Exoskeletons In My Closet: What Raytheon’s Robotic Suit Really Means for the Field
Last week, Waltham, MA-based defense contractor Raytheon (NYSE: RTN) unveiled its latest prototype “exoskeleton.” This is a powered robotic suit that a soldier or worker could strap on in the field to enable them to load heavy equipment faster, carry supplies or munitions using less energy, or—let’s face it—just look ultra-cool. Raytheon said in a statement that the new robotic suit is “lighter, stronger, and faster than its predecessor, yet it uses 50 percent less power.” The device is powered by high-pressure hydraulics and gives its wearer some degree of super strength.
In the demo, which took place at Raytheon’s Sarcos subsidiary in Salt Lake City, UT, an engineer wearing the suit (which includes arms and legs) punched through some boards, did pushups, and lifted weights with little effort. The news was reported fairly breathlessly by media outlets including CNET, Wired, Scientific American, and the L.A. Times. And I understand why—it’s a sexy technology that conjures up visions of “Iron Man” and mythical references to superhuman strength. Plus it’s far more accessible than all the top-secret stuff Raytheon does that is actually useful for the military—radar systems, cybersecurity, missile defense, and so forth.
But I wondered how much progress has really been made in exoskeletons—in the fundamental robotics, sensing, control, and energy technologies necessary to make a robot suit powerful, safe, and reliable to move around in. Raytheon declined to be interviewed for this story, but I did some digging around.
After all, I’ve been following the field since 2001, when I worked in the old Leg Lab at MIT, which was home to robots that could walk, run, hop, and keep their balance. Back then, the main problems with designing a robotic exoskeleton were how to make it powerful without being clunky, how to control it safely, and how to supply enough energy to it.
In 2002, I attended a private meeting of the U.S. Defense Advanced Research Projects Agency program on “exoskeletons for human performance augmentation” (from which I still sport a nifty backpack, though it doesn’t give me super strength). The program manager was Ephrahim Garcia, a professor at Cornell University, who later handed it off to engineer John Main. At the time, about a dozen universities and research groups were competing to build exoskeleton technologies for DARPA, and Sarcos had one of the designs that eventually won out. A couple years later, I visited Sarcos (which Raytheon acquired in 2007) and got a tour of the Utah lab and an early demo from its leader, Steve Jacobsen, for a photo essay in Technology Review.
My first impression from the demo last week was that not much has changed in the field in the past decade. That’s a bit surprising, since other kinds of robots—Predator aerial drones, PackBots, Roombas—have become increasingly sophisticated as they’ve been commercialized and deployed by the military. As it turns out, though, my first impression of the Raytheon device was not entirely correct.
“They’ve clearly demonstrated increases in strength,” says Hugh Herr, a professor who leads the biomechatronics group at the MIT Media Lab, which works on things like smart prosthetic devices to help disabled people walk better. (Author’s note: Herr was my faculty advisor in the Leg Lab when I was a postdoc.) “If the mission is to carry very heavy objects, or to lift an object from the floor to a table, then it clearly works,” he says.
So here’s my bottom line. A few things have improved: the hardware is cheaper, stronger, and lighter; the control is a little smoother; and the devices are more powerful. But two fundamental problems remain, and they are somewhat related.
One is mobility. Part of the original vision of the DARPA program was to create a suit that soldiers could wear to augment their movement capabilities in the field—think running 10 miles without breaking a sweat, or carrying around 200 pounds of equipment effortlessly. But when it comes to exoskeletons, Herr says he is “not aware of any improvements in locomotion” as of yet. Presumably that’s because strapping on bulky equipment changes people’s natural gait and makes it harder to walk smoothly.
For his part, Herr’s research group is working on a lightweight, low-power, lower-body exoskeleton focused not on increasing strength but on “allowing humans to get from point A to point B with less energy,” he says. Applications of such a device might include rehabilitation, defense, and sporting equipment—but it certainly won’t be easy to make it work right.
“Humans have evolved over millions of years,” Herr says. “Something like walking is highly refined. It’s exceedingly difficult to get a machine to improve on those efficient dynamics.”
The other big challenge in exoskeletons is how to power the device. Hydraulics is Raytheon’s preferred method. But hooking up a soldier or worker to a high-pressure fluid line won’t work in the field. So the company has devised a backpack-size internal combustion engine running on gasoline (or other liquid fuel); that power source drives hydraulic fluid to control the robotic joints. Clearly, a soldier won’t want to carry something noisy and smelly that could blow up in the field. But it’s very hard to do better than an internal combustion engine, in terms of delivering power efficiently. So this remains a great unsolved problem for exoskeletons—and robots in general.
And, much more broadly, for alternative energy. If scientists ever solve the portable energy problem, they will find many other things to apply it to besides exoskeletons. Whether it’s a fundamentally new kind of battery or energy storage device, or a new kind of fuel, it could potentially change the cleantech landscape. Indeed, the societal context around alternative energy and green technology has changed a lot more in the past decade than the issues around personal robotics.
The Boston area should help lead the way on these fronts. Besides Raytheon and MIT, Massachusetts is home to robotics companies like iRobot, Boston Dynamics (Leg Lab founder Marc Raibert’s company), CyPhy Works, Harvest Automation, Heartland Robotics, iWalk (Herr’s company), Myomo, and U.S. defense facilities such as the Natick Soldier Systems Center.
For my money, though, I would not expect to see a soldier, or anyone else, wearing an exoskeleton for at least another decade. Though it might make a nice Halloween costume in a few weeks.