HKU Develops Autonomous Drone That Navigates Forests at High Speed Without GPS or Light

Researchers from the University of Hong Kong, led by Professor Fu Zhang, have developed an advanced drone named SUPER (Safety-Assured High-Speed Aerial Robot). Unlike traditional drones that rely on GPS and pre-mapped routes, SUPER mimics birds’ natural agility and can autonomously navigate through complex environments like dense forests, even in darkness. Weighing 1.5 kg with a 280 mm wheelbase, SUPER achieves flight speeds of over 20 meters per second while avoiding extremely thin obstacles, such as twigs and power lines, using only its onboard sensors and processors.

The drone features a lightweight 3D LIDAR sensor with a 70-meter range and a dual-trajectory planning system—one route for speed in unknown areas, and another prioritizing safety. This technology enables fast, real-time decision-making without external data input. It has been tested in both indoor and outdoor scenarios, including archaeological site exploration and search-and-rescue simulations.

Lead author Yunfan Ren highlighted the drone’s potential for applications like emergency rescue, delivery, and environmental monitoring. Its ability to operate under challenging conditions, including nighttime, marks a significant advance in autonomous aerial technology.