Tumbleweed Rover for Marathon Martian Journeys

An anonymous reader writes "A prototype Mars rover, the Tumbleweed, has completed its 40 mile trek across the Antarctic, driven only by winds even in rough terrain over eight days. While the current rovers are designed for flat, equatorial regions, the tumbleweed design is geared to cover longer distances across what many consider the more interesting and dangerous polar regions on Mars."

Tumbleweed design?

[James A. H. Joyce]

Sure, a wheel is efficient at getting maximum distance out of minimum impulse, but doesn't there need to be some kind of ratchet-like one-way mechanism so it doesn't get blown backwards?

Re:I expected to see some kind of rover but...

[werfele]

I suspect it was inspired by this rover [tvacres.com].

Difficult task

[Doesn't_Comment_Code]

I built a robot for a class once. And after completing it, I have much more respect for the people trying to design these things.

It's not so hard to make a robot that works. It's all the possible problems that make it hard. And if your robot happens to be on Mars when it breaks... you're SO screwed!

All the little things... unanticipated terrain, sensor malfunction, wheel gets stuck... the list is endless. Of course, during my project I kept to the smooth carpet and just moved the robot whenever something happened. But these guys have got their work cut out for them.

No pressure, its just a 3 million dollar robot!

Broad purpose robot

[darth_MALL]

I would imagine the lack of directional control would make this unit perform very generic atmospheric readings/mapping whatever. What does that solve that an orbiter and beagle type unit combo couldn't?

What about crevasses?

[Hayzeus]

Or even relatively shallow craters? These would appear to end such missions prematurely. Is the plan, then, to drop multiple such probes (IE, faster/cheaper)?

Pelt the planet with these things...

[mobiux]

I saw we make a bunch of them and launch them shotgun style at the planet. and let them wander around collecting data. Temp, random soil tests etc.

Then have a mothership satelite or two orbiting trying to pic up a ready to transfer signal from the units and relay the data back to earth.

Serendipity

[Tablizer]

I like this part: "[image of] The spherically-wheeled rover [tricycle with 3 ballon wheels] that inadvertently gave birth to the idea for a giant tumbleweed ball. When one of the wheels broke loose during a test, it traveled across the terrain only too well."

Re:Not a short-term solution

[Doesn't_Comment_Code]

If the rover could switch modes - blow up like a balloon or deflate to be like a more traditional rover - it could do all its tests in regular mode and only use the balloon to cover large distances between testing points.

Non-wheeled rovers are not new

[csoto]

There are many interesting designs that take "unguided" tours of terrain. These include the "tumblers" like the one pictured, as well as a number of "glider" or "floater" designs. The gliders are basically "hover around over a spot and go where the winds take you" approach. They often have a *little* bit of navigation capability (but are mostly at the mercy of winds). The balloon idea is pretty smart. Basically, attach a bunch of sensors along a serpentine "spine" that is attached to a helium balloon. During the day, the balloon warms up and the rover flies to a new destination (wherever the winds take it). Sensors useful in the air (radar, atmospheric, etc.) take over. At night, the helium cools and it touches back down. Sensors useful on the ground (spectrometers, RATs, etc.) now fire up. Brilliant idea because of the simplicity and cost effectiveness.

Re:Not a short-term solution

[smack_attack]

Wouldn't you run the risk of having it get buried in sand/ice and then be completely immobile?

Re:Tumbleweed design?

[mobiux]

The point is that it goes wherever the wind takes it, so there is no backwards, just "where the wind takes it"

Forwards and backwards imply a destination, but it is more of a journey.

Some improvement suggestions

[mnmlst]

First off, to get some of the positional data in a fashion similar to using the Earth's Iridium satellite network, NASA could drop some RF triangulation devices on Mars. This seems like a cheap, viable option to me.

I would think NASA would like to have their "tumbleweeds" stay parked when something fantastic was found. Perhaps the probe could drop anchor now and then. I'm thinking a magnetized metal disc could be dropped outside the ball and attract to an electromagnet inside the ball. When the engineers are ready to let the ball move again, they switch off the electomagnet's current. Given six disposable discs, a ball could be anchored six times and no need for a motor or drilling system.

Also, how about a kite and harness rig? If the ball/probe needs to make a long run, it could have a harness around it attached to an axle running through it. A kite, attached to the harness, would then be launched from the probe and set it off on a faster run than just having air blowing against a ball on the Martian surface. The harness and kite could be dropped if the probe needed to "be free".

BTW, I highly recommend actually reading the links referenced as I am already seeing a lot of duplicate comments here as in previous discussions. Moderators in particular should check those links, unless you like modding up dupes...

Re:Does this mean...

[Short Circuit]

I hope they set up some sort of GPS system, then. You're going to need to find out where you are after every dust storm.

Fortunately, it should be possible to do it with relatively few satellites. Just three, in fact, for a base system, if all are geosynchronous.

And the ball should have some means of getting itself out of a bowl. Maybe a high-RPM motor inside that runs off a large capacitor to give a sudden burst of torque.

Re:Does this mean...

[dekashizl]

can't they do any better with the incredible lag due to the distance between Earth and Mars?

The lag (~10 min) affects our real-time control ability, but much of what makes the current rovers (Spirit and Opportunity) so advanced is their autonomous navigation abilities. Ground control says something like "move 100m north" and the rover figures out how to get there, drives itself, so lag isn't an issue at all. They've just been fairly cautious so far utilizing this. As the mission gets further along and past 100% success point, you'll likely see longer autonomous drives, since there is less to lose at that point of rovers fall off a cliff.

--
For news, status, updates, scientific info, images, video, and more, check out:
(AXCH) 2004 Mars Exploration Rovers - News, Status, Technical Info, History [axonchisel.net].

Re:Does this mean...

[Tablizer]

That the rovers that go there will move faster than a few feet per day?

That is an exaggeration. The rovers can go at least about 100 feet a day, but often stop to look around or poke rocks. Look how far Spirit [nasa.gov] has gone. It's lander is little more than a spec in this scene. I bet Opportunity will go even further per day when it moves out of the crater because there are less rocks in the way to study or stump the rover.

Re:Does this mean...

[joeljkp]

Their current plan is that if it gets stuck, then it becomes a "stationary sensor" (which could be valuable as well). These things would be cheap, light, and small... you could drop 50 of them. So if a few get stuck in weird places, it won't do any harm. And you get to study those weird places for a long time.

Re:Does this mean...

[shokk]

"Oh, look...there went that interesting outcropping we could have explored. Wait, wait...no, wind still hasn't died down. Damn this is the third time around the planet and we still haven't gathered any useful data."

What do you do for solar panels when the thing is round? Could a round panel-covered object still gather sufficient power to run the computers that will be inside, or is the wind supposed to power that as well? Off to read the article... =)

Lots more info

[Jafa]

Nasa's site has a lot more info, especially if you do a search [nasa.gov] on their site for 'tumbleweed'.
Some early research [nasa.gov]
Video [nasa.gov] from June 2001.

J