Chinese researchers have developed a new and much more efficient navigation system, allowing a swarm of 10 lightweight drones to fly together without colliding or hitting obstacles, including in places as congested and perilous as forests.
Much lighter and more agile devices
even though drones can calculate their position, determine the optimal route and identify potential obstacles using a whole range of sensors, the latter often prove to be expensive and cumbersome. Therefore, developing more compact versions of such devices generally involves getting rid of these types of components, which impacts their ability to evolve safely within their environment.
In the context of work published in the journal Science Robotics, Xin Zhou and his colleagues from theuniversité du Zhejiang have developed a new software approach to reduce the size and hardware requirements of the drone without sacrificing its agility.
Able to fit in the skin of a hand, their new drone incorporates commercially available electronic components (including a camera that transmits images in real time to its central processor), powered by a 100 gram battery giving it autonomy in 11 minute flight.
A high-performance algorithm
The localization algorithm developed by the team creates a 3D image where the position of obstacles (including other devices in the swarm) is regularly updated, allowing the flight pattern of each drone to be readjusted in real time. The most efficient route to pass through the area safely is then planned.
While such a program drains most of the drone’s processing power, it has the advantage of not requiring specialized (more power-hungry) processors or GPS signals, meaning these drones might theoretically be deployed anywhere.
« The proposed system is similar to birds able to fly freely through the forest while avoiding obstacles and other moving creatures “, Explain Zhoupointing out that such swarms might eventually carry objects (including within urban environments) and perform aerial mapping in hard-to-reach areas.