NASA Testing Linux-Based Exploration Robots

SeenOnSlash.com writes "This week NASA is testing a Linux-based lunar rover called K-10 in the Arizona desert. To cut costs and promote maintainability the K-10 runs Linux and uses commercial off-the-shelf parts where possible. The robot rover's control and communications system is based on an IBM Thinkpad X31 and attaches to subsystems with standard PC interfaces. Real-time tasks such as fine-grained motor control are offloaded to a distributed network of microcontroller-powered control boards. Maneuvers can be watched through a live webcam."

Comments

Victory!

[Anonymous Coward]

Once again, Linux proves itself as the OS of choice whenever you couldn't care less about the OS.

The robots are driven by custom robot-specific software that has nothing to do with the OS underneath. The main reason Linux gets used in such an application (or in supercomputers, clusters, etc) is simply that the OS doesn't matter enough for anyone to bother, so they'll grab the nearest thing on the shelf.

It's not like there's some feature of the OS that makes it especially robot-friendly.

Re:A great Contest

[KokorHekkus]

The use of Linux and commercial grade gear for the space program is really quite cool stuff. It makes me think that a really cool contest for NASA would be to have grad students desgn and build a rover/probe and the winner (once vetted by NASA) is actually launched into space. It is probably cost prohibitive but it would be very cool if it happened. It may be a way to break NASA groupthink, and re0invigorate designs with some fresh minds. Not that I'm critisizing NASAs robotics programs, the Mars rovers are a smashing success.

I suspect that the idea here is to cut down development costs, not the build costs for the final implementation. Commercial grade gear just wouldn't cut in space with the extrem temperature swings, cosmic radiation messing with memory or if you end up on Venus god knows what kind of weird atmosphere.

But someone could actually build a proof of concept design (like they're doing now) that could become the basis for the actual rover. I can't imagine it takes that much more work using Linux than some other solution since there's bound to be a lot of non-standard stuff going on with the rover software.

Of course there is

[Mateo_LeFou]

Namely: it can be dismantled, optimized and modified to your heart's content. Which is a lot, when every millimeter of length and every gram of mass has to be accounted for.

Higher Quality vs days of Yore

[johnBurkey]

The days of NASA designing their own systems from scratch using proprietary stuff should be just about over. While in days of yore, there really wasn't much in the way of "off the shelf" parts, and NASA (had to/ could get a way with) large R&D budgets for designing tech gear, at this point its better they focused their thing (exploration/safer rockets/advanced propulsion) and let the tech community do the computer hardware/software design. Its a function of ROI- Apple/MS/Intel/IBM/Moto/Linux etc. have spent billions, and more importantly, thousands of man hours perfecting stuff- I would rather NASA spent its man hours doing user applications, device drivers, etc., than designing a new computing platform whose relative youth could cause a mishap. That simple user app from Linux's view point, protected from doing harm to the OS or itself by micro kernels, memory protection, etc., might actually be a rover AI. That simple device driver might be a USB rover mandible. Its all about fighting the common foe of all engineers : Complexity. There is enough of that in NASA's domain to keep them busy for a long time :)