MULTI AI SENTINEL: MULTI-AGENT AI FOR PREDICTIVE FLOOD RESPONSE AND AUTONOMOUS RESOURCE ALLOCATION

Abstract

Floods are common natural disasters that affect people, infrastructure and the availability of essential resources, and identifying affected areas and providing the required support at the right time is challenging. To address this issue, this paper introduces Bharat AI Sentinel, a software-based system designed to support flood monitoring and emergency resource management. The system uses flood, weather, population and geographical data to identify areas affected by flooding and assess the level of impact. Based on this information, it estimates the number of people who may require assistance and the number of rescue teams needed, and it supports the allocation of essential resources, including food, water, medical kits and rescue teams, according to the estimated requirements. After allocation, the system records the updated quantities so that the remaining resources are available for subsequent allocation. Python is used for the artificial intelligence (AI) and machine learning (ML) processes, while PostgreSQL is used for data storage and management. In a case application to Kilakarai, the system classified flood severity as medium, estimated 500 of 1,000 residents as affected, allocated all required resources without shortage and identified an accessible evacuation route. The proposed system is intended to provide useful information to emergency response teams and help them plan rescue activities and resource distribution in a more organised and efficient manner.