In just a month, teams will guide this 300-pound, 6-foot-tall humanoid robot in a series of simulated trials as part of NASA’s Space Robotics Challenge.
Valkyrie, NASA’s humanoid space robot. She stands on her own, the yellow harness is only to catch her if she falls. For this demo, her hands were replaced with equally weighted dummy versions. (Click image to enlarge.)
S&T: Monica Young
Meet Valkyrie: She’s 6 feet and 2 inches tall, weighs about 300 pounds, and cost $2 million — and one day this humanoid space robot, or more likely her much more advanced descendant, might help humans colonize Mars.
Although that ultimate goal is still a long ways off, Valkyrie, a prototype developed by NASA’s Johnson Space Center, will take her first Martian-specific test next month, as 20 teams guide a simulated version of the robot through a set of scenarios. Winners may take home a hefty prize: $1 million is on the table.
Valkyrie: A Tour
At first sight, Valkyrie looks not unlike Iron Man, the glowing circle on her chest marking her status (blue when motors are engaged, for example). But draw back the infrared-transparent faceplate and instead of Tony Stark, you’ll find a whirring LIDAR sensor that’s constantly scanning the surroundings for objects and obstacles.
Cameras and sensors abound on Valkyrie — in addition to a MultiSense SL camera on her head, which combines laser, 3D stereo, and video to get a sense of the environment around her, additional “hazard cameras” look ahead and behind from her torso. On each three-fingered hand, 38 sensors help maintain dexterity. Numerous small motors (actuators, in engineer-speak) control the robot’s 44 degrees of freedom, including seven-jointed arms.
The brains of the robot are two Intel Core i7 computers, which combine the sensors’ input and determine the best course of action. And, while testing is generally done with a power cord in place, Valkyrie also has a battery pack that’s good for about an hour.
All of that sensing, computing, and moving will be essential to Valkyrie’s future mission on Mars. While NASA initially developed Valkyrie (or, as she was initially called, R5), for disaster relief, in 2014 the agency changed course to reconfigure her for deep space. As one of a class of « caretaker » robots, Valkyrie might help set up living compounds on Mars, maintaining power and life support systems until the humans arrive.
Space Robotics Challenge
To that end, Valkyrie and the humans developing her are now facing space-based trials, such as the the Space Robotics Challenge that takes place in June. Here’s the setup:
In the not too distant future, R5 has arrived on Mars along with supplies ahead of a human mission. Overnight a dust storm damaged the habitat and solar array, and caused the primary communication antenna to become misaligned. R5 must now repair an air leak in the habitat, deploy a new solar panel, and align the communication antenna.
So basically, Valkyrie will replay the plot of The Martian, minus the part about growing potatoes (because, of course, she doesn’t eat potatoes).
Valkyrie navigates a doorframe. Out of view, a researcher controls Valkyrie using HTC Vive, a virtual reality interface that provides users with a 3D view of the world around the robot. The black cord provides Valkyrie’s power.
S&T: Monica Young
Twenty teams will compete in the challenge, which is entirely simulated: teams will issue computer commands to a virtual Valkyrie, who will operate within a virtual Martian backdrop. But virtual doesn’t mean easy — teams will have to deal with strictly defined tasks, limited bandwidth, and a built-in latency designed to mimic the communications delay between Earth and Mars. You can follow the teams’ progress on Twitter.
But the Space Robotics Challenge is only one step of a marathon. NASA has also provided University of Massachusetts, Lowell, MIT, and Edinburgh University in Scotland, with three Valkyrie prototypes. These three teams are teaching Valkyrie complex subjects, such as how to walk across uneven surfaces and grasp different-shaped objects. They’re also learning how robots and humans can best communicate with each other.
Why a Humanoid?
Whether you follow Valkyrie’s progress on Twitter or see a live demo of her abilities, as I did at the NERVE (New England Robotics Validation and Experimentation) Center in April, one thing soon becomes clear: making a humanoid robot is hard. On the one hand, it’s pretty incredible to watch a person-shaped machine walk up and down a ramp or navigate a doorway. On the other hand, it’s also incredible how much work it has taken to get here.
“Valkyrie can currently walk over flat terrain by balancing her weight at all times,” says Taskin Padir (Northeastern University). “That’s why it looks slightly different than human walking.” The ability to walk on soft or uneven terrain, he adds, is the next goal.
Watch Valkyrie show off her balancing skills in this video:
Replicating humans’ ability to walk has long been a robotics goal, but it’s still not foolproof*, which is why space-based robots (such as Spirit, Opportunity, and Curiosity) have so far still relied on wheels to get around. But when it comes to, say, building human habitats on Mars, robots will have to be human-shaped — because ultimately, it will be humans, not robots, living and working on Mars.
* At least one other humanoid robot has made strides in walking ability, but Atlas, created by Boston Dynamics, is different from Valkyrie in some key ways. It can navigate freakily well across uneven terrain thanks in part to its hydraulic power, but fluid pressure (and possible accompanying leaks) is not particularly well-suited to space. So Valkyrie uses less powerful but more reliable electric actuators to keep her balance.