In the future Jonathan Hurst envisions, robots will be able to help soldiers carry more equipment for longer distances. They’ll act as an exoskeleton to help disabled people get around in ways they couldn’t before. They’ll go jogging on rough terrain. They’ll be embedded with touch screens so they can follow us around the house while we do chores and simultaneously video-chat with distant loved ones, or watch some futuristic iteration of “Law and Order.”
It’s not a passive dream. The assistant professor of mechanical engineering has been passionate about robots since he was seven and building Star Wars Imperial AT-AT Walkers out of Legos. A couple of decades and a Ph.D. in robotics from Carnegie Mellon later, Hurst runs a lab at Oregon State University, where he is focused on creating a two-legged robot that can maneuver over uneven ground like a human can.
“One thing I can say about the goals of this lab is we want robots moving and doing things,” he says. “I want to be in here when a robot is hopping around and every week we’re taking things out to the practice football field and running tests.”
The theme of Hurst’s lab, as he explains it, is to create robots that not only move like humans, but also save energy by using the momentum from their swinging parts—rather than relying solely on more energy-intensive mechanized controls. While building ATRIAS, the latest robot creation in the lab, Hurst and students used fiberglass springs—the same kind used in archery bows—to mimic human tendons. The springs are stretchy and flexible, allowing ATRIAS to hop up and down and move like humans.
“If you’ve ever played capture the flag and it’s gotten dark, and you run across a big dip in the ground and you’re still running, you might feel you should have fallen. But your body can soak up those big changes in terrain,” Hurst says. “Robots can’t do that yet. But our robots will be able to.”
Researchers are trying to create a robot that uses a similar amount of energy a human does. Robots like Boston Dynamics’ “Big Dog,” for example, perform eerily well over rough terrain, but use a considerable amount of energy to do so. According to Hurst, the focus on energy efficiency is something that sets his lab apart from many others.
Hurst’s lab wouldn’t function, though, without the help of OSU engineering students, who have been vital to the ATRIAS project. Hurst needs students who can build devices, design electronics, and write software. Eight undergraduates and four graduate research assistants are working with Hurst currently, and their work spans mechanical, electrical and computer engineering. He has been impressed with their contributions.
“OSU students know how to build stuff,” says Hurst. “This place is really great for applied engineering. Students can do not only the analysis, but they know how to weld and machine, and put together a device that’s really going to work. They come up with ideas I’ve never thought of.”
Ultimately, Hurst wants to share his lab’s discoveries with other universities, so that other people can build on his foundation of knowledge. When a two-legged version of ATRIAS is complete, he plans to build two additional copies and send them to colleagues at Carnegie Mellon and the University of Michigan, where engineers are working on the same “legged locomotion” robotics. It will help him get to his version of the future even faster.
“Once you have a machine that could demo on a similar level as an animal or a human, then it’s going to be an industry,” he says. “It’s going to be like the automotive industry, where different companies make improvements over time to make things better and better.”