Gyeonggi Province Uses AI to Predict Forest Fire Paths Every 30 Minutes, Pinpointing Damage Down to Smallest Administrative Units

Gyeonggi Province Develops 'G-DAPS', Pilot Launch Set for April

AI Collaboration Drives Innovation in Forest Fire Disaster Response

Real-Time AI Model Predicts Forest Fire Movement Paths

System Supports Alert Issuance and Disaster Text Targeting

Prediction Model to Be Shared with 31 Cities and Counties, and Other Local Governments

A new civil defense alert prediction model has been developed that uses artificial intelligence to predict the movement path of a forest fire when it occurs, allowing this information to be utilized for disaster text alerts or warnings.

Demonstration of the wildfire prediction model. Provided by Gyeonggi Province.

The Emergency Planning Office of Gyeonggi Province announced on March 30 that its employees had voluntarily formed an AI collaboration team and independently developed the "Civil Defense Alert Prediction Model" (tentatively named G-DAPS: Gyeonggi-do Disaster Analysis & Prediction System) for rapid disaster response.

The developed prediction model is a system that predicts in real time the expected path, arrival time, and estimated alert issuance time from the initial outbreak of a forest fire until the alert is issued. Employees directly utilized generative AI tools such as Claude and Gemini to code the prediction and analysis system and to analyze case studies.

The prediction model was built using data from the Korea Meteorological Administration's short-term forecast, the Korea Forest Service's forest fire risk forecast, the Ministry of Land, Infrastructure and Transport's Digital Twin Land open API, and the audible area information for 589 civil defense alert facilities across the province.

The system visualizes weather conditions, forest fire risk forecasts and incident history, and the audible range of civil defense alert facilities on a web map. Based on this, the risk of a forest fire is analyzed in 30-minute intervals to identify affected municipalities down to the smallest administrative unit. Employees can use this information to issue alerts and send disaster text messages for specific areas.

The Emergency Planning Office of Gyeonggi Province is currently in discussions with the Gyeonggi Province Forest and Greenery Division and the Korea Forest Service about introducing the civil defense alert prediction model, with pilot operation expected to begin as early as April.

Cho Gwanggeun, Director of the Emergency Planning Office of Gyeonggi Province, stated, "Starting with forest fires, we will continue to improve the performance of the prediction model so that it can also be applied to major natural disasters such as floods and heavy snowfall, as well as responses to North Korean trash balloon provocations." He added, "We will work with the AI Bureau of Gyeonggi Province to share the system with all 31 cities and counties in the province and with other local governments to help build a robust disaster safety network."

This content was produced with the assistance of AI translation services.