Research published in Science Robotics presents a key advance in autonomous drone navigation. It is a system called SUPER (Safety-assured High-speed Aerial Robot) , developed by researchers at the University of Hong Kong in collaboration with aerial robotics experts, which allows MAVs (Micro Air Vehicles) to reach speeds of over 20 m/s in unknown environments without compromising safety.
SUPER is able to navigate through unstructured and unfamiliar forests. Source: MARS LAB HKU
The development of the SUPER autonomous drone
SUPER’s development responds to the need for drones capable of moving at high speed in search and rescue missions or urban environments without supporting infrastructure. To do so, researchers have integrated a long-range 3D LIDAR sensor , which scans the environment with high precision and feeds a dual trajectory planning algorithm. SUPER’s technology includes:
- Compact design and superior maneuverability : Its thrust-to-weight ratio of more than 5.0 allows maneuvers in tight spaces.
- Advanced sensing with 3D LIDAR : Detects objects up to 70 meters away with high precision.
- Dual trajectory algorithm : Simultaneously calculates two routes: a safe one in known space and a fast one in partially explored areas.
- Reduction in failure rates : The system reduces collisions by 35.9 times compared to previous methods.
Testing in real environments
Tests showed that SUPER can avoid thin obstacles such as power lines , outperforming commercial drones that often fail to detect such objects due to limitations in optical sensors, and navigate accurately in dense forests or dark spaces. Its combination of optimized hardware and intelligent software positions it as an ideal candidate for operations in hazardous or GPS-deprived areas.
This development represents a significant advance in drone autonomy, with potential applications in rescue missions in disaster zones, exploration in inaccessible environments and surveillance in areas without communication infrastructure and opens the door to new applications in rescue, surveillance and exploration of challenging environments.
Follow us on social media and don’t miss any of our posts!
YouTube LinkedIn Facebook Instagram X ( Twitter ) TikTok
Source: Science Robotics
Photo: University of Hong Kong