EPFL researchers created a claw-like device that enables a flapping-wing robot to settle independently on a horizontal platform. The advancement has the potential to dramatically broaden the spectrum of robot-assisted jobs. A bird landing on a limb makes the action seem simple, yet the process of perching requires a highly precise balance of timing, high-impact forces, speed, and accuracy. It’s a manoeuvre so difficult that no flapping-wing robot (ornithopter) has ever mastered it.
Raphael Zufferey, a postdoctoral scholar at the School of Engineering’s Laboratory of Intelligent Systems (LIS) and Biorobotics ab (BioRob), is the first author on a new Nature Communications study revealing the unique landing gear that allows for such perching. He constructed and tested it alongside colleagues from the University of Seville in Spain, where the 700-gram ornithopter was created as part of the European project GRIFFIN.
“This is the beginning of a greater effort. Once an ornithopter has mastered landing independently on a tree limb, it will be able to do certain tasks, such as collecting biological samples or measurements from a tree invisibly. It may eventually settle on man-made buildings, opening up new avenues of use “According to Zufferey.
He adds that the ability to land on a perch might give a more efficient approach for ornithopters, which have limited battery life like many unmanned aerial vehicles (UAVs), to recharge using solar energy, possibly making them perfect for long-range missions.
“This is a significant step toward the use of flapping-wing robots, which can currently only fly in free flight, for manipulation tasks and other real-world applications,” he adds.
Increasing strength and accuracy while decreasing weight and speed
The technical challenges of landing an ornithopter on a perch without any external orders necessitated handling several elements that nature had previously flawlessly managed. The ornithopter has to be able to slow down considerably when perched while still flying. The claw had to be powerful enough to grab the perch and sustain the robot’s weight, but not so heavy that it couldn’t be kept aloft. “That’s one of the reasons we decided with a single claw instead of two,” Zufferey says. Finally, the robot must be able to recognise its surroundings as well as the perch in front of it in connection to its own location, speed, and trajectory.
All of this was accomplished by outfitting the ornithopter with a completely on-board computer and navigation system, which was supplemented by an external motion-capture device to assist it in determining its location. The leg-claw appendage of the ornithopter was highly tuned to compensate for the up-and-down oscillations of flight as it sought to focus in on and hold the perch. The claw was intended to absorb the robot’s forward velocity and shut swiftly and securely to maintain its weight. Once perched, the robot stays there without using any energy.
Despite all of these variables, Zufferey and his colleagues were successful, eventually developing not one, but two claw-footed ornithopters to reproduce their perching findings.
Zufferey is already considering how their gadget may be extended and enhanced, particularly in an outdoor context.
“The flying studies are currently being conducted inside since we need a regulated flight zone with exact localization from the motion capture system. We want to strengthen the robot’s autonomy in the future so that it can execute perching and manipulation tasks in a more uncertain environment.”
