Kookmin University Students Win Grand Prize for AI Forest Management Tool

A team of students from Kookmin University's Department of Forest Environment Systems has won the top AI Service Creation Award (Grand Prize) at the first Forest Science AI Utilization Competition. The competition was hosted by the National Forest Science Institute of Korea.

The winning team, named 'Forest Citizens,' comprised Master's student Choi Hee-do as team leader, along with Na Jeong-woo, Ha Su-beom, and Kim Min-seok from the Department of Forest Environment Systems. Guided by professors Im Cheol-hee and Kang Yu-jin, they developed 'MOFOM: Multi-purpose Forest Management Decision Support AI.'

MOFOM is a web service designed to assist forest owners by comparing the value of different management strategies for their specific forest plots. Users input their forest plot number, and the AI estimates timber volume and carbon content using satellite imagery. It then predicts forest growth over the next 30 years, incorporating data from the National Forest Inventory, carbon sequestration standards, and mountain weather data.

“

It took a lot of time to connect the forest data scattered across various institutions into one service, but it was meaningful to have it evaluated after implementing it into a usable form.

Based on these predictions, MOFOM presents comparative values for six different forest management scenarios, ranging from immediate logging to thinning and subsequent nurturing, alongside estimated carbon sequestration amounts for each. The team validated their results using actual forest plots in Boeun County, Chungcheongbuk-do, and data from carbon offset projects, ensuring accuracy.

To minimize the risk of numerical errors, the service uses a language model to explain the results only after calculations are complete. The team's ability to provide a fully functional service for direct demonstration by judges was recognized for its technical completeness and practical applicability. The students leveraged their academic knowledge in forest science and management, including satellite image analysis, growth prediction, economic calculations, and web service development, to create MOFOM.

Team leader Choi Hee-do expressed that connecting dispersed forest data into a single service was challenging but rewarding. He hopes to expand the model, validated in Boeun, nationwide so that forest owners across the country can utilize it. The competition, held for the first time this year, aimed to solve forest-related issues and discover new forest policies and services using big data and AI, attracting 259 participants from 41 universities and 33 institutions.

“

I was able to apply what I learned in the department, such as forest growth and management, to solve real-world problems, and I want to develop it into a service that forest owners can actually use by expanding the model validated in Boeun, Chungcheongbuk-do, nationwide.