent building analytics offer credible improvements in energy efficiency at the campus scale, and under what technical and governance conditions. Peer-reviewed hospital-specific investigations, building-energy prediction and fault-detection studies, field evaluations of predictive control, and research on digital twins and clinical indoor environments were critically synthesised. Evidence available through 12 August 2026 was considered. The literature supports a clear distinction between accurate energy forecasting, detection of actionable faults, and measured energy savings attributable to closed-loop interventions. Hospital data studies demonstrate promising predictive performance, whereas direct, prolonged, independently evaluated deployments across entire healthcare campuses remain uncommon. Model predictive control has a stronger real-building implementation record than autonomous reinforcement learning, but aggregate benefits from ordinary commercial buildings cannot be assumed to apply to operating theatres, isolation rooms or critical services. In addition, normalisation by clinical activity, sensor calibration, counterfactual measurement, interoperability and cybersecurity materially influence the validity of claimed savings. This review proposes a safety-constrained, hierarchical approach in which transparent benchmarking and diagnostics precede supervised optimisation and carefully governed campus-level coordination. Future investigations should report whole-campus and zone-level outcomes, seasonal persistence, uncertainty, clinical environmental compliance, equipment reliability and carbon consequences under reproducible evaluation designs. Intelligent analytics should therefore be understood as a means of improving the quality of operational decisions rather than an independently established guarantee of hospital energy savings.