Does NASA Have A Robot That Can Do This?
Mars Rover Views Spectacular Layered Rock Formations
“Curiosity took the images with its Mast Camera (Mastcam) on Sept. 8. The rover team plans to assemble several large, color mosaics from the multitude of images taken at this location in the near future. “Curiosity’s science team has been just thrilled to go on this road trip through a bit of the American desert Southwest on Mars,” said Curiosity Project Scientist Ashwin Vasavada, of NASA’s Jet Propulsion Laboratory, Pasadena, California. The Martian buttes and mesas rising above the surface are eroded remnants of ancient sandstone that originated when winds deposited sand after lower Mount Sharp had formed.”
Keith’s note: Boston Dynamics has robots that can do things that NASA’s R5 and Robonaut are simply incapable of doing. Yet NASA continues to pour money into their antiquated in-house hobby shop efforts when the private sector would happily sell them vastly more capable devices – devices that constantly improve. Look at the Mars Curiosity images that NASA featured today. They were taken by a rover with a limited ability to traverse terrain. A robot like the ones that Boston Dynamics makes could scramble up these scree slopes with a rock hammer and get samples. NASA’s broken R5 robot can’t even walk without a hoist to keep it upright.
– NASA Challenges People To Use Its Broken Robot To Fix Things on Mars, earlier post
– The Droid That NASA Should Be Sending To Mars, earlier post
– Earlier posts
While I don’t disagree with the thrust of your post, Keith, building a robot that can perform in a lab and building one that will survive & operate effectively on Mars (after launching & flying to Mars) are two different things. Perhaps they could build one that could perform there as you describe, but it’s certainly not going to be as easy as your phrasing suggests.
NASA’s robots can’t even match the performance of Boston Dynamics’ robots – in the lab.
The BD robots are on YouTube, walking around on the real world in the great outdoors and even surviving a guy trying to kick them over. Power issues aside, they’re pretty good!
There are different performance metrics to consider. Boston Dynamics robots require a lot of power. NASA’s rovers, since they have been proven to work on Mars, out-perform the Boston Dynamics robots on power requirements and possibly many other metrics.
Agreed, although the Boston Dynamics robots need a fair bit of internal power. BigDog required about 11 Kilowatts’ worth of on-board power to work, never mind the power it would need for instrumentation. You’d probably have to have it operate near a robotic fuel-making plant that could re-fill it between excursions.
Atlas is a bit better in that regard – the latest version had a 3.7 KW-hour battery pack that gave it about an hour of operations. It would still need a pretty substantial source of power for re-charging between walks.
They are well in advance of what NASA has and they are not looking back. NASA has choked.
At 11 kW for the vehicle, safely say, “never mind” about the instruments. An entire flagship mission’s science doesn’t draw more than a few hundred watts.
3.7 kW-hours per hour of operation might not be so bad. 10 m^2 of solar array could probably support a few hours of operation per day. You might even be able to boot-strap it: Could this Atlas robot set up a 10 m^2 array in an hour or two? If so, you could land it fully charges, a spare battery pack, and the unassembled power station.
More power would likely require a reactor. While this would require $1-2 billion to design and test, additional units duplicating the design would be less expensive to build, process and launch than an RTG of similar power as it would use uranium, which is not biologically toxic, for fuel rather than plutonium, which is extremely toxic.
A high performance robot also needs much greater computing power than any space probe or rover to date. This requires development and testing of high performance radiation tolerant computers. So far the company working hardest on this is SpaceX with its 4×2 redundant flight computer.
I know metalic plutonium is pretty nasty, but that isn’t relevant. RTG’s use plutonium dioxide, which isn’t water soluble, and plutonium in general isn’t readily absorbed by the body. There are much nastier radioactives around. But I agree: A reactor, if you launched it cold, would be less of a radiation concern. As soon as you turn it on, it would start producing daughter products which would probably be worse than plutonium dioxide. Launching it cold means launching it untested.
In any case, an RTG using a Sterling engine (or other heat engine) instead of thermocouples would probably give you a factor of five or more increase in power output (and specific output, since this needs to be a portable system.)
In some sense every rocket ever launched was “untested”- test on earth, send a copy to space. The same could happen for the pile you imagine.
I remember when the crazys came out for launch of Galileo, though only at first; when it was finally launched things had calmed down a bit. Perhaps launching a nuclear device now would not engender much opposition.
Just look at how many RTGs and RHUs have been launched since Galileo without enough protest for them to replace Galileo as your go-to example of Teh Crazies.
Who protested MSL? New Hoz?
There were protests at the Cassini launch and over the Earth flyby. However, at the launch, the Planetary Society organized a counter-protest. I’ve been told they had more people turn out than those opposed to the launch.
However, that was all before Fukushima (or only a few months after, in the case of MSL.) I’m not sure what the public reaction would be like today.
On the other hand, we could take one of Mr. Spencer’s favorite ideas to its logical conclusion. Why put the fuel in the reactor and incur whatever risk launching it entails? There is plenty of uranium beyond Earth orbit.
Well, yes, all rockets (to date) have been somewhat untested prior to launch. But that’s exactly what companies like SpaceX are trying to change. Shifting to a power system that can’t be tested before use isn’t what I’d call a step in the right direction. But there are ways around that. You could launch to a parking orbit, or one near a space station, test out the reactor, and be in a position to fix it (if something doesn’t work.)
Honestly, that’s close to one of the things I never understood about the Space Shuttle and the Galileo mission. Why didn’t they open the antenna while the spacecraft was still in the Shuttle’s cargo bay? If that had been done, the failure to open correctly could have been fixed. Why launch it, and then open the antenna when it was too late for anyone to kick it and unstick a pin?
I’ve wondered the same thing about the high gain dish.
>>Launching it cold means launching it untested.
Previous studies have suggested a so-called zero-length criticality would be adequate. The complicated parts can be tested without actual fission,
>>plutonium in general isn’t readily absorbed by the body
It doesn’t take much, and it does not have to be absorbed. A single 5 micron respirable particle can induce lung cancer.
I’m not saying plutonium is a fun thing to toss around. But you can make similar statements about many things. For example, you don’t really want to inhale 1 to 5 micron particles of powdered (non-radioactive) beryllium.
Maybe I’m overreacting, but I complain about comments which are along the lines of the “plutonium is the most deadly substance on Earth” rhetoric. That’s been commonly used by the people who protest against using RTGs at all.
On one occasion, Edward Teller (not necessarily the sanest person in history) once offered to swallow a teaspoon of PuO2 (238) if a person who had just made such a “deadliest substance on Earth” statement would swallow a teaspoon of arsenic. The person in question declined to do so.
Gee, everyone talks about ISRU to make methane and oxygen for rocket fuel. Here on Earth we call that Natural gas and use it for all kinds of engines ….
Carrying around cryo-Methalox at the appropriate ratio is a little different than sticking a gasoline tank on there.
Only if you need high power for a short time. Combustion engines are good at that. For lower power over a longer time (i.e. a flatter power profile), why bother? The ISRU plant would be run using solar power. What’s the point of converting electric power (from solar) to chemical, only to burn the fuel and convert it back to electrical power to operate a robot?
There are various approaches but until we have dozens or more likely hundreds of people on Mars ISRU doesn’t seem practical to me because of a wide range of logistical and support requirements that are seldom mentioned in the rather optimistic studies.
There are simpler strategies that could work on Mars. maybe a stationary array of solar panels, high capacity batteries with the panels, a self-driving vehicle, also with batteries, and an automated charging system to transfer power to the vehicle. Wait a minute, those are (or soon will be) all Tesla products.
Apples and oranges, maybe? Are the Boston Dynamics products designed for hard-radiation environments?
Is NASA’s R5 designed for hard-radiation environments? No.
The problem as I see it with NASA in-house vs the like of Boston Dynamics, Space-X etc, is that NASA’s in-house activities have become become so bogged down with excessively bloated management and design review practices and self serving political interference that what come out at the end is a kludge of several years old technology often chosen to appease various political and management sectors instead of state of the art technology chosen for it’s ability to do the job in the most efficient and effective way.
That covers a lot of what they are doing these days.
That is because they are trying to run an R&D program like a flight program.
Yes indeed, and the biggest problem is that’s not even how to run a proper flight program in today’s word.
This is why I don’t understand why NASA, who stared the COTS program, with so far great success. Why not reform and unveil a new Robotic Moon mission but with cooperation from the private sector such as Boston Dynamics, possibly even Google’s X Labs. The R&D rights alone would be more than lucrative from a business standpoint.
That wouldn’t even need true robotics. The moon is close enough for direct teleoperation from Earth.
[Aside: I’ve suggested before a cheap NASA contest where teams would control rovers, racing over simulated lunar terrain, behind a 2.7s signal lag. Look for methods of optimising control (such as locally predicting response, as used in video games to deal with variable network lag). Different classes for schools, unis, self-funded hobbyists, for wheeled vs legged, etc.]
I’ve thought of the same thing, but suggesting that it could be taken further. I think a lunar rover race could tap into college athletics funding. A Sojourner or Yutu class rover isn’t much bigger than a CubeSat. I could imagine development, transport to the lunar surface and operations for under $25 million. Do you think there are enough rich alumni to cover that, for their university’s entry in the First Lunar Grand Prix?
For a Mars mission (well, pretty much for any mission) a bipedal robot would be stupid. A quadruped, like some of BD’s other robots (Big Dog/LS3, Spot, etc), would be much better at dealing with terrain issues.
(They would also be able to carry heavier power systems, or Keith’s suggestion of fuel tanks, or even just solar panels on their backs splayed out like a Pegasus’ wings.)
I honestly don’t see any justification for the obsession JSC has with humanoid robots, even if they were any good at it.
(Something like JPL’s RoboSimian makes sense for space-station ops, the spider-like ability should work well in micro-g. But JSC’s Robonaut is a stupid design.)
I don’t know this for a fact but perhaps the JPL efforts are simply research?
I’m not sure a walking robot is the best idea. Either rolling or walking rovers will encounter some obstacles they cannot traverse. But even over fairly rough terrain a wheeled chassis is faster and more energy efficient than a walker and has greater passive stability. Take a look at the DARPA off-road autonomous vehicle race; imaging systems have made a tremendous difference but the winners are wheeled vehicles.
In weightless environments neither wheels nor legs are needed and three or four hands and some reaction jets might be best.
NASA is currently building a very similar rover, allegedly identical except for carrying a different set of instruments, for launch in 2020. So the state of the art for NASA Mars rovers hasn’t changed much in the past decade. (And, before someone points out that the used the same design to save cost, I’ll point out that this is keeping the cost “down” to $1.8 billion. I might be interesting to see how that compares to the income of this Boston Dynamics company.)
Keith isn’t criticising rovers, he’s criticising JSC’s Robonaut.
To be fair, the mainstream anti-nuclear people do also protest against the oil and gas (and coal) industry. Against off-shore drilling, against fracking, against waste-dams, etc. And most of them don’t protest against the launch of space missions with RTGs.
Yes, I totally agree that Boston Dynamics attracts the best researchers in humanoid robotics, however NASA as supposedly one of the most preeminent agencies for space exploration and associated technologies in the world should also be attracting the very best in the many fields in which they work, if they are not attracting (and keeping) the best then something is clearly very wrong.
I would like to see what that Robot Looks like after 10 years on Mars.
They look cool, but only after an test in a desert for a few weeks I would trust in them.