Traditional drone navigation systems are based on pre-mapped environments or actual calculations in real time, limiting their ability to dynamically react. However, researchers from the University of Hong Kong have developed The high -speed aerial robot assured safety (Super) – A compact MAV with a wheelbase of 280 millimeters and a push / weight ratio greater than 5.0. This new generation autonomous drone can make fractional-secret decisions using perception based on Lidar, real-time mapping and a two-trajectories strategy.
Its advanced navigation system allows great navigating in high -speed complex environments while avoiding obstacles in real time. By integrating Lidar detection and an intelligent planning framework, the drone ensures both agility and security – even in completely unknown land.
The key innovation behind Super is its two -trajectories planning system, which allows it to explore new paths while retaining a safeguarding plan for safety:
- Exploratory trajectory – traces a quick and effective path towards its objective, even through unknown spaces.
- Safeguard trajectory – guarantees that the drone can always return to a safe space known if the exploratory path encounters obstacles.
This system recalculates the trajectories 10 times per second, allowing Super to instantly react to changes in its environment and avoid collisions – even at high speed. If a replan fails, Super will execute the last trajectory initiated – guarantee security by ensuring that the drone never enters unknown or dangerous areas without a way to return.
One of the biggest challenges in drone navigation in real time is to quickly and accurately understand the environment. The Super System tackles this with a sophisticated lidar perception module:
- The LIDAR sensor of the drone captures 200,000 points per second, detecting obstacles at more than 70 meters.
- A dynamic mapping system guarantees that only the relevant data is stored, preventing the size which could slow down the calculations.
- A sliding point strategy deletes obsolete information (for example, moving objects that are no longer present), preventing the detection of false obstacles.
This level of perception allows super to navigate effectively through forests, urban environments and disaster areas, where obstacles are unpredictable and change constantly.
What makes this navigation system really unique is its generation of real -time backup corridor – an approach first of security which guarantees that the drone always has an evacuation path without collision.
Using a new CIRI algorithm (Iterative regional inflation of the configuration space), the system extracts the safe flight corridors directly from Lidar data. Unlike the old methods based on pre-mapped environments or an expensive occupation cartography in calculation, CIRI allows a generation of safe areas in real time, considerably reducing the general treatment costs.
- Ciri guarantees that each backup corridor is fully contained in the known free space, preventing navigation failures.
- The system accumulates recent Lidar analyzes (1-2 seconds) to improve the precision of the corridor.
- Even with the limited Lidar field, the CIRI dynamically adjusts the backup area to guarantee safe reservation.
By eliminating the need for an occupation cartography by radius department, this breakthrough allows high -speed UAV operations in difficult and unpredictable environments.
The ability to generate safe and real -time backup trajectories marks a major jump forward in autonomous drone technology. This innovation is ideal for:
- Drings of autonomous drones in complex urban environments.
- Research and rescue missions in disaster areas.
- Military recognition in unknown or high -risk areas.
- Precision agriculture, where the UAV scanned large fields with a minimal delay.
By combining high -speed flight, real -time cartography and smart navigation, super pushes the limits of what autonomous drones can achieve. While these capacities continue to evolve, we can soon see drones navigate in complex environments with the same agility and the same consciousness as real pilots – if not better.
