POSTECH LabCumentary Seungchul Lee (Mechanical Engineering)
Industrial AI Lab
Industrial AI Lab
Seungchul Lee (Mechanical Engineering)
The fusion of Artificial Intelligence (AI) into existing technology is bringing about an impactful transformation in our daily lives. Thanks to AI technology, rescue drones have advanced in ways which were previously not possible: they are now capable of distinguishing such intricacies as the voice of an individual lost in the mountain from any other background noise to filter out irrelevant sounds and pinpoint the exact location of the individual on a map. Video footage from capsule endoscopies of the small intestines are read by AI, not by doctors, and laryngeal cancer is diagnosed by AI, which requires just a voice audio sampling of the patient to determine.
The Industrial AI Lab, or iAI Lab, directed by Professor Seungchul Lee at the Department of Mechanical Engineering, POSTECH, aims to apply AI technology to a broad array of industrial systems and machines. At present, AI cannot easily be applied to machinery: machines are designed to achieve 100% accuracy in their functioning, and this inevitably leads to the lack of data required to identify abnormal conditions occurring in machines, making it difficult to operationalize AI on machines. Also challenging is the fact that sensor data gathered from machine systems are different in their linguistic structure from those processed by computers.
The Lab took a selection and focus approach that first applies AI to mechanical engineering to solve a series of issues. Researchers at the Lab developed technology to combine AI with manufacturing equipment and applied it to such instances as the detection of vibrations generated by equipment to identify defects and assess the integrity of such components as bearings and gears. The Lab has also made significant achievements in the materials sector – such as its development of technology that increases the image resolution of electron microscopy with the help of AI as a materials analysis tool. This enhanced the quality of images in the materials analysis process, and reduced the time and cost considerations taken for materials development. The findings of such achievements were published in the international academic journal of Acta Materialia in June of this year.
Through careful analysis of several possible industries that rely heavily on machine power – such as healthcare and robotics – the Lab is increasingly expanding the breadth and depth of AI applications. To acquire the high-quality data required of machine-based AI learning, the Lab has ensured it is equipped with experiment devices to collect data on vibrations, accelerations and movements as well as ultra-high speed cameras and thermal imaging cameras to gather video footage.
Recently, ‘AI+X’, which searches for industrial applications for AI, has not only garnered heightened attention, but it has also enjoyed significant increases in government support. This emerging field, however, demands the study of two completely heterogenous disciplines, and the rapid advancement of AI technology has proved it difficult for researchers to stay current on the latest developments. The iAI Lab takes pride in its potential to revolutionize mechanical engineering with its continuous exploration of applying AI+X to this field. The ultimate aim of the iAI Lab is to create AI-Aided Engineering (AIAE) as an emerging discipline in its own right, just as the invention of Computer-Aided Engineering has broadened the application of computing across a wide range of industries.
-
Head of Lab
-
Location
Science Building Ⅴ 223
Visit Other Labs