Abstract
Unmanned aerial vehicle (UAV) swarms offer transformative potential for disaster response, enabling rapid situational awareness, search and rescue, and resource delivery. However, effective coordination in dynamic, uncertain environments remains a critical challenge. This article proposes a multi-agent coordination framework for autonomous UAV swarms in disaster response, integrating decentralized task allocation, collision avoidance, and adaptive communication. A simulation-based study evaluates the framework against baseline approaches under varying disaster scenarios (earthquake, flood, wildfire). Results demonstrate that the proposed framework reduces mission completion time by up to 32% and improves coverage efficiency by 28% compared to centralized and random coordination strategies. The framework also exhibits robustness to communication disruptions, maintaining 89% task completion under 30% link failure. These findings highlight the viability of decentralized swarm intelligence for real-time disaster operations and offer design guidelines for future deployments.