Robotics engineers are creating new designs for autonomous extraterrestrial rovers that can roll, step, and scuff their wheeled “feet” up sand piles as steep as 28 degrees.
Not that the goal is to hunt out the steepest sand dunes on other planets. Soft soil was a problem for NASA’s Spirit rover, which operated on Mars for five years before becoming trapped in just such a surface.
“It wasn’t able to free itself,” says Siddharth Shrivastava, a student at Georgia Institute of Technology, US, who notes that this eventually led to the vehicle’s death, nearly a decade before its twin, Opportunity, succumbed to the Martian weather.
And it isn’t just Mars where sand and other soft surfaces can pose deadly traps. Similar problems, Shrivastava says, may be encountered in future lunar missions, especially because some of the most geologically interesting and potentially resource-rich areas lie near the moon’s south pole, where the regolith is far less consolidated than that encountered by NASA’s Apollo missions.
“Any rover that goes to this region might encounter the same issues that Spirit did,” he says.
As anyone who’s ever tried to power a bicycle through loose sand knows, it’s not a medium that favors wheeled vehicles. What’s needed is the ability to pick up your feet and walk, rather than just trying to spin your wheels until you reach firmer terrain.
The same principle, Shrivastava said last week at a meeting of the American Physical Society in Boston, Massachusetts, US, applies to rovers.
To test how best to attack this problem, he built a 2.1-kilogram scaled-down version of a rover NASA was considering for its now-abandoned lunar Resource Prospector mission — a design in which the rover had the ability to independently lift each leg and swivel its wheels sideways in what he calls a shoveling or paddling motion — though in a human it might be more like scuffing your feet.{%recommended 8608%}
“The rover we came up with can initiate a walking pattern if it encounters a risky medium,” he says.
In its most basic such gait, the rover lifts its wheeled legs one at a time, while using the others to push forward.
It looks awkward, but it allows the rover to traverse soft surfaces at something close to its normal driving speed, Shrivastava says.
It also gives the rover much greater propulsive force than it can get simply by trying to drive forward.
“If you initiate the leg motion, you’re able to have a propulsive force that’s almost twice as large as that of just a wheel,” he says.
But that’s just the beginning.
Shrivastava tested his model rover on steep slopes of poppy seeds, chosen partly because they don’t produce fine particulates that pose a health risk to researchers, and partly because they are soft enough not to cause damage if they get in the robot’s gears.
Doing so, he found that it was possible to design an even more potent mode of locomotion that allowed the rover to power its way up loose surfaces as steep as about 28 degrees — far steeper, he says, than any past rover has ever been able to climb.
It works partly by using the front wheels to dig into the slope ahead of them. This shoves sand … or poppy seeds … backward, toward the rear wheels, an effect that can be magnified by having the front wheels lift and re-plant in walking-like behaviour.
Then, via a combination of spinning and sweeping, the rear wheels shove this loose material back behind the rover, producing a mound against which it can now push off.
Shrivastava calls the process a “granular conveyor belt” or “terrain remodeling via controlled avalanches” — a technical mouthful describing something that has never been tried before.
His coworker, Andras Karsai, also of Georgia Institute of Technology, says that, in effect, the rover is carving out a localised platform in the slope it’s trying to climb, about eight or nine degrees less steep than the slope itself.
That makes “a mess of the terrain”, he says, but it allows a rover that could normally climb a slope of about 20 degrees to work its way up a steeper one on which it would otherwise have no hope.
“It uses [this] to bring the local slope down to about 20 degrees and climb it successfully,” he says, making it a useful advance, whether it is ultimately deployed on the moon, Mars, or worlds beyond.