NASA is exploring the potential for human robot design with Robonaut 2, a humanoid machine currently assisting astronauts on the International Space Station. The chief design goal for the robot -- a collaboration between the space agency and General Motors -- was manual dexterity, as it was built to perform simple, repetitive, or potentially dangerous tasks for astronauts.
To show progress the agency has made since it sent Robonaut 2 to the International Space Station aboard the Space Shuttle Discovery in February 2011, NASA recently unveiled a video on its website of Robonaut 2 that features the robot shaking an astronaut’s hand on the space station much like a human would, as well as performing other manual tasks. (Watch a video of Robonaut 2 in action below.)
NASA has designed Robonaut 2 to perform a series of tasks with its hands much like humans would. The robot is currently helping astronauts aboard the International Space Station. (Source: NASA)
“That was an amazing moment,” Robonaut deputy project manager from NASA’s Johnson Space Center Nic Radford said of the handshake. Radford speaks on another NASA
video in an interview with NASA public affairs officer Dan Huot. Radford said Robonaut 2 was designed specifically to be put to work alongside astronauts, and in order to accomplish the tasks NASA has planned, it needed “to have very capable hands and very sensate hands.”
Robonaut 2’s ability to feel with its hands -- which have a grasping force of about five pounds per finger -- comes from touch sensors at the tips of its fingers. The robot is controlled by station crew members or through telepresence with controllers on the ground. However, it was also designed to act autonomously to perform tasks, with astronauts needing only to provide periodic status checks. It can move its arms up to 2 m/s, with about 12 degrees of freedom in its hands and two degrees of freedom in its wrist. It has a 40-pound payload capacity, and overall, includes more than 350 sensors and 38 PowerPC processors.
NASA has divided the robot’s experiments on the International Space Station into two stages -- free-space motions, such as waving its hands; and contact operations, such as flipping switches on machines and using objects in an intelligent way, according to Radford. He said Robonaut is unique in that it’s one of the first payloads on the space station than can actually impart forces on its environment. Safety was a major factor in its design, which includes a feature to shut down automatically if it comes into unexpected contact with an astronaut or equipment.
In fact, before the handshake between Robonaut 2 and astronaut Dan Burbank, another astronaut floated by and accidentally brushed the robot, Radford said. The robot immediately shut itself down and restarted as a safety measure. “We have a hair trigger safety system right now -- that is by design,” he said. In that particular situation, he said, “it reacted in the way it was supposed to.”
Wow, between this development and Chuck's slide show on intelligent highways and cars, it's quite a wake-up call to the 21st or maybe even 22nd century!
I definitely applaud the idea of sending robots into space to perform the tasks that humans can't or shouldn't. I'm assuming a lot had to go into the design to enable the humanoid machine to function properly despite the laws of gravity. Too bad we're pulling back on space exploration research at a time when we have all this new technology to help uncover valuable insights.
This is very cool. A great first step. These can also be used in deep sea diving reseach. SInce 1960, we've been to the bottom of the Mariana Trench 4 times. We could move faster and even stay down longer if decommpression sickness isn't an issue.
I thought the GM co-branding, on the chest in Nascar-style, was interesting.
Considering how complex the human body is, the limited mobility it has is impressive.
Beth, activities like space exploration stimulate the economy much more than the construction industry and local government as was done in the recent stimulus. It also stresses engineering and gets innovation into the "civilian" economy fairly quickly.
You are correct also that we are steadily moving forward (21st century) despite the economic issues. This is some cause for optimism.
I agree Beth, it is a shame that space exploration research has been curtailed. Not only did it generate new technology and bring people a level of enthusiasm and solidarity in past decades that little else could come close to - it also created invaluable spin off technologies that both improved life and stimulated the economy.
That is another important aspect of STEM, keeping space in front of our kids so that they still grow up with a sense of wonder that only the stars can bring about. We are frequent visitors to the McDonald Observatory near Fort Davis and brought our son on his 13th birthday for a special viewing that is only held a few times a year through the 109" telescope. I am guessing there were about thirty people in our group and our son was the only kid...
2 degrees of freedom in the wrist, and about 12 degrees of freedom in the hand ? Is this supposed to be the equivalent of 'axes of movement' ? I guess the wrist can rotate and bend = 2 axes of motion. The 'about 12' in the hand may be the finger segmants. Is it about 12 because they aren't sure how many ? Or because the individual joints have some interferences in certain movements ?
I can understand configuring a new robot to look like a human only to the extent that it will be operating equipment designed to be operated by humans. Human shape and configuration evolved under the strong effect of gravity.
If we're building robots and other equipment for use in space, it may be far more practical to omit robot design features used to deal with gravity such as legs, feet, and toes for transport and arms, hands and fingers for manipulation. I would think robots should look more like an octipus that evolved in near weightlessness. Equipment and robots would best be designed to work with each other, eliminating the physical human factors.
It almost looks like they're trying to put HAL's brain into a modified R2D2 body. Heaven help us when it becomes self aware.
jhankwitz has an interesting point--how much do robots in space need to have human parts or features when gravity isn't an issue? I think part of the answer is that gravity is an issue in a space station, and that fingers or some such appendage for manipulating is needed, at least when Robonaut 2 needs to flip switches, or when surgical robots are being deployed to service or refuel satellites: http://www.designnews.com/document.asp?doc_id=237609
jhankwitz, I can accept HAL in an R2D2 body. Your image of an autonomous octopus is the thing of nightmares.
Human interaction with automation is branching out in many interesting ways. Engadget.com (sorry for the reference to another technology site) has numerous articles about studies of ever-more-realistic human-form robots.
An octopus is absolutely a smarter, more efficient form factor. It may not be accepted by its users though.
Talk about the stuff of nightmares ... but given the issues that Ann mentioned, an octopus design might have more applicability in terms of serving up more "hands on deck" for jobs that require dexterity when it comes to small motor skills.
We did a story with GM on this, and part of their motivation is to explore the possibilities of humanoid robots being used in assembly areas. That would require working closely with human workers which creates interesting issues related to safety and productivity. Interesting technology.
Beth, It's interesting to me that the creep factor is a major consideration for you. Clearly there is a trend to mobile robots interacting more with human workers, from autonomous vehicles that are transporting materials in tire manufacturing plants, for example, to surgical assistants helping with organization and sterilization of instruments. In any of these applications where there is human interaction, I guess there is an adjustment to working with the robot. I guess it's the upper torso design that makes the difference in this case. It's interesting that GM sees the possibility long-term of service robots used in assembly areas, working in conjunction with human workers as a possibility.
What I've noticed with this one is that they didn't try to give it a face - which I think tends to be a major failing with a lot of the humanoid robots. The "helmet" look prevents the "creep factor" of something that looks "almost" human.