The 'Sohwakhaeng Data Team,' composed of students Kim Do-won, Kim Dong-jin, Kim Su-hwan, Kim Yoon-ji, Song Si-eun, Song Yoo-dam, Lee Ye-jin, Choi Go-woon, and Han Do-hee from the Department of Environment and Energy's AI Remote Sensing Laboratory (supervised by Professor Yeom Jong-min), won the Minister of Climate, Energy and Environment's Award in the livestock category at the 2026 Weather Big Data Competition held on August 5, 2026.

Now in its 12th year, the competition was organized to increase the value of meteorological and climate data and to identify talent with data-driven problem-solving capabilities. Co-hosted by the Korea Meteorological Administration, Korea Electric Power Corporation, and the Livestock Products Quality Evaluation Service, the competition drew a total of 125 teams—80 teams in the disaster and safety category and 45 teams in the livestock category. Ten teams in each category advanced from the preliminary round to compete in the finals.

The Sohwakhaeng Data Team combined Hanwoo individual, pedigree, and farm information with meteorological and climate data to analyze the major factors affecting carcass grade and the relationships among the datasets. They used analytical models with different characteristics, including XGBoost, CatBoost, and FT-Transformer, to compare and synthesize results. They applied validation methods reflecting real-world conditions to prevent data leakage and to enhance the reliability of their analyses.

They were particularly highly evaluated for proposing that meteorological data may be used not as a direct determinant of carcass grade but to identify production management and business risks for livestock farms—such as mortality risk from heatwaves and cold snaps, electricity usage, and feed supply.

The research team stated, 'Through this study, we confirmed the potential of AI models that combine meteorological information and livestock data.' They added, 'We thank our professor for his generous guidance and advice during the research, and we will further develop the work by adding stage-specific rearing data to provide practical support for farm production management and marketing decision-making.'