Intelligent Rescue Rover for Human Detection using Transfer Learning and Embedded Systems*

The increasing frequency and severity of natural disasters highlight the urgent need for advanced search and rescue technologies to enhance emergency response efforts. This work focussed on developing an Autonomous Search and Rescue Rover (ASR-Rover), which is an AI-powered robotic system designed to autonomously navigate disaster-stricken and hard-to-reach areas to locate survivors. Equipped with state-of-the-art sensors, including LiDAR, thermal imaging, and an Arducam 64MP camera, the rover efficiently detects human presence and differentiates between environmental elements using deep learning algorithms running on a Raspberry Pi 5. Real-time data processing enables AI-driven decision-making, optimizing search and rescue operations by reducing response times and improving situational awareness. A GPS module records pre-cise survivor locations, while the HC12 S144B3 transmitter ensures long-range wireless communication with rescue teams for timely intervention. The rover's rocker-bogie mechanism provides stability on rough terrains, and ultrasonic sensors aid in autonomous navigation by detecting and avoiding obstacles. Powered by a lithium-ion battery, the ASR-Rover offers a robust, intelligent, and adaptable solution for search and rescue missions, significantly improving efficiency and effectiveness in disaster response scenarios.