A bird landing on a branch is no big deal; we see it happen hundreds, if not more, of times per day. But a flapping-wing robot (ornithopter) performing the same feat? Unheard of — until now.
In what could drastically increase robot-assisted tasks, Dr. Raphael Zufferey and a team of researchers have developed and tested at the University of Seville (Spain) a 700-gram (1.5-pound) ornithopter.
“This is the first phase of a larger project,” Zufferey said. “Once an ornithopter can master landing autonomously on a tree branch, then it has the potential to carry out specific tasks, such as unobtrusively collecting biological samples or measurements from a tree. Eventually, it could even land on artificial structures, which could open up further areas of application.”
In addition, the perching could provide a way for ornithopters to recharge using solar energy — thus making them perfect for long-range missions.
“This is a big step toward using flapping-wing robots, which as of now can really only do free flights, for manipulation tasks and other real-world applications,” he added.
The engineering issues that come with perching an ornithopter sans external commands required mitigating many variables: It had to be able to slow down prior to perching while still maintaining flight; its claw needed to be strong enough, yet not too heavy, so as to grasp the perch and support the robot’s weight; and it needed to be able to perceive its environment and the perch in relation to its own position, speed, and trajectory.
The researchers achieved all of that by equipping the ornithopter with a fully on-board computer and navigation system, which was complemented by an external motion-capture system to help it determine positioning. In addition, its leg-claw appendage was finely calibrated to compensate for the oscillations of flight as it attempted to home in on and grasp the perch. Once perched, the robot remains without energy expenditure.
Zufferey and the team wound up building two claw-footed ornithopters to replicate their perching results.
Here is a Tech Briefs interview, edited for clarity and length, with Zufferey.
Tech Briefs: What inspired the research?
Zufferey: It’s part of a larger project. We’re trying to do what we call manipulation with flapping robots. The idea here is to have robots that can physically interact with the environment — trees, infrastructure, industrial applications, stuff like that. The first step toward being able to do any kind of manipulation would be to physically perch.
Tech Briefs: What were the biggest technical challenges you faced on the journey?
Zufferey: The biggest challenge was to integrate everything into one platform. By everything I mean we have a claw system that has to hold the weight of the robot once we land; we need to compensate the oscillations; and we need to actually get to the branch in the first place. Integrating all that and mastering all those aspects into one flying robot that has to execute everything, and be very precisely timed, was difficult.
Tech Briefs: Would you mind explaining in very simple terms how the technology works?
Zufferey: We did the perching maneuver, which starts by launching the robot with a catapult to a fixed speed. At that point, the wings take over and flap about 60 degrees at around four times per second. The robot then follows a trajectory toward the branch, and as we close into the branch, we detect it with a sensor that’s on board, and compensate for the oscillations that are generated by the flapping wings during flight. Once that is compensated, we then reach very close to the branch and, mechanically, the claw latches onto the branch.
The claw has been designed so that it has sufficient strength to hold the robot on the branch without using any energy. That was particularly important; we want to be able to stay on the branch without having any motors enabled, for example.
Tech Briefs: What’s the next step and how many phases are there altogether?
Zufferey: The next step, in my opinion, is flying outdoors. Right now, the maneuver was performed inside in controlled conditions. For example, there’s no wind, the light condition never changes, everything is very controlled, and we have a motion-capture system inside. It’s a system that allows us to know, with high precision, where the robot is at every point in time.
Once you move this whole task outside, you’re faced with a lot of new challenges — you have to rely on satellite data, GPS, which is slightly less accurate. You’re dealing with light changes, wind, and all those different things. The next step is definitely to be able to have a bird that’s able to fly, detect the branch, and land outside.
Tech Briefs: How long do you think it will be before the tech will be ubiquitous or commercialized?
Zufferey: There is a lot of improvement that is happening on that kind of aspect with standard drones that you can buy commercially. So now it’s a matter of transferring this technology a bit toward flapping robots. So hopefully not too far in the future.
I’m hoping that in the next decade there will be a lot of new, exciting developments with those flapping robots. It’s quite an active field of research; I think we’ll see a lot more results soon.