A drone-like aerial robot capable of navigating unknown environments has been developed by a team of engineers from the University of Hong Kong.

Specifically, SUPER includes a lightweight 3D lidar sensor for precise, long-range obstacle detection.
Enabling SUPER to achieve safe high-speed flight is a planning framework that generates trajectories directly from lidar point clouds. The team explained that each re-planning cycle generates two trajectories: one prioritizing safety within known free spaces and another optimizing speed across all spaces. This approach reportedly reduced failure rates by 35.9 times, cut planning time in half and allowed SUPER to navigate at speeds exceeding 20 meters per second while successfully avoiding thin obstacles as well as confined spaces.
“SUPER represents a milestone in autonomous MAV (micro air vehicle) systems, bridging the gap from laboratory research to real-world applications,” explained the researchers in the study abstract.
During trials of SUPER, the robot was capable of flying 20 meters per second across a forest-like obstacle course without incident, successfully avoiding trees and branches as it followed a target, like a human.
The team envisions that SUPER will be used in applications such as exploration, logistics, inspection and search-and-rescue operations.
The SUPER is detailed in the article, “Safety-assured high-speed navigation for MAVs,” which appears in the journal Science Robotics.