Forgot your password?
typodupeerror
Android Hardware Hacking NASA Open Source Space Build Linux

Can Android Revolutionize Spacecraft Design? 110

Posted by timothy
from the but-wait-for-the-monthly-bill dept.
An anonymous reader writes "NASA's Ames Research Center is working on a new project designed to drastically cut the cost of launching and operating small satellites in Low Earth Orbit (LEO). The project, known as PhoneSat, will see the Android powered Nexus One and Nexus S phones command their very own small scale spacecraft this year in a first of its kind research mission."
This discussion has been archived. No new comments can be posted.

Can Android Revolutionize Spacecraft Design?

Comments Filter:
  • Or are they? (Score:1, Flamebait)

    by EGSonikku (519478)

    We're the Nexus One and One S one of the devices found infringing in the Samsung V Apple trial? Apple is having an injunction hearing soon to get the infringing devices taken off shelves.

    Not trolling, just curious. Wouldn't be an issue though I think, as there are no shortage of cheap non-infringing Android phones they could use instead.

    • by Anonymous Coward on Saturday August 25, 2012 @05:02PM (#41124823)

      We're the Nexus One and One S

      Wow. Take that, Siri. Nexus One and One S have advanced to the point where they can troll Slashdot.

    • The Nexus One was made by HTC, so I doubt Samsung got into any trouble for it.
    • Re:Or are they? (Score:5, Insightful)

      by MacDork (560499) on Saturday August 25, 2012 @08:18PM (#41125909) Journal

      They don't have to buy them in the US. Other markets will still have access to these smartphones. In the US, your choices will be limited by decisions made in kangaroo courts.

      Not that you have much choice in the US anyway. Notice how the rest of the world had quad core phones back in April... Maybe Apple will "invent" quad core phones later this year. I wouldn't hold my breath on it though. They tend to deliver old hardware that's at least a year or two out of date on launch.

      • They don't have to buy them in the US. Other markets will still have access to these smartphones.

        For the moment maybe.

        But Apple is basically following the well trodden path established by the RIAA/MPAA lawyers. They'll use the USA and its government as a base to extend their influence worldwide.

        While we probably won't see Kim Dotcom style takedowns of manufacturers in the next few years, we will see raids of black-market phone resellers in countries who've been coerced into whatever the SOPA equivalent will be called.

    • by Type44Q (1233630)

      Wouldn't be an issue though I think, as there are no shortage of cheap non-infringing Android phones they could use instead.

      It's a non-issue regardless; the recent court decision in Apple's favor won't apply, due to jurisdiction. :p

  • by Anonymous Coward on Saturday August 25, 2012 @04:38PM (#41124655)

    Perhaps try Dr Chris Bridges at Surrey Space Centre in the UK?
    http://www.sstl.co.uk/divisions/earth-observation-science/science-missions/strand-nanosatellite

  • No. (Score:2, Insightful)

    by Anonymous Coward

    No.

    • Yes.

    • Yes if it's plural "androids". We need truly autonomous robotic space vehicles and probes with the intelligence and mobility of an Apollo-era astronaut. ATM our space probes, already remarkable pieces of hardware, don't have the capability to fix themselves, much less build newer copies. The Martians rovers move at a pace that would shame a snail and they take such unimaginative pictures that, had they been taken on Earth, wouldn't merit a second glance on Facebook. As for the satellites above our heads, ma

    • by Anonymous Coward

      What idiots/macfags have modded this insightful?

      Android is establishing itself as an exceptional option for controlling UAVs and other autonomous vehicles.

      http://developer.android.com/tools/adk/index.html [android.com]

      http://code.google.com/p/androuav/ [google.com]

      http://code.google.com/p/aeroquad/ [google.com]

    • by mcgrew (92797) *

      Wow, such horrible moderations today, why do some days seem to be "give points to morons" day? An AC says "No" to the question, and nothing more; a completely uneducated assumption that flies in the face of what TFA says and gets an "insightful"? WTF? Whoever modded this, please undo your idiotic moderation by commenting. Lacking that, someone with both mod points and a brain mod it "overrated". I have yet to see a single comment here today that was more overrated than that one. It would even be overrated a

  • Really? (Score:4, Insightful)

    by fuzzyfuzzyfungus (1223518) on Saturday August 25, 2012 @04:46PM (#41124709) Journal

    Please, do correct me if I'm wrong; but I was under the impression that the overwhelming majority of the cost of doing space work was in launching the things, with the relatively high salary and R&D costs of building sophisticated precision instruments in very short runs.

    Is the cost of computing anywhere near that significant(especially in situations where you are willing to skip serious rad-hard gear), to the point where you would be better off using a commodity phone(with screen, consumer-pocket-resistant chassis, more GPU than you need for Quake3, etc.) rather than a slightly more expensive, but by no means all that esoteric, ARM SoC board designed for embedded applications? In the same vein, is there an advantage to using an Android environment(whose virtues lie primarily in UI and 3rd party applications) rather than a standard embedded linux or other OS?

    • Re:Really? (Score:5, Informative)

      by icebike (68054) * on Saturday August 25, 2012 @05:13PM (#41124887)

      Please, do correct me if I'm wrong; but I was under the impression that the overwhelming majority of the cost of doing space work was in launching the things, ...In the same vein, is there an advantage to using an Android environment(whose virtues lie primarily in UI and 3rd party applications) rather than a standard embedded linux or other OS?

      Android IS Linux. Its just that it has already been trimmed down to the bones and runs on very powerful, but low energy consuming hardware. So it saves them the need of making a linux version to do the same thing.

      As for your other points, it seems from TFA that they will not actually be using the phones as we know them, but rather stripped down models sans screen, case, battery, etc. Probably wrapped in something to compensate for lack of radiation hardened componentry. But then again, the types of short duration, low orbit, small payload missions they are planning this for can probably risk the radiation for the duration.

      In the case of our intrepid Nexus phones, the issue is being tactfully ignored. As these are test missions, NASA isn’t concerned about the long-term viability of these craft, and only expects them to last a few weeks or months. Due to their low orbits and lack of thrusters to increase their altitude during the mission, the Nexus-powered satellites will be falling back down to Earth within a year anyway. Even if they were built to better withstand the extremes of space, they would still just burn up in the atmosphere before too long.

      I doubt they launch these things singly, they probably piggyback on other launches, but being the size of a coffee cup, they might be able to use cheap sounding rockets [nasa.gov]. There are a number of different models [nasa.gov] of these relatively cheap rockets that can launch low earth orbit small payloads.

      • by gbjbaanb (229885)

        but still, I've found alot of Linux systems very reliable, never-reboot type things, but Android?! I love my Galaxy S but the number of times I've had to pull the battery to reboot the thing because its unresponsive is just too much. Choose a different distro!

        • by Pseudonym (62607)

          I haven't had it for long, but my Galaxy S III has only ever hung when the UI shell hung. Oddly enough, the cause of most crashes so far has been Mongloid.

          • by gbjbaanb (229885)

            It hung on me after a couple of months, but its getting a lot worse recently - particularly after I upgraded to the ICS update, now I have less free memory and I think that this is the problem - Java apps either swapping (as it can be dreadfully slow too at times) or just not realising its run out of ram and throwing a wobbly.

    • by Anonymous Coward

      Generally speaking yes, hardware is a very significant part of the cost. While I don't have a breakdown of how much is salary vs instrumentation vs bus vs computing and so on I can say that the price of the things we are launching, especially for scientific missions, is massively higher than the launch costs.

    • Re:Really? (Score:5, Informative)

      by evilviper (135110) on Saturday August 25, 2012 @06:08PM (#41125243) Journal

      Please, do correct me if I'm wrong; but I was under the impression that the overwhelming majority of the cost of doing space work was in launching the things

      Nope. A quick search shows that cubesats cost $40k to launch, and developing a cubesat reportedly runs from $25-50k, easily a significant fraction of launch price.

          http://www.space.com/308-cubesats-tiny-spacecraft-huge-payoffs.html [space.com]

      Is the cost of computing anywhere near that significant [...] rather than a slightly more expensive, but by no means all that esoteric, ARM SoC board designed for embedded applications?

      You really should read TFA. Cell phones are perfect because they include a compass, gyro, camera, etc. A LOT more than just an ARM SoC. Hell, they can probably sell the screens on eBay and make back a significant portion of their purchase price.

      Thought TFA didn't say so, the power management in Android phones is probably better than what you'll get anywhere else... Standby and talk time are major advertised features, so manufacturers make sure it's working as well as possible. And with a satellite, electrical limits are a major issue.

      In the same vein, is there an advantage to using an Android environment(whose virtues lie primarily in UI and 3rd party applications) rather than a standard embedded linux or other OS?

      Android is a standard embedded Linux OS... It's basically just got a custom UI instead of X11. From the command-line, you wouldn't know the difference. Many people install Debian on their Android phones...

      • Re:Really? (Score:4, Insightful)

        by ceoyoyo (59147) on Saturday August 25, 2012 @06:39PM (#41125425)

        Still seems like a gimmick. You can drop by an off the shelf electronics supplier and get a processor, gyro and accelerometer for a fraction of the cost of a cell phone. Then put something on it smaller and more reliable than Android.

        • by AmiMoJo (196126)

          More reliable than hardware that has been extensively tested by tens of millions of users over a period of years?

          Phones get a lot of expensive EM and RF testing too, as well as very detailed power analysis. You could spend tens of thousands of dollars doing that, or you could just use some highly reliable off-the-shelf hardware someone else has already spent that money on.

          Plus Android is a well known and understood platform that is easy to develop for, providing standard APIs for things like gyroscopes whic

          • by Microlith (54737)

            Plus Android is a well known and understood platform that is easy to develop for, providing standard APIs for things like gyroscopes which raw Linux does not.

            Please. Sensors like that are exported via /sys nodes. If accessing those scares you then you probably shouldn't be writing software for satellites. Android adds nothing except a layer of complexity.

          • by ceoyoyo (59147)

            Has your phone ever crashed and required a reboot? That's going to be a big oopsie for a satellite. Phones certainly haven't been tested in the conditions faced by a satellite. RF testing? For the cell phone component that's going to be completely useless (and hopefully you can turn off completely to save power)?

            Phones are engineered to work in a fairly coddled environment, for a fairly short period, with a (catastrophic) failure rate dictated by how expensive it is to replace them under warranty and a

          • by ceoyoyo (59147)

            PS: I read the article. I'm disagreeing with it.

            Gyroscope APIs? If you can't figure out the datasheet for a a solid state gyro you shouldn't be designing satellites.

      • by Timmmm (636430)

        developing a cubesat reportedly runs from $25-50k

        Yeah in engineers salaries, solar panels, machining work, testing, etc. The cost of the computer is only a part of that.

        Cell phones are perfect because they include a compass, gyro, camera, etc. A LOT more than just an ARM SoC.

        The hardware in a phone doesn't cost anywhere near $25k. In fact buying the parts separately is probably cheaper than most phones (because you can miss out the screen, 3G, GPS, wifi, etc. on a satellite). Plus if you build it yourself you will save weight and have more flexibility. Something like gumstix combined with and one of invensense's new IMU chips would be much better than an Andro

        • by evilviper (135110)

          The hardware in a phone doesn't cost anywhere near $25k.

          You're borderline illiterate if you think I said or implied anything like that.

      • by Microlith (54737)

        Android is a standard embedded Linux OS...

        Compared to what's out there, it's rather non-standard. It doesn't use uClibc or glibc, it's got a custom IPC method nothing else uses, a wonky filesystem layout, etc.

        It's basically just got a custom UI instead of X11.

        There are many, many more differences.

        From the command-line, you wouldn't know the difference

        Yes you would. I've had to wander around the Android filesystem and wonder where the hell they put things. But that's what happens when a platform is built fro

      • Well, a CubeSat might cost $40k to launch, if you can find a ride-along as a secondary payload. More often, it costs the user nothing, since someone like NASA or NRO picks up the tab. The problem is finding a launcher that has space available. Many satellites sit on the shelf for years.
    • The cost is often 10 to 20% of the cost of the mission, but it has a lot to do with the cost of everything else. A fairly large launch vehicle can cost in the $100M range, so it makes it important that whatever you're putting on it is going to work, and it's worth spending a lot of money to do that. Think of it in terms of domestic travel-- if it cost $1M to fly from LA to New York, you'd probably spend more than that making sure you had the whole trip planned out in incredible detail. With it costing on

    • Please, do correct me if I'm wrong; but I was under the impression that the overwhelming majority of the cost of doing space work was in launching the things

      You're wrong. Launch costs rarely, if ever dominate.

    • that significant(especially in

      fix

      commodity phone(with screen

      your

      Android environment(whose virtues

      space key

  • ... will see the Android powered Nexus One and Nexus S phones command their very own small scale spacecraft ...

    For how long? These are consumer devices. The hardware and software are not flight rated and not radiation hardened.

    That said its a really cool hack but hardly something that will radicalize design. Its not like the space program wasn't already on the path of smaller, lighter and less power consuming electronics. Our modern computers and devices are a direct result of space research.

  • Many Android users have to remove the battery now and then to restart their phones when an offending application completely freezes Android. IMHO, if you need to run a custom (and only) app, it is not worth the hassle. As for the cost of the hardware, there are many cheap SBCs that could do the job running an OS more fit for the job, like Linux or any other free OS, maybe with real-time scheduling and proper GPIO to wire-up a satellite.

  • by Anonymous Coward on Saturday August 25, 2012 @04:51PM (#41124751)

    The spacecraft's battery will die half way through its rated time, a simple grid UI will lag for its input, your spacecraft will get hijacked by malware from the NASA-Store, each spacecraft will have a different version driving costs up, and software updates in space will be at the mercy of At&t and Verizon?!

    • mod parent up, made I chuckle.

      There's a reason NASA use purpose-built, radiation-hardened hardware in flight systems, a reason why they use in-house and real time flight system software, and a reason why they use older *proven* technology on the ground. It's because they don't want to have to sit there worrying about some rogue bit of software taking the kernel down at any point in the data chain, because mitigating against that after the fact, if even possible, would be obscenely expensive not only in term

      • by icebike (68054) * on Saturday August 25, 2012 @06:12PM (#41125269)

        Commodity mobile phones aren't safe for flight science systems, end of story, because they're a: not hardened, b: not real time, c: not reliable and d: vulnerable to script kiddy software hacks.

        Really?
        If all your assertions are true, then explain why is it that this program is being done at NASA’s Ames Research Center where they employ real world class ROCKET SCIENTISTS, rather than kids who play them on slashdot?

        Script kiddies in space? Really? You are going with that?
        Not reliable? Did you even read TFA? Its a coffee cup sized payload, in low earth orbit, expected to re-enter in less than a year. How reliable does it have to be?
        Real Time? Seriously? Its monitoring a sensor package, and radioing it back to earth. It has no thrusters to control. Just how much real time is needed?
        Hardended? In low earth orbit (WAY lower than the ISS) protected by the Van Allen belts. Costing maybe a thousand dollars to construct. Who cares if it takes a direct hit? They are cheap enough to send up two dozen on a single sounding rocket.

        At least READ TFA before pontificating on things you obviously know nothing about.

        • NASA employs great people and okay people, just like other places. They're not magical.

          • by icebike (68054) *

            But ALL of the employees given big budgets to do this type of thing are significantly more magical than your average self appointed expert posting on slashdot under the name of Tastecicles.

        • saywhatnow?? This is SLASHDOT, we don't read the fucking articles! You're lucky I skimmed the fucking summary.

        • At least READ TFA before pontificating on things you obviously know nothing about.

          You must be new here.

          (I have never said that tired meme before... but someone had to in this instance. Yes, I get the ironing(sic) of a 6 digit UID telling a 5 digit UID that meme. Shall we persue a rediculous ... never mind.)

  • All I can say is these devices are not built to be radiation tolerant to say nothing of radiation hardened. Keep in mind the laptops (not used for safety critical things) on the ISS have to rebooted daily because of Single Event Upset (SEU)s that lock them up.

    • by X0563511 (793323)

      Please tell me those things at least have ECC.

      • by gbjbaanb (229885)

        of course they have ECC, but they run Windows 95 because that was the only OS available at the time NASA started the ISS OS-validation program, :)

        • by wvmarle (1070040)

          They run Win95, have to reboot once a day, and blame it on some external factor?

          For starters they should be happy they need to reboot just once a day!

  • There's an app for that.

  • I guess someone's already building prettier iRocket. With more RAM and way more intuitive interface for space inhabitants.

  • Man, what can't these amazing phones do?! I mean, steampunk carrot ship designs are a dime a dozen, but Revolutionary War themed Spacecraft? Hell yeah!

    Unfurl the solar sails and set a course for the Kaiser Sea of Mars!

  • Neil Armstrong is probably spinning in his grave...

    What, too soon?

  • by Compaqt (1758360) on Saturday August 25, 2012 @05:15PM (#41124899) Homepage

    1. They use touchpads in 2001.

    2. They use PADDs in Star Trek.

    3. Apple copies from #1 and #2.

    4. Android is used to build real versions of #1 and #2.

    5. Apple sues #4.

    6. Profit!

  • Phonesat 1 will post Instagrammed photos of Earth and tweets about how lonely it is. Phonesat 2 will retweet them.
  • I'd say, judging by the number of MacBooks at the NASA Curiosity control center, that a more appropriate question would be to replace "Android" with "MacOS".
  • If they're using Android, why not a RaspberryPi?
  • by bmo (77928)

    Oh, look, it's another person suggesting that consumer level electronics be used for space-rated applications and saying it would be cheaper. Despite all those labels and warnings saying "this won't work below freezing or significantly above 50C.

    You send certified space rated electronics up because once you send it up on the rocket, you're not going to go up and repair it if it doesn't work. If it doesn't work, you're out the full amount of money you spent, plus you have a hunk of junk orbiting the Earth

    • oh look, another person who can't read:

      As most people know, space is a pretty nasty place to hang out. Between the monstrous temperature extremes, crippling radiation, and risk of impact with all manners of space junk and micro-meteorites, it’s hard enough for a full scale spacecraft to survive, let alone a smartphone. While these risks can be mitigated by adding shielding to separate the internal electronics from the hellish environment outside, this unfortunately adds considerable weight to the vehicle. Added weight means it takes more energy to put the craft into orbit, which in turn means higher costs.

      In the case of our intrepid Nexus phones, the issue is being tactfully ignored. As these are test missions, NASA isn’t concerned about the long-term viability of these craft, and only expects them to last a few weeks or months. Due to their low orbits and lack of thrusters to increase their altitude during the mission, the Nexus-powered satellites will be falling back down to Earth within a year anyway. Even if they were built to better withstand the extremes of space, they would still just burn up in the atmosphere before too long.

      Assuming the success of this upcoming mission, future smartphone based spacecraft will likely be designed for longer duration missions, with better shielding and redundant systems. But first we need to see if it will even work...

  • I'd hate to see someone have to trust their life in an OS that attracts malware like Android does.
  • We wouldn't give a stupid Apple fanboy wankfest a pass if these phones were running iOS, we shouldn't give this a pass because these phones run Android. Quite disappointed to see this on Slashdot.

  • It should be one with unlimited data, and good coverage, at least in outlying areas.

  • Q) How many Androids does it take to fly a satellite?

    A) Two. One to fly the satellite, and the other to power the robotic finger that dismisses the "Battery Full" notification and relaunch the flight control app whenever the solar panels are temporarily obscured by a shadow.

    Not really a joke, but based on my experience using an Android phone as a 24x7 webcam. Every time there is a power glitch, the "Battery Full" notice takes over the screen and foobars the IP Webcam app. The OS just wasn't designed for sta

    • There's a watchdog system to reboot the phone in case of a glitch. Not really a problem. Most CubeSats operate the same way, since they don't use rad-hard electronics due to the cost and limited performance.

"Indecision is the basis of flexibility" -- button at a Science Fiction convention.

Working...