Imagine a robot that ships as a flat pack and assembles itself on demand. Search-and-rescue missions might use them to squeeze into and scout the inside of a collapsed mine. They could be deployed into space for autonomous exploration, or dropped into disaster zones for rapidly-created shelters. And wouldn’t it be great if your next piece of IKEA furniture could put itself together? While we don’t rely on robots for such tasks today, it’s primarily because they’re expensive, and can take years to develop.
Now, a collaboration between engineers and computer scientists from Harvard University and the Massachusetts Institute of Technology hopes to change that, laying the foundation for wider access to at least certain kinds of robots. The group, led by Robert Wood, a professor of engineering at Harvard and founder of its Microrobotics Lab, describes the method in a paper published today in Science. Here’s what it looks like in practice:
Just as interesting as a self-folding robot is the inspiration for its design: Origami, the Japanese art of paper-folding. The idea isn’t revolutionary—the team has used computational origami to build a self-folding inchworm and lamp, but both needed human intervention. “We wanted to build a complex machine that could fold itself and operate without any human support,” says Sam Felton, a doctoral candidate in Wood’s Harvard Microrobotics Lab. The effort, the team says, proves that high-functioning robots can be assembled autonomously.
So how does it work? The robot is made from composite paper and Shrinky Dinks (plastic sheets that shrink on baking—yes, these) and equipped with two motors, two batteries, and self-folding hinges that are activated by heat. A microcontroller—which Felton refers to as “the robot’s brain”—heats the hinges via electrical signaling. “When electricity is applied [to the hinges], they heat up—like turning on a light bulb—causing them to contract,” Felton explains. After just four minutes, the robot has gone from flat to functional, and stiff enough to crawl away. (It can’t yet reverse the process and unfold).
Currently, the materials limit what the robots can do. The Shrinky Dinks are simple and cheap, but come only in standard letter-sized sheets and can’t support much weight. There are stronger materials out there, Felton notes, but not yet in the size they would need to build a bigger robot.
Much smaller robots would also be tricky. Right now, the team uses a commercial laser system to make the robot parts. It works to a resolution of one millimeter (0.04 inches), which was already imprecise enough to pose problems. “One of the biggest challenges for making this robot was packing all of the necessary components onto the sheet in such a way that the heat from the hinges didn’t affect the other electrical components,” Felton says.
Still, with better materials, these limitations could be overcome. Since the method doesn’t require complex parts and construction, it’s cheap—just $100 for the raw materials—and quick. It might not necessarily produce the kinds of powerful and sophisticated machines that build cars and perform surgery, but the team believes there’s a lot of scope to increase the foldable robots’ computational power and complexity of movement. Their ultimate goal is a “24-hour robot shop,” where you could walk in today asking for a playmate for your cat, and pick up your feline’s friend tomorrow.