Smart drones for human detection in disaster response

Abstract

Natural disasters require swift, coordinated responses to minimise human casualties and infrastructure damage. This paper presents a novel AI-assisted drone system designed to enhance disaster relief efforts through advanced human detection and a distributed emergency Wi-Fi network. This drone system, equipped with state-of-the-art machine learning algorithms and thermal imaging, excels at locating and identifying individuals even in challenging conditions, such as smoke, debris, or low visibility. The drone fleet operates autonomously, dynamically forming an ad-hoc network that adapts to the evolving needs of the disaster zone. By integrating real-time data processing with efficient network management, our system provides a critical lifeline for communication and a powerful tool for rescuers to navigate and respond effectively. The detection and communication times are observed to be within the range of 5 to 8 seconds for this proposed system, which is widely used in disaster response.