Foldable Drone Changes Its Shape in Mid-Air

This quadrotor can alter its shape in flight depending on where it needs to go

4 min read
Foldable Drone Changes Its Shape in Mid-Air
The foldable drone and its X and O morphologies.
Photo: University of Zurich and EPFL

Quadrotors are fast, cheap, and capable, and they’re getting smarter all the time. Where they struggle a little bit is with adaptation. Many other kinds of robots can change their structure to better perform different tasks: Humanoids do it all the time, with all those conveniently placed limbs. Hey, wouldn’t it be cool if drones had movable limbs too? Yes, it would. Someone should figure out how to do that.

We’ve seen some drones in the past that can alter their shape in flight, but this Folding Drone (developed by roboticists from University of Zurich and EPFL) is different in several important ways. Each of its four arms has a servo motor at the base that can rotate one propeller independently, with some height differences between the arms making sure that the Folding Drone doesn’t immediately blenderize itself. While there are some arm location combinations that are particularly handy, like completely unfolded (“X” morphology), completely folded (“O”), straight line (“H”), and partly folded (“T”), the drone isn’t limited to those shapes, and it remains fully stable and controllable wherever its arms happen to be, even if the configuration is asymmetric. It’s not easy to do this—it requires “an adaptive control scheme able to cope in real-time with the dynamic morphology of the vehicle.”

“We exploit the morphing to adapt the vehicle’s size to tasks such as traversing gaps, inspecting surfaces, or transporting objects. However, we believe that a morphing quadrotor can tailor its shape to more dynamic tasks, like for example flying at high speed, where it can improve its performance by folding to change its aerodynamic properties. This would allow very fast flight in time-critical scenarios.”

While the Folding Drone doesn’t have quite as many degrees of freedom as that crazy flying dragon robot from ICRA, it’s much less complex and expensive, and can transform very quickly, as you can see in the video. It may not be able to wrap itself around anyone’s neck and slowly strangle them, but we’ll let that slide just this once. The (relatively) simple design helps the Folding Drone maintain both efficiency and versatility, allowing it to operate autonomously with onboard sensing and computing. It doesn’t sound like the drone can autonomously decide how to reconfigure itself to get past an obstacle yet, but the researchers are definitely working on that.

They’re also working on some other cool things, and for more about that, we spoke with Davide Scaramuzza at the University of Zurich via email.

IEEE Spectrum: Why is a drone that transforms like this better than a drone that’s simply small enough to pass through small gaps?

Davide Scaramuzza: Small drones have several disadvantages: They are less stable in case of external disturbances (e.g., wind), they have very limited flight time, and cannot transport large payloads.

Why did you choose these particular degrees of freedom for the transforming ability?

We chose these degrees of freedom because they allowed us not to increase the complexity of the vehicle both from in terms of mechanical design and from a control perspective. At the same time, rotating the arms around the main body leads to very compact shapes, reducing the overall size of the platform significantly.

Foldable Drone Close-up of the foldable drone and its components: (1) Qualcomm Snapdragon Flight onboard computer, equipped with a quad-core ARM processor, 2 GB of RAM, an IMU, and two cameras. (2) Qualcomm Snapdragon flight electronic speed controller. (3) Arduino Nano microcontroller. (4) Servo motors used to fold the arms. Image: University of Zurich and EPFL

What are the disadvantages of a drone that can morph?

The main disadvantage of morphing is the potential reduction of stability and flight time compared to the standard X configuration. Some morphologies are significantly less stable than others, and reduce the flight time significantly. Therefore, morphing comes at the cost of sacrificing this two aspects. The amount of such sacrifice depends on the desired morphology. For example, the O morphology allows the robot to become very compact (each dimension becomes roughly 20 percent smaller than in the X morphology), but can produce lower angular accelerations (which translates into less robustness against external disturbances) and a flight time approximately 60 percent lower than in the X shape.

The T and H configurations, instead, are capable of providing higher angular accelerations around one of the body axes, but sacrifice their agility around the other axis. This shows that the standard X morphology is the most efficient and therefore should be used as long as a different morphology is not strictly required by the task at hand.

The paper mentions that you’re working on “exploitation of the morphology for improved flight at high-speed, and novel, bio-inspired mechanical designs.” Can you elaborate?

At the moment, we exploit the morphing to adapt the vehicle’s size to tasks such as traversing gaps, inspecting surfaces or transporting objects. However, we believe that a morphing quadrotor can tailor its shape to more dynamic tasks than those shown in the paper, as for example flying at high speed, where it can improve its performance by folding to change its aerodynamic properties. This would allow very fast flight in time-critical scenarios, allowing the vehicle to navigate through large areas in very little time.

“The Foldable Drone: A Morphing Quadrotor that can Squeeze and Fly,” by D. Falanga, K. Kleber, S. Mintchev, D. Floreano, and D. Scaramuzza from University of Zurich and EPFL, appears in IEEE Robotics and Automation Letters.

[ Foldable Drone ]

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman
LightGreen

“I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

This article is part of our special report on AI, “The Great AI Reckoning.”

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less