FOREST CHASE WITHOUT GPS: Drone Tracks Runners Through Dense Woods Autonomously

Imagine a drone chasing a runner through a thick forest where GPS is useless and there's no internet connection. Purdue University's IDEAS Lab says it has built exactly that. In a demonstration, the lab showed an autonomous drone with onboard LiDAR that follows moving targets, such as runners, through dense, GPS-denied woods. It builds a real-time 3D map of its surroundings and runs obstacle-avoidance AI entirely on its own chip, with no cloud support.

The drone can reportedly weave under tree canopies and predict where a target is heading when trees block its view for a moment. That matters for search and rescue, where every minute counts and rescuers can't always count on satellites or networks. The work also fits the lab's Department of Defense-funded research on air-ground robotics and navigation in complex terrain. The demo drew plenty of online attention, with viewers excited about rescue uses and edge AI, and some comparing it to camera-based systems like Tesla's. Palmer Luckey also joined the conversation with a joke about the drone's usefulness for exercise. The details come from the lab's demonstration, and independent performance data, such as speeds, ranges and success rates, hasn't been published yet.