Life-Size Humanoid Robot Is Designed to Fall Over (and Over and Over)

Why try to develop a humanoid robot that doesn't fall over when you can instead just develop an armored one that can fall over and get up again?

3 min read
Life-Size Humanoid Robot Is Designed to Fall Over (And Over and Over)
Image: University of Tokyo

Roboticists worldwide are spending an obscene amount of time and effort trying to teach large humanoid robots how to not fall over. We rejoice every time there is even the smallest incremental bit of progress towards success, because not falling over is super hard, especially if you want your robot to be doing something useful. And even though some large humanoid robots can occasionally survive falling over, most of them don’t enjoy it very much.

Led by Kei Okada and Masayuki Inaba, a team from the University of Tokyo and Kawasaki Heavy Industries is working on their own life-sized humanoid robot, and they’ve come up with a new strategy for not worrying about falls: not worrying about falls. Instead, they’ve built their robot from the ground up with an armored structure that makes it totally okay with falling over and getting right back up again.

The concept behind Robust Humanoid Robot (RHP2) is, as the title of the paper suggests, that it should be able to work for a long time in hazardous situations like “a disaster site, a fire site or a wet environment.” The robot is tethered so that you don’t have to worry about power or connectivity, and since it can fall over and get up without any trouble, you won’t have to go rescue it and then waste time on repairs.

The researchers came up with a new strategy for not worrying about falls: not worrying about falls. Instead, they’ve built their robot from the ground up with an armored structure that makes it totally okay with falling over and getting right back up again

The fall protection system is designed to do three things: prevent direct mechanical breakage from contact with the ground, prevent indirect mechanical damage (like messed up joints and motors) that might otherwise be sustained during a fall, and prevent all of the fragile and expensive stuff like sensors and computers from fall damage. A lot of this is taken care of by an armored metal frame that encloses the entire robot, meaning that there are no weak points where environmental contact could cause damage. Places on the robot that are especially likely to experience stress (including knees, hands, chest, hips, elbows, and back) are particularly reinforced.

This version of RHP2 is powered by electric motors, although the researchers will eventually upgrade the whole system to hydraulic to give it more power. Rather than use conventional geared joint mechanisms, RHP2 uses linear actuators wherever possible because they’re much more robust. On top of all its metal armor, the robot can also put on different suits depending on what kind of environment it’ll be operating in:

RH2 robust robot that can fall The Robust Humanoid Platform (RHP) with a suit (top left), falling down (top right), and rendering of scenarios where the robot could be deployed (bottom). Image: University of Tokyo

RHP2 runs ROS and has a few autonomous behaviors, like being able to reposition its limbs during a fall and get itself standing again afterwards, but so far the researchers haven’t had a chance to emphasize practical, operational autonomy. Since the robot relies on that big tether anyway, it’s possible that realistic use-cases may involve telepresence, or DRC-style assistive telepresence.

It may be true that a robot that falls over all the time and just gets up again is not quite as elegant as a robot that doesn’t fall over at all, but we’re very far from the latter, and the former is ready to go right now. In the context of disaster relief, capable humanoid robots can’t come soon enough, so at least in the short term, a robust humanoid robot sounds like a good idea to us.

“Development of Life-Sized Humanoid Robot Platform with Robustness for Falling Down, Long Time Working and Error Occurrence,” by Yohei Kakiuchi, Masayuki Kamon, Nobuyasu Shimomura, Sou Yukizaki, Noriaki Takasugi, Shunichi Nozawa, Kei Okada, and Masayuki Inaba from the University of Tokyo and Kawasaki Heavy Industries, was presented this week at IROS 2017 in Vancouver, Canada.

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman

“I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

This article is part of our special report on AI, “The Great AI Reckoning.”

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less