A Study on an Edge AI-Based Disaster AX Platform Architecture for Climate-related Disaster Response in Island Regions

Abstract

Climate-related disasters in island regions are increasing due to climate change and geographical characteristics. Jeju Island is particularly vulnerable to marine, coastal, and mountainous disasters such as coastal accidents, low-salinity inflow, flash floods, and localized heavy rainfall. Existing disaster management systems mainly rely on post-event monitoring and manual decision-making processes, making it difficult to provide rapid responses and predictive actions. This paper proposes an Edge AI-based Disaster AX platform architecture for Island Regions. The proposed architecture integrates Edge AI CCTV systems, environmental sensors, disaster prediction models, and an AI-based integrated control platform to support real-time monitoring, prediction, and early response. The architecture is designed to process data at edge devices while utilizing cloud-based integrated analysis for large-scale decision support. The proposed approach aims to reduce disaster response time and improve operational efficiency in island environments.