Mechanics-informed AI laser 4D frameworks for adaptive materials processing and sustainable manufacturing: a quantitative review

The convergence of artificial intelligence (AI), laser-assisted materials processing, and four-dimensional (4D) smart materials is transforming the design and manufacture of adaptive structures across aerospace, biomedical, energy, and sustainable engineering domains. Stimuli-responsive materials such as shape-memory alloys, shape-memory polymers, hydrogels, and liquid-crystal elastomers enable programmable actuation and multifunctional behaviour; however, their performance remains highly sensitive to nonlinear deformation, thermal gradients, and microstructural evolution during fabrication. This review synthesises recent advances in AI-driven process modelling, inverse design methodologies, and laser-based fabrication techniques to address these challenges. A unified AI–Laser–4D framework is introduced, integrating surrogate learning, multi-objective optimisation, and closed-loop feedback within additive manufacturing and laser post-processing workflows. Evidence from recent studies indicates that AI-assisted parameter tuning can significantly reduce trial-and-error iterations while maintaining high predictive accuracy in surface integrity and actuation performance. Despite these advances, key challenges persist, including the lack of standardised datasets, limited uncertainty quantification, and difficulties in scaling multi-material architectures. To guide future research, a forward-looking roadmap is outlined, emphasising physics-informed AI models, open materials databases, and real-time adaptive control architectures aimed at enabling intelligent, sustainable, and reproducible 4D material manufacturing for next-generation industrial applications.