Professor Hyun Jong Yang of POSTECH's Department of Electrical Engineering has developed the world's first AI-controlled electron microscope with UNIST, Egovid Co., Ltd., and COXEM Co., Ltd.

Electron microscopes offer much higher magnification and more varied information than optical microscopes. However, obtaining good images is difficult for beginners because it requires knowledge of electron beams, specimen properties, vacuums, and other topics. Industrial users, particularly in production and quality control, therefore face constraints such as lengthy specialized training or hiring experts.

To address these difficulties, Yang's team was the first in the world to propose and successfully develop technology that uses AI to learn expert microscope operation, enabling beginners to obtain expert-quality images. Although academic applications of AI to electron microscopy have been reported several times and are increasing, Yang's team is the first to successfully apply and develop reinforcement-learning-based electron microscope control.

The technology is expected to make electron microscopes much easier to use and broaden their adoption in industry and education. COXEM is currently conducting further development with the goal of becoming the first company to commercialize the jointly developed AI-controlled electron microscope.

The team's findings were published in successive years, 2019 and 2020, in Microscopy and Microanalysis, a leading microscopy journal, and presented at the 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems, one of the world's largest robotics conferences.

Yang joined POSTECH's Department of Electrical Engineering on July 15, 2020, and is actively conducting research through the Advanced Information Systems Laboratory.