Related Experiment Video
Updated: Jun 9, 2026
05:58
Plasma Lithography Surface Patterning for Creation of Cell Networks
Published on: June 14, 2011
Neural Network Enabled Real-Time Plasma Imaging for Inverse Designed Fabrication of Micro/Nano Structures with
Rui Han1,2, Jiaqun Li1,2, Jianfeng Yan1,2
1Department of Mechanical Engineering, Tsinghua University, Beijing, China.
Advanced Materials (Deerfield Beach, Fla.)
|June 8, 2026
Summary
This study introduces a neural network strategy for real-time monitoring of ultrafast laser micro/nano fabrication. It enables precise quality control by imaging through plasma plumes and designing structures with high accuracy.
Area of Science:
- Materials Science
- Optical Engineering
- Artificial Intelligence
Background:
- Real-time monitoring is crucial for high-precision micro/nano structure fabrication using ultrafast lasers.
- Laser-induced plasma plumes obscure sample surfaces, hindering direct imaging during fabrication.
Purpose of the Study:
- To develop a neural network-enabled strategy for real-time plasma imaging and inverse design fabrication.
- To overcome limitations in monitoring micro/nano fabrication processes obscured by plasma plumes.
Main Methods:
- Utilized a conditional generative adversarial network (cGAN) for real-time plasma imaging from intensity profiles.
- Employed a multilayer perceptron (MLP) model for inverse design of micro/nano fabrication parameters.
- Validated the imaging strategy through dual-pulse and sequential single-pulse processing experiments on various materials.
Main Results:
- The cGAN achieved high-fidelity imaging of feature structures with a latency of 1691 milliseconds.
- The MLP model accurately predicted fabrication outcomes, achieving a coefficient of determination (R²) of 0.97 for both forward prediction and inverse design.
- Fabricated structures closely matched the designed target values.
Conclusions:
- The proposed neural network strategy enables effective real-time process monitoring in ultrafast laser micro/nano fabrication.
- This advancement contributes to the development of intelligent, high-precision laser manufacturing systems.
