Optical metasurfaces, planar artificial media capable of controlling light propagation, are transitioning from laboratory curiosity to commercial applications. This shift requires advanced meta-atom and metasurface designs, considering manufacturability and enhancing optical performance with post-processing algorithms. Artificial-Intelligence(AI), particularly machine-learning(ML) and optimization, offers solutions to these demands. This perspective systematically reviews AI’s potential impact in three critical areas: AI-enabled metasurface design-for-manufacturing(DFM), design beyond the classical local phase approximation, and AI-empowered computational backend.

Metasurfaces, which consist of arrays of ultrathin planar nanostructures (also known as “meta‐atoms”), offer immense potential for use in high‐performance optical devices through the precise manipulation of electromagnetic waves with subwavelength spatial resolution. However, designing meta‐atom structures that simultaneously meet multiple functional requirements (e.g., for multiband or multiangle operation) is an arduous task that poses a significant design burden. Therefore, it is essential to establish a robust method for producing intricate meta‐atom structures as functional devices. To address this issue, we developed a rapid construction method for a multifunctional and fabrication‐friendly meta‐atom library using deep neural networks coupled with a meta‐atom selector that accounts for realistic fabrication constraints. To validate the proposed method, we successfully applied the approach to experimentally demonstrate a dual‐band metasurface collimator based on complex free‐form meta‐atoms. Our results qualify the proposed method as an efficient and reliable solution for designing complex meta‐atom structures in high‐performance optical device implementations.

From the whole plants of Clinopodium gracile five new saponins, called clinoposaponins I-V, were isolated together with two known and four artifact saponins, and their structures elucidated from spectroscopic data and chemical evidence. These saponins were composed of saikosaponin a and additional glucose moieties.

Additive manufacturing (AM) is employed for fabricating industrial products with complex geometries. As topological optimization is suitable for designing complex geometries, studies have combined AM and topological optimization, evaluating the density optimization of lattice structures as a variant of topological optimization. The lattice structures of components fabricated via AM comprise voids. Models designed using topological optimization should be modified to ensure structures suitable for AM. As the lattice unit can be easily fabricated using AM with fewer design modifications, this study uses lattice density optimization for an industrial AM product. We propose a method of optimizing the lattice distribution for controlling the surface temperature uniformity of industrial products, such as molds. The effective thermal conductivity of the lattice is calculated using the homogenization and finite element methods. The effective thermal conductivity changes depending on the internal pore sizes. The proposed methodology is validated using a 3D example; the minimization problem of surface temperature variations in the target domain is considered. The variable density of the embedded lattice in the target domain is optimized, and we experimentally validated the performance of the lattice unit cell and optimal 3D structure using metal powder bed fusion AM.

A catalyst-assisted HF gas etching process enables high-aspect-ratio SiO ₂ meta-atoms with smooth, vertical profiles. A constant-gap design enables scalable, high-efficiency all-glass metalenses for visible wavelengths.

We developed a deep-learning-enabled, fabrication-friendly metasurface design framework and experimentally demonstrated a dual-band metasurface collimator operating at 650 nm and 780 nm wavelengths concurrently.