GeoRescue Drone
Su. Suganthi, J. Ilakkiya, C Risheeth, V D Madhumitha, V Subbiah · 2025
Earthquakes, landslides, and floods are natural disaster induced catastrophes resulting in victim detection delay and greater casualties. The traditional rescue processes like manual digging,thermal imaging, and sonar scanning fail tolocate buried survivors under the rubble. In response to this, an Artificial Intelligence-based Unmanned Aerial Vehicle (UAV) with Ground Penetrating Radar (GPR), GeoRescue Drone, is introduced to scan subsurface in real-time to identify buried victims accurately and minimize rescue time. The drone emits high-frequency electromagnetic waves, sweeps reflected waves, and distinguishes human bodies from other objects with a machine learning-based classifier. GPS navigation and AI-driven autonomous flight offer high coverage with low human intervention.Performance testing shows 92.8% detection rate, 90.5% precision, and 1.2 seconds response time per sweep. Future plans involve multi-sensor fusion, onboard LiDAR, and optimized AI-driven signal processing for enhanced accuracy.