These Drones Can Perch and Dangle Like Birds and Bats

As drone technology continues to improve, Cuberg has ceated a lithium metal battery that can extend flight time.

This article originally appeared in Smithsonian Magazine.

Randy Rieland – When Parisian firefighters desperately sought to save Notre-Dame from total devastation, they relied on drones to show them where they needed to focus their efforts and position their hoses.

Meanwhile, UPS has started using drones, formally known as unmanned aerial vehicles (UAVs), to transport medical samples to and from buildings in a hospital network in Raleigh, North Carolina.

The U.S. Department of the Interior recently reported that it launched more than 10,000 drone flights last year, twice as many as in 2017. Their use in response to natural disasters spiked dramatically.

There’s not much question that drones have become a go-to tool for our times, a technology whose uses will only keep expanding. Yet, for all their potential, UAVs still face a big challenge—limited battery power. Most models can stay airborne for no more than 20 minutes before they run out of juice. Some flights can last 30 minutes, but that’s generally the limit.

Birds do it

Much research has focused on the batteries themselves. A startup named Cuberg, for instance, says it has developed a lithium metal battery that can extend flight time by 70 percent.

But an international team of scientists has taken a different approach, instead looking at ways to allow drones to save battery power by being able to “rest” during flights. Specifically, they’ve designed UAVs with landing gear that enables them to perch or balance on objects like birds.

“We have a few different perching strategies,” says Yale researcher Kaiyu Hang, lead author of a study recently published in Science Robotics. “Where it is totally perched, where it is grasping around something, like a bat, we can stop all the rotors and the energy consumption would become zero.”

Another option is what Hang calls “resting.” It involves using a landing device that enables a drone to balance on the edge of a surface, such as a box or a ledge. In that position, it would be able to shut down two of its four rotors, cutting consumption roughly in half. Another alternative makes it possible for a drone to sit on top of a small surface, such as a pole, a tactic that cuts energy use by about 70 percent, according to Hang.

The concept of perching drones isn’t new, but this research, Hang says, expands the types of surfaces on which UAVs can rest. The design of the landing gear resembles a gripping claw, with three fingers. What gives the device its versatility are different attachments that can be mounted to the fingers, depending on what kind of surface will be used for resting.

Hang compares it to changing the lens on a camera to adapt to different conditions. “It’s super difficult to design a landing gear that could work with every kind of environment,” he says. “But if you make it modular, it’s much easier to design grippers that will work with the surfaces the UAV is going to interact with. It’s providing different solutions instead of a single best solution.”

Neil Jacobstein, a noted Silicon Valley artificial intelligence and robotics expert who was not involved with this research, acknowledges its potential benefits. He says that while he wouldn’t necessarily describe it as a “breakthrough,” he thinks it’s “useful because of the low energy density of drone batteries. The capability of perching and resting enables drones to conserve power.”

Next steps

The goal is for these drones to use artificial intelligence to survey an environment and then choose the most suitable landing surface, Hang says. So far, all the research has been done in a laboratory so the scientists were able to use an external camera instead of installing them on the drones. They also didn’t have to deal with currents and other weather conditions that will make it more difficult for UAVs to land and stabilize on real-world surfaces.

“Outside, we’d have a lot of aerodynamic issues to deal with,” Hang says. “That’s one of the challenges of future development.” The first step, he notes, was to create a prototype that could show what was possible using modular components with drone landing gear. The team, however, has not applied for a patent. It’s been more of an academic project than a commercial one, Hang notes.

But Hang is enthusiastic about how these design innovations can have an impact on enhancing what drones can do. By being able to stabilize them more securely on different surfaces, for instance, it would make them capable of lifting objects, something a hovering UAV can’t do very well.

“With ropes, a drone could actually act as a pulley,” he says.

Hang also imagines a day when a drone could land at your window to make a delivery. “You wouldn’t have to allow drones to come into your house,” he says. “You’d be able to reach out and grab what they’re delivering. It would be like a bird sitting on your windowsill.”