Jul 2026· International Journal of Advanced Research in Science, Communication and Technology· 0 citations
Abstract
Natural disasters continue to create significant challenges for emergency response teams due to delays in incident assessment, inefficient allocation of rescue resources, and lack of coordinated decision making. Conventional disaster management systems mainly depend on manual operations, making it difficult to respond effectively when multiple emergencies occur simultaneously. This paper presents a Multi-Agent Disaster Management Simulator that automates the disaster response process using intelligent software agents, machine learning, graph-based routing, and generative artificial intelligence. The proposed system consists of four autonomous agents responsible for incident detection, resource allocation, rescue coordination, and performance evaluation. A Random Forest model predicts disaster severity using environmental and demographic features, while the A* search algorithm identifies the shortest rescue path between response teams and disaster locations. Google Gemini API is integrated to generate human-readable explanations for resource allocation decisions. A React-based dashboard provides live visualization of incidents, rescue routes, and performance metrics through WebSocket communication. Experimental results demonstrate that the proposed approach improves response efficiency, reduces manual intervention, and provides better transparency during emergency management. The simulator also supports offline execution through rule-based fallbacks, making it suitable for academic research, training, and disaster response simulations.
Natural and man-made disasters continue to extract a catastrophic toll on human lives, infrastructure, and economies across the world. Between 2000 and 2022, disaster events claimed over 1.9 million lives and caused economic losses exceeding USD 2.97 trillion, with climate change accelerating the frequency and intensity of hydro-meteorological events such as cyclones, floods, and droughts. In India, a country that ranks among the world's most disaster-prone nations, the lack of an integrated, technology-enabled disaster response coordination system has repeatedly resulted in delayed relief operations, duplicated resource deployment, and inadequate victim tracing. This paper proposes a comprehensive Disaster Response and Emergency Management System (DREMS) — an AI-driven, IoT-integrated, multi-agency coordination platform designed to support all four phases of disaster management: mitigation, preparedness, response, and recovery. The DREMS architecture comprises six functional modules: a multi-hazard early warning subsystem using satellite and ground sensor fusion, an AI-based damage and casualty assessment engine using satellite imagery analysis, a real-time resource allocation and dispatch optimizer, a decentralized mesh-network communication infrastructure for connectivity-denied disaster zones, a victim registration and family reunification portal, and a post-disaster recovery tracking dashboard. The system was evaluated through a simulated large-scale flood disaster scenario modelled on the 2015 Chennai floods, demonstrating a 34% reduction in resource dispatch response time, 91.4% accuracy in satellite-based damage classification, 87.6% victim registration coverage in simulation, and communication continuity in connectivity-denied zones with mesh network node density of 1 node per 0.8 km². The proposed DREMS framework offers a deployable, scalable model for integrated disaster management aligned with India's National Disaster Management Authority guidelines.
Keywords — Disaster Management, Emergency Response, Early Warning System, Satellite Imagery, AI Damage Assessment, Mesh Network, Resource Allocation, NDMA, IoT, Flood Response
Agesta Jenifer A Agesta Jenifer A, Saran S Saran S, Potrivel K Potrivel K et al.· International Scientific Jou...· 0 citations
The increasing frequency, intensity, and complexity of natural and human-induced disasters have exposed critical limitations in conventional emergency response systems, creating an urgent need for rapid, intelligent, and resilient disaster management solutions. Recent advances in autonomous technologies, including unmanned aerial vehicles (drones), autonomous ground vehicles, marine robots, and artificial intelligence (AI)-enabled decisionsupport systems, have transformed disaster operations by enhancing real-time situational awareness, search and rescue, infrastructure assessment, medical supply delivery, and humanitarian logistics. This narrative review critically synthesizes the current evidence on the evolution and applications of autonomous systems across the disaster management cycle, encompassing preparedness, emergency response, recovery, and humanitarian logistics. It examines recent technological advances while evaluating the technical, operational, regulatory, ethical, and economic barriers that continue to limit large scale implementation. Building on these insights, the review proposes the Adaptive Autonomous Disaster Response Ecosystem (AADRE) Framework, a novel conceptual model that integrates AI-driven decision support, heterogeneous autonomous platforms, human expertise, humanitarian logistics, and continuous learning into a unified disaster response ecosystem. The framework provides a scalable roadmap for improving coordination, interoperability, adaptive decision-making, and disaster resilience. The findings highlight the need for harmonized governance, interoperable digital infrastructure, multidisciplinary collaboration, and implementation-focused research to facilitate the safe and effective integration of autonomous technologies into future disaster management systems, with important implications for policymakers, emergency management agencies, humanitarian organizations, and researchers.
R. Ohaka, Ayomide Daniel Akinyemi, Ugochukwu Udonna Okonkwo· Journal of Computers and App...· 0 citations
The increasing frequency and intensity of natural and man-made disasters have highlighted the necessity for intelligent disaster management systems capable of providing rapid response and accurate situational awareness. Conventional disaster management approaches often rely on manual observations, fragmented communication infrastructures, and delayed reporting mechanisms, which can significantly reduce the effectiveness of emergency response operations. The emergence of the Internet of Things (IoT) has introduced new opportunities for real-time monitoring, data acquisition, predictive analytics, and automated decision-making. This paper presents an IoT-enabled disaster management system that integrates distributed sensors, wireless communication networks, cloud computing platforms, and machine learning techniques to improve disaster preparedness, detection, response, and recovery. The proposed framework continuously monitors environmental and structural parameters, analyzes collected information through intelligent algorithms, and generates early warnings for emergency authorities and affected communities. The system aims to minimize casualties, reduce property damage, and enhance coordination among disaster response agencies. Experimental evaluation demonstrates improved prediction accuracy, reduced response time, and enhanced operational efficiency when compared with conventional disaster management systems. The proposed solution offers a scalable, reliable, and cost-effective approach for building resilient smart cities and disaster-resistant communities.
Keywords— Internet of Things, Disaster Management, Smart Cities, Early Warning Systems, Machine Learning, Cloud Computing, Emergency Response.
Kasiraju Rajvardhan Kasiraju Rajvardhan, Islavath Meenakshi Islavath Meenakshi, A. M. A Mamatha· International Journal of Cre...· 0 citations
In recent years, long-term communication systems for emergencies using UAVs have been developing rapidly in particular situations, such as disasters in remote regions. Previous machine learning approaches exhibit several limitations, including limited communication range, data loss in transmission systems, limited bandwidth availability, and abrupt communication failures, which collectively hinder overall system performance. To address this limitation, propose a hybrid, optimization-based UAV-assisted communication framework that integrates Federated Learning with swarm intelligence algorithm. The disaster-aware UAV deployment uses Federated Learning for decision-making to identify critical communication zones. A hybrid algorithm combining federated reinforcement learning with a graph attention-based UAV communication framework for consistent, low-latency data communication. The UAV network's lifecycle securities constant communication, energy efficient resource allocation and load balancing. In experiment analysis, 74.2% reduction in end-to-end latency (248 ms to 62 ms), 53.8% reduction in energy consumption, and a packet delivery ratio of up to 94% under varying network densities. The proposed system delivers reliable, scalable, and intelligent communication for emergency response in remote and disaster-affected areas.
C.Alakesan, Anthony Johnson A, M. M et al.· 2026 4th International Confe...· 0 citations
This paper presents a modular decision support system that infers the primary location of the user or the reported incident and a situation-aware risk level from multi-turn Turkish disaster dialogues between a help-seeking user and an AI-supported emergency assistant. The assistant guides the user with follow-up questions about health status, number of affected people, structural damage, environmental hazards, and known nearby landmarks to complete missing information. The system manages the dialogue with a finite state machine, determines the location by linking user cues to a local GeoJSON point-of-interest database and by landmark verification, and produces explainable risk scores with Multi-Criteria Decision Analysis. The key novelty is treating landmarks as an evidence layer that verifies the current location hypothesis through proximity and clustering instead of directly replacing candidates based on a landmark signal. In a ten-scenario pilot evaluation, accuracy, end-to-end latency, and token usage are reported for three configurations.
Eren Varlıker, Yusuf Sinan Özmen, Selim Balcisoy· Signal Processing and Commun...· 0 citations
The assignment of a limited number of search and rescue (SAR) personnel to multiple, geographically dispersed disaster sites is a critical decision problem that directly determines the effectiveness of the initial response. Although this problem extends the classical assignment problem, the multidimensional nature of disaster operations cannot be adequately captured by single criterion distance minimization. In this study, the problem is modeled around a unified objective function (Φ) that integrates personnel competence, travel proximity, disaster demand, coverage ratio, and operational team cohesion. Under this common objective, Mixed Integer Linear Programming (LP/MILP) and four nature-inspired metaheuristics (Grey Wolf Optimizer, Genetic Algorithm, Particle Swarm Optimization, and Ant Colony Optimization) are evaluated within a fair comparison framework. The method's dynamic incremental data mechanism also allows for the addition of reinforcement personnel arriving after the initial assignment and newly reported crash areas, while previously applied assignments remain locked. The method is validated on an urban earthquake scenario for the Çukurova district of Adana province, inspired by the 2023 Kahramanmaraş earthquakes. Across four scenarios representing a gradual transition from initial response to full capacity containing three disaster types, a total of 600 runs are evaluated using descriptive statistics, non-parametric hypothesis tests (the Friedman test and the Nemenyi post-hoc test), and convergence and sensitivity analyses. The principal finding is that the incremental solution (Φ = 0.8471) yields a higher objective value than the static approach solving the same data in a single pass (Φ = 0.8255), showing that the locked field state preserves operational continuity without sacrificing solution quality.
Nurettin Havutçu, Mevlüt Ersoy· Advances in Artificial Intel...· 0 citations