Skip to main content

Winching is a Cinch for These Wasp-inspired Robots

Early FlyCroTug prototype, shown lifting 600 grams. | ©EPFL / Laboratory of Intelligent Systems

A new design for tiny flying robots takes inspiration from wasps' ability to transport prey much heavier than themselves, by latching onto surfaces and dragging their catch. Mimicking the insects' latch-and-tug technique, the micro-drones described in this study can pull heavy objects toward themselves with a force up to 40 times their own weight, without compromising their quick and nimble flight capabilities.

The new feature described in the October 24 issue of Science Robotics helps bring small aerial robots, or micro air vehicles (MAVs), one step closer to real-world applications. MAVs could be used in search and rescue operations or as utility robots for performing tasks like closing a valve inside a factory without humans present.

"It will enable new applications in hard to reach areas, such as in cities or off shore where working at height is a major health hazard," said Mirko Kovac, director of the Aerial Robotics Lab at Imperial College London, who is unaffiliated with this study. He wrote a piece on advancing aerial robots in the 20 May 2016 issue of Science.

Currently, "search-and-rescue tasks are usually accomplished with larger, ground-based robots, but might well be addressed using small, specialized air vehicles working cooperatively," said Matthew Estrada, a recent Ph.D. graduate from Stanford University and lead author of the study.

For example, when faced with the ragged terrain of an earthquake disaster zone, or the cluttered ruins of a collapsed building, robots that fly would be able to navigate the environment more smoothly and reach more remote locations than terrestrial robots. Equipped with additional gadgets like sensors or cameras, aerial robots could serve as eagle eyes for vast stretches of land.

However, robotic flyers are still limited in flight times and weight restrictions. MAVs are ideal for accessing tight spaces or closely interacting with humans, but they currently fail to produce as much thrust or force as their larger counterparts. This means they cannot carry sensors or additional instruments that would allow them to be more useful in their environment.

"One of the challenges for current MAVs is that they are focused on flight only," said Kovac.

Estrada's team is interested in creating MAVs that can intelligently interact with their environment and use it to their advantage. They took inspiration from small, flying insects and the natural structures they use to accomplish a plethora of actions, including perching, climbing and dragging large prey back to a nest.

The researchers built wasp-inspired attachment mechanisms — microspines to mimic a wasp's claws and gecko-inspired adhesives to mimic sticky pads on the wasp's body — onto their palm-sized MAVs, named FlyCroTugs. These attachment tools were linked to an inner robotic winch for reeling in objects.

Estrada and his team demonstrated the capabilities of a FlyCroTug in a partially collapsed building, where it maneuvered through small crevices, adhered onto a concrete surface, and lifted a sensor 20 times its own weight.

As well, the researchers demonstrated that a team of two FlyCroTugs could coordinate with each other to open a heavy door. One MAV would slide a hook under the door and adhere to the floor's carpeting using microspines. The other would fly up to the door handle, lasso it, and stick to the door. While the first FlyCroTug pulls horizontally, the second robot stuck on the door pulls down on the handle, and together, instigating pull forces of up to 40 times their own weight, they open the door.

"Our design expands MAV capabilities by enabling them to exert large forces," said Estrada. One important breakthrough that they made in their study, he noted, was to show the feasibility of the hardware.

The future of their FlyCroTugs will consist of implementing a higher level of autonomy and planning capabilities on the robots, Estrada said. His team is working on preparing their MAVs to carry out more open-ended tasks in unknown environments, such as perceiving on the spot whether a surface is suitable for attachment, or calculating how many drones would be needed to lift a certain weight.

[Credit for associated image: ©EPFL / Laboratory of Intelligent Systems]


Juwon Song

Related Focus Areas