Integrating AI and Multimodal Sensor Fusion with Swarm Robotics for Survivor Detection in Collapsed Structures

Multiple studies have been conducted in the fields of both swarm robotics and machine learning, with ongoing research on integrating these two areas, particularly in disaster management. However, current research primarily focuses on navigating and mapping volatile terrain rather than detecting survivors in disaster areas such as those affected by earthquakes. This paper provides a holistic overview of sensors that can be equipped on robots for detecting survivors in collapsed structures, offers insight on leveraging artificial intelligence to control swarms of robots, and presents a framework for multimodal fusion of sensor data to enhance survivor detection probability. A detailed proposal combining relevant and evaluated strategies to maximise survivor detection using various sensors and algorithms has been included for further advancement.