Rescue operations, particularly during natural disasters like floods, often require the rapid deployment of Unmanned Aerial Vehicles (UAVs) to assist in search, coordination, and logistics. However, effectively managing a group of UAVs in such dynamic environments presents several challenges, including decision-making under uncertainty, optimizing resource allocation, and ensuring seamless coordination among autonomous agents. This paper examines the complexities of multi-agent systems in UAV-based rescue operations and highlights the need for adaptive decision-making algorithms to address environmental changes and resource limitations. The key to this approach is balancing the UAVs’ coverage capabilities, energy consumption, and real-time operational adjustments using collaborative decision-making methods. The paper also explores the specific roles of different UAV types − reconnaissance, transport, communication, and monitoring − in achieving efficient mission execution. A multi-level coordination system is proposed, where UAVs autonomously distribute tasks, adapt to new information and ensure continuous communication. This framework aims to enhance the reliability and scalability of UAV group operations while maintaining flexibility in the face of rapidly changing conditions. By improving these systems, rescue operations can be carried out more effectively, even in remote or challenging environments, reducing the risks to rescuers and victims.