Korea University of Technology Research Team Publishes 'AI Fine Dust Prediction System' in International Journal

Precise Response to Fine Dust Issues Becomes Possible

The research on an AI (artificial intelligence)-based fine dust prediction system developed by a research team at Korea University of Technology and Education (KOREATECH) has been published in a prestigious international academic journal.

According to KOREATECH on the 12th, Professor Donghyun Lee's research team at KOREATECH, in collaboration with Korea AI, developed an AI-based fine dust prediction system that was published in the internationally renowned SCIE journal, Journal of Cleaner Production.

The system developed by the research team can predict fine dust concentrations more precisely and reliably than existing fine dust prediction systems, enabling faster and more accurate responses to fine dust issues.

The AI model developed by the research team learns fine dust and meteorological big data using the Monte Carlo method, producing precise prediction results even under various conditions, thereby providing highly reliable forecasts.

While existing fine dust prediction systems often showed discrepancies from actual measurements, using the research team’s AI model, PM10 can be detected with 88.2% accuracy, reducing false alarms to 0.1%, and PM2.5 can be detected with 91.2% accuracy, lowering false alarms to 0.5%.

Professor Donghyun Lee, Korea Polytechnic University. [Photo by Korea Polytechnic University]

The research team’s AI-based fine dust prediction system will be officially published in the journal on the 1st of next month.

Professor Lee also developed AI-based algal bloom prediction technology in 2022 and published a paper in Technological Forecasting and Social Change (IF=12.9), the world’s top journal in urban planning. He holds AI-based prediction technologies in environmental AI and infectious disease modeling and is engaged in international research and activities. Domestically, he has served as an environmental health expert on the Environmental Economics Subcommittee of the Ministry of Environment’s Central Environmental Policy Committee.

Professor Lee stated, "The newly developed AI model reduces uncertainty and enables precise predictions," adding, "It is expected to play an important role in protecting public health and air quality policies, and we are considering export through technology commercialization."

Korea AI, selected as a promising startup technology company for developing AI-based systems in the environment and health sectors, is a school-affiliated company established within KOREATECH. It has been selected as an official supplier for the Ministry of Science and ICT’s AI Voucher program. Using proprietary AI technology, it supplies systems for fine dust, water quality, environment, and infectious disease modeling domestically and internationally, and is also commercializing technology services such as AI dashboards, MLOps, LLM, and AI chatbots.

This research was conducted with support from the Korea Research Foundation’s Excellent Young Researcher Program and the Korea Environment Institute’s Environmental Big Data Analysis and Service Development Project.