Autonomous drone tracks targets in the forest without GPS

Researchers from Purdue University have presented an autonomous drone capable of pursuing a running person in the forest without using satellite navigation.

The project was developed independently by the IDEAS Lab, including both hardware and software.

The drone uses LiDAR and inertial odometry to build a real-time 3D map of the terrain. The system predicts moments when trees might block the view of the target and autonomously navigates around obstacles. All calculations are performed onboard.

This technology eliminates the need for external computers or satellite signals. Forested areas no longer guarantee concealment from aerial surveillance of this level.

Currently, this is a research demonstration solution. Safety issues and operational efficiency in real-world conditions require further verification.