Unlike birds that navigate with an environment of significant speed and agility, drones often rely on external guidance or pre-map routes. But the pioneering development of the Department of Mechanical Engineering at the University of Hong Kong’s School of Engineering and researchers at the University of Hong Kong’s Department of Mechanical Engineering have made UAVs and Micro Aviation Vehicles (MAVS) follow the example of a flying capability that imitates birds more closely than ever before.
The team has developed a safely preserved high-speed aerial robot (SUPER), capable of flying at 20 meters per second, and avoiding thinner 2.5 mm (such as power cords or branches), using only onboard sensors and compute power. With a wheelbase of just 280mm and a take-off weight of 1.5kg, Super’s compact design demonstrates extraordinary agility, browsing dense forests at night and cleverly avoids thin lines.
Professor Zhang described the invention as a game changer in the field of drone technology, “the picture ‘robot bird’ quickly maneuvered in the forest, effortlessly avoiding branches and obstacles at high speeds. This is an important step in autonomous flight technology. Our system can drive the highway on the highway, making it the same range as before, and form a certain range. It can avoid obstacles in real time while competing towards its goal.”
The breakthrough lies in the complex integration of hardware and software. Super uses a lightweight 3D light detection and range (LIDAR) sensor to detect obstacles up to 70 meters with precise accuracy. This is paired with an advanced planning framework that produces two trajectories during flight: one optimizes speed by risking into unknown spaces, and the other by retaining priority security in known accessible areas.
By directly processing LiDAR data as a point cloud, the system greatly reduces computing time and allows quick decision making even at high speeds. The technology has been tested in various real-life situations, such as autonomous exploration of ancient sites and exhibits seamless navigation both indoor and outdoor environments.
“The ability to avoid thin obstacles and navigate narrow spaces opens up new possibilities for applications such as search and rescue. Every second is calculated. In a variety of lighting conditions, including nighttime, the super robustness makes it a reliable tool for all-weather operations,” said Yunfan Ren, lead author of the research paper.
The research team envisions a wide range of applications of this innovative technology, including autonomous delivery, power line inspection, forest monitoring, autonomous exploration and mapping. In search and rescue missions, super-technical calfs can quickly drive disaster areas (such as collapsed buildings or dense forests), day and night with current drones to locate survivors more effectively or assess hazards more effectively. In addition, in disaster relief programs, they can provide critical supplies to remote and inaccessible areas.
