SETU-AI: A GIS-Based Decision-Support Framework for Proactive Disaster Relocation Planning

Communities living in disaster-prone regions often face repeated risks from hazards such as floods and landslides. Although existing disaster management systems provide valuable information for hazard monitoring and emergency response, relocation planning can involve additional questions, including identifying vulnerable communities, evaluating suitable alternative locations, and determining which communities require priority. This paper presents SETU AI, a proposed GIS-enabled decision-support framework designed to connect these stages within a single workflow. The framework integrates multiple factors, including hazard exposure, geographic conditions, population vulnerability and historical disaster information, to support Red Zone identification. It further demonstrates a relocation-site evaluation process based on factors such as safety, capacity, infrastructure, livelihood, accessibility, and cost. The proposed prototype also includes community prioritization and future scenario comparison to support proactive planning. Assam was selected as the demonstration context because of its recurring flood events, vulnerable settlements, and complex geographical conditions. The framework was implemented as an interactive web-based prototype using demonstration data to visualize the proposed workflow and provide transparent, factor-level explanations of its outputs. The current system has not been operationally deployed or validated using real-world datasets; therefore, its outputs should be considered illustrative rather than predictive. Future development would require verified datasets, real-time information, model validation, and field testing. Overall, SETU AI demonstrates the potential of integrating disaster-risk assessment and relocation planning within a transparent decision-support framework.