Intelligent phase change material envelopes for adaptive and predictive energy storage in smart buildings

Abstract Phase change materials (PCMs) can substantially reduce building energy demand through latent thermal energy storage, but conventional systems are constrained by passive operation and limited adaptability to varying climate, occupancy, and grid conditions. This field has been driven forward by recent advances in adaptive PCMs, predictive control, artificial intelligence, and digital twins, but the existing reviews cover these developments separately. This review summarizes 102 studies published from 2003 to 2026 and proposes a unified three-pillar framework of Hardware Intelligence, Control Intelligence and Data Intelligence for intelligent PCM building envelopes. It is demonstrated that passive PCM systems are inherently constrained by seasonal mismatch, and experimentally confirmed energy savings are on the order of 10–25%. However, with the integration of adaptive hardware and predictive control, reductions in operating cost of 20–57% and peak load shifting of up to 80% are possible. Emerging techniques such as reinforcement learning and differentiable predictive control provide improved adaptability but need more substantial experimental validation. The review concludes that the performance of intelligent PCM depends on the coordinated integration of material, control and data, rather than on isolated technological advances. Integrated material-control co-design, physics-informed artificial intelligence, digital twins and standardized evaluation frameworks are the critical priorities to accelerate grid-interactive, low-carbon building envelopes.