A comprehensive review published in iOptics reveals how artificial intelligence is fundamentally transforming the design and implementation of optical metasurfaces, overcoming longstanding challenges that have hindered their practical application. Optical metasurfaces, with their ultra-thin and lightweight properties, are driving the miniaturization and planarization of optical systems, but their development from unit-cell design to system integration has faced significant obstacles. The review article demonstrates how AI provides solutions for metasurface technology to transition from unit optimization to system-level integration, marking a pivotal advancement in photonics research.
Led by Professor Xin Jin from Tsinghua University, the review outlines how AI addresses challenges at each design stage of metasurface development. At the unit-cell level, AI-driven surrogate modeling accelerates electromagnetic response prediction, while inverse design frameworks explore complex solution spaces that traditional methods cannot efficiently navigate. Robust design methods enhance stability against manufacturing variations, addressing a critical practical concern in metasurface fabrication. "For metasurface optimization, AI methods like graph neural networks model non-local interactions between densely packed meta-atom," explains Jin. "Multi-task learning resolves conflicting performance objectives, and reinforcement learning enables real-time dynamic control."
At the system level, AI provides a unified differentiable framework that integrates structural design, physical propagation models, and task-specific loss functions. This represents a fundamental shift in how metasurfaces are designed and implemented. "This end-to-end optimization directly links nanostructure design to final application goals, overcoming incompatibility between metasurface design and backend algorithms," adds Jin. "AI is shifting metasurface design from traditional, staged methods toward intelligent, collaborative, and system-level optimization." This approach enables more efficient development of advanced optical systems that were previously difficult or impossible to create using conventional design methodologies.
The implications of this AI-driven approach are substantial for multiple technology sectors. Application areas benefiting from AI-driven metasurfaces include compact imaging systems, augmented and virtual reality displays, advanced LiDAR systems, and computational imaging platforms. These technologies stand to gain significantly from the enhanced design capabilities that AI enables, potentially leading to smaller, more efficient, and more capable optical devices. The review also identifies future research directions, including developing AI methods integrated with electromagnetic theory, creating unified architectures for multi-scale design, and advancing adaptive photonic platforms. The original research is available at https://doi.org/10.1016/j.iopt.2025.100004.
This advancement matters because it addresses fundamental limitations in optical engineering that have constrained device miniaturization and performance. By enabling more efficient design of complex metasurfaces, AI accelerates the development of next-generation optical technologies with applications ranging from medical imaging to consumer electronics. The integration of AI with metasurface design represents more than just incremental improvement—it enables fundamentally new approaches to optical system creation that could transform multiple industries. The work received support from multiple funding sources, including the Shenzhen Science and Technology Program under Grant JCYJ20241202123921029, the Natural Science Foundation of China under Grant 62131011, and the Major Key Project of PCL under Grant PCL2023A10–3.
