과학기술정보통신부 산하 정부출연연구기관인 한국재료연구원(KIMS, 원장 이정환)이 인간과 인공지능과의 협업으로 재료의 미세조직 분석을 쉽고 빠르게 할 수 있는 방법을 개발해 다양한 분야에서 널리 활용될 것으로 기대되고 있다.
Materials Research Institute, Pohang University of Science and Technology, Development of a Framework for Classifying Microstructures
The Korea Institute of Materials Science (KIMS, President Lee Jeong-hwan), a government-funded research institute under the Ministry of Science and ICT, has developed a method to quickly and easily analyze the microstructure of materials through collaboration between humans and artificial intelligence, and it is expected to be widely used in various fields.
The Korea Institute of Materials Science announced on the 28th that the research team of Dr. Sejong Kim and Dr. Juwon Na from the Materials Digital Platform Research Headquarters Materials Data Center, together with the research team of Professor Seungchul Lee from Pohang University of Science and Technology, developed a technology to automatically distinguish and quantify the microstructure of materials from microscope images through interaction between humans and artificial intelligence (AI).
Microscopic imaging systems provide visualization of material structure information at various scales, from nano to meso scale. Quantitative analysis of microstructure is the process of extracting quantitative information by distinguishing microstructure phases from these microscopic images.
On the other hand, due to the complexity and diversity of microstructures, there have been many limitations for humans or artificial intelligence to perform this alone.

▲Microorganizational distinction based on human-AI collaboration
The research team developed an integrated quantitative microstructural analysis framework that can be universally utilized in the process of developing new materials by effectively integrating human and artificial intelligence capabilities.
The technology allows artificial intelligence to perform microstructural classification using just a single microstructural image and a rough expert sketch of it.
Additionally, AI actively interacts with humans by asking experts to sketch. Through extensive experiments, the research team confirmed that the human-AI collaboration framework can be applied universally across a variety of materials, microstructures, and microscopic systems.
While previous studies have required the collection of labor-intensive, large amounts of learning materials, this study is evaluated to have greatly improved practicality in that it replaced this with simple sketches using an expert's pen and mouse.
In the future, this technology will be embedded in the microstructure automatic quantitative analysis system (TIMs) that Materials Research Institute is developing so that it can be easily used by general researchers.
“This study is the result of improving the existing subjective and time-consuming quantitative microstructure analysis work into an objective and automated work,” said Na Ju-won, a senior researcher at Materials Research Institute.
In addition, Professor Seung-Cheol Lee of Pohang University of Science and Technology said, “The artificial intelligence framework that interacts with experts is expected to be widely used as a core analysis technology in industries and research fields,” adding, “Through this, we expect to drastically reduce the cost and time of new material research and development and even greatly improve reliability.”
This study was conducted with the support of the Ministry of Science and ICT through the Basic Project of the Korea Institute of Materials Science, the Mid-career Researcher Support Project of the National Research Foundation of Korea, and the Alchemist Project of the Ministry of Trade, Industry and Energy.
Additionally, the research results were published on August 15 in Acta Materialia (first author: Dr. Na Ju-won), the most authoritative journal in the field of metal materials.
.jpg)
▲(From left) Kim Se-jong, Senior Researcher at Materials Research Institute, NajuWon Senior Researcher, Professor Seung-Cheol Lee, Pohang University of Science and Technology