Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
Space

Space Shuttle Displays Go Glass 164

Posted by timothy from the I-need-this-in-my-'95-Escort dept.
cloudscout writes: "NASA has finally decided to bring the space shuttle up to date with a 'glass cockpit.' Until now, the space shuttle cockpit has used a system of gauges and dials designed in the early 70's. They now have full-color computer displays and controls. Pictures and details are available in this article at WESH Channel 2000. So how long until someone ports MAME to this thing?" Can anyone shed light on what sort of operating system will drive all those screens?
This discussion has been archived. No new comments can be posted.

Space Shuttle Displays Go Glass More

Space Shuttle Displays Go Glass

Comments Filter:
  • what they really need is plexiglass cockpits...ahhh. imagine it. mmmmm
  • Glass screens, full colour, brand new... it's obviously Windows CE. Or Pocket Windows. Or whatever it's called this week.

    NASA Administrator: "Sure, we want a calendar and a phone list in Atlantis -- but what about videos and MP3s??"

    NASA Worker: "Uhh, I'll get right on that."

    NASA Administrator: "Can you put in that paperclip thing too?"
    - Cam MacLeod

  • this was in the same news story as the Shuttle going up to tow the ISS. The "glass cockpit" is being live tested on the mission. We gotta have a way to moderate stories as well as comments.

  • Why retrofit these things? (Score:4)

    by Grant Elliott ( 132633 ) on Saturday April 22, 2000 @03:38PM (#1116286)
    First of all, the shuttles are supposed to be replaced within a few years (hopefully). Secondly, I can think of a lot of other things that should be updated first. Anyone know how the computer system onboard works?

    There are five computers- with tape drives! The tapes aren't big enough to hold everything, so there are seperate tapes for take-off, landing, orbit, etc. The tapes need to be changed by hand on the first four machines. The fifth machine is permanently running the emergency landing routine. If one of the first four computers disagrees with the others, it is shut down. If the shuttle is ever down to two computers, the fifth machine kicks in and takes the next available landing window- with or without anyone onboard. It's never happened.

    I always assumed the reason that the system never got updated was because they were planning on replacing the shuttles soon. Makes you wonder how confident NASA is about the X-33.

    (This information is a few years old. Anyone know if NASA has upgraded the computers yet?)

  • The Display (Score:2)

    by Anonymous Coward
    I think all the shuttle displays are done by a company in Albany; they use a Photon drawstream without QNX's rendering libraries. I don't think that most of the companies producing these technologies are allowed to be named due to their contract with government.

    The shuttle requirements are high; they need a mission critical OS with real time performance for *many* of the systems. Good for QNX, bad for M$. :)

    It's a case of both security through obscurity and the government not wanting to be known for publically endorsing one vendor over another. (What would be the impact if we heard that the government started using linux or microsoft over QNX?)

  • More cost cutting (Score:4)

    by djrogers ( 153854 ) on Saturday April 22, 2000 @03:41PM (#1116288)
    NASA refused to comment on the OS question, however one source has told me that they got a $7,600 savings thanks to a $400 per unit rebate from MSN.....
  • Hmm. If NASA installed a Win32 OS in the Atlantis shuttle, chances are, the astronauts on this upcoming mission will be the first to see the Blue Screen of Death (TM) from space. Though, they also might die in the process. HeH!
  • Must be linux, because in space, no one can hear you scream.. when you blue screen :o)

    Segfault

    segfault@bellatlantic.net [mailto]
  • So they've used this old stuff for 19 years now. The technology has been throughly tested, NO WAY is it possible that there would be bug somewhere. Now I'm sure the engineers and programmers at NASA know their shit, but nevertheless there is risk (no matter how small) that this new system has some flaw in it.

    Does this really bring so much new funtionality that the risk is worth taking?

  • So do the shuttles still have iron core memory? I remember that this choice was 1) made in the 1970s and 2) core memory is remarkably resistant to EM radiation, gamma rays, etc.

    I also remember that the shuttle was apparently running on Z80 clones. Forget WinCE, they're still on CPM.... (which is more advanced that what runs under Win98).

    Karen
  • Everything NASA does is done in-house.that's why their budget is 12 some billion and they manage so many shuttle launches a year. Like the time they built their own video camera for something like 180k for the shuttle and it didn't work and one of the crew hooked up a sony and it was amazing. This probably cost a half mil and could have been bought at radio shack and set up by a kid like Jason Fox for 1% of that.Typical.
  • I just looked at the picture. They must have done some serious retrofiting. I seriously doubt they could run all that off the old computers. This probably cost a pretty penny...
  • Possibly not. Previous shuttle missions have had Thinkpads in use, and for some reason I think even a powerbook has gone on at least one mission.

    Can't believe they've had laptops up there without experiencing that beautiful blue screen at least once... :)
  • Comment removed based on user account deletion
  • For better or for worse, they aren't getting rid of them anytime soon. I can't back this up with links, but I know I've read at least one article recently that mentioned new engines and fuel tanks. And I know the frames were built to last 30-40 years, which would mean 2010-2020.
    ~luge

  • Hmmm... (Score:3)

    by Ravagin ( 100668 ) on Saturday April 22, 2000 @03:56PM (#1116298)
    I think it's unlikely that Windows would be used. Stability issues aside, for a spaceship, you want the maximum ability for customization. I'm no Linux expert, but it sounds like it would be a good OS to use. And hey, we had that story earlier about building your own distro. That would be neat: NASA writing its own Linux distribution for the space shuttle
    I'm just speculating, mind you. They could use Windows or DOS or something like that...and get all sorts of fun errors.

    Life Support has caused an illegal operation and will be shut down.

    c:\>Close bay doors
    Bad command or file name.
    c:\>Close "bay doors"
    Too many parameters.
    c:\>Close (bay doors)
    Error reading bay doors.
    Abort, Retry, Ignore?

    [astronaut clicks on "Extend Landing Gear"]
    Not enough memory to complete this operation. Close some programs and try again.

    Oh dear....
  • Yeah. It sure would be a terrible shame, right? If they hadn't yet experienced any BSoD aboard a shuttle, they wouldn't be able to suffer and endure the frustration which millions of computer users go through every day, right here on planet Earth.
  • As to what software it's running... it's software version OI-27. (According to the specs) [shuttlepresskit.com]

    Now, I don't know exactly what it means.. but oh well. Maybe they just loaded one of the crays on there.. hehe.
  • I don't remember which platform it was (pretty sure it was Apollo, but maybe something later) that NASA required not just multiple copies of crucial software but multiple copies written completely independently, so that bugs in one copy would be visible when compared against the others.
    ~luge
  • First of all, the shuttles are supposed to be replaced within a few years (hopefully).

    Nope. The video in the link says that the glass cockpit is just one of the updates to make the shuttles last for another 20 years (been about 20 already).

    Maybe they will build a new shuttle in the meantime, but the ones we have now (Atlantis, Columbia...) will stay in service.

    Spyky

  • ShuttleOS: 1 bug in 420,000 lines (Score:5)

    by MikeApp ( 151816 ) on Saturday April 22, 2000 @04:03PM (#1116303)
    Read this article in Fast Company:

    They Write the Right Stuff [fastcompany.com]

    There have been 17 bugs total in the last 11 OS revisions of the Shuttle code, approximately 420,000 lines delivered each time.

  • While installing the software, and monitors, and throughout all the comotion, let's just hope we DON'T USE THE METRIC SYSTEM!!!
  • Actually, what with the joysticks and all, it might be a great system for first person or flight games...Mmmmm, X-Wing on the Space Shuttle...
  • Right. That's what I was referring to when i said, "Though, they might die in the process."

    That would suck for Bill Gates if the astronauts aboard one of these shuttles die because of a computer malfunction. But, then again, he has more money than God. So, it wouldn't matter much, anyway.

  • Re:Shuttle still has core memory? (Score:2)

    by Anonymous Coward
    The Z80 was and still is a workhorse for embedded space systems. The primary reason for this is that it was one of the few cpus that are available in a radiation hardened version. As of the early 90s, the rad6000(used in the mars rover, int space station, many others) from ibm has pretty much taken over that market. This is a first generation rs6000 that is radiation hardened. You can get this puppy at up to a whopping 35mghz. Lockheed Martix produces the development and flight boards for these for the resonable price of $30k and $300k a piece and it generally runs vxworks for the os. In the mid 90s intel introduced an i386 that was hardened and just recently a i486 (think hubble upgrade not so recently). The rad6000 is due for a much needed upgrade to a full blown power pc. It'll be nice to get some documentation for your $300k (i'm sure that'll go up with the upgrade ;) board that has a date that doesnt read 1990 on it ;) The main reason that they antiquated beasts are still in use is primary because compared to newer chips they drawing alot less power, and in the embedded world you can get away with it because you are generally doing 1 specific duty and will never have to compete for the cpu.
  • Hm. It could be Palm OS 3.5...Isn't that what's running on the Palm IIIc?
    ;)

  • I work on military aircraft avionics, which are (theoretically speaking) not really a lot different than space navigation systems. Right now most of our aircraft are still using 60's and 70's technology in the avionics systems.

    Anyway, in a system with a glass cockpit, they will usually have a separate computer for every main function (navigation, guidance, heading, etc), and each computer will run on its own custom hardware. Software usually comes from dozens of banks of ROM chips. Although, if this system is still being designed from scratch, or was recently, then I don't see why they wouldn't use some sort of small RAID arrangement. If it even runs an OS (at least, how *we* define OS), chances are it will be very low level, and custom-coded by either NASA or a high-profile aviation company like Lockheed. Remember, custom (and very expensive) hardwear.

    If they *were* to run a PC-type OS in this glass cockpit, I forsee them choosing either MS-DOS, (hey, it's been proven stable) or some Unix variant or clone known to be very stable (BSD?).

    I'm almost certain they aren't going with Linux because don't you think they would have announced it by now? I mean, Open Source == Publicity these days.

    And to you guys worring about Wince... 1) NASA is not really that stupid and I doubt it could handle shuttle avionics anyway 2) There are always two or three backup systems for every major system. :P

    NOTE: This is all pure speculation which therefore means that it will probably be moderated down as flamebait. :P
  • I'm trying to find some documentation of the following. Somebody (Geeknews, Slashdot...somebody) picked up the story a few months afterwards, but I don't remember who.

    Anyhow, I was watching the NASA channel -- I don't remember the official name -- a few months back. It's cool -- just old tapes of missions, and a bunch of live broadcasts, too. So I'm watching this live broadcast of a recent mission. It's really cool -- you can see a map showing where the shuttle is in relation to the globe, as well as the latitude and longitude, as well as a camera shot from the shuttle. There was some other stuff, too. The screen was all divided up like Bloomberg.

    Oh, and you can hear all of the radio transmissions, too. So I hear a woman on the shuttle say something about a computer problem, and could Mission Control help her out? She describes the problem to MC, and a tech helps out. She reads off the message on the screen, as well as on a printout. It's a BSOD. MC chuckles, and you can hear laughter in the background, and MC says to reboot the computer.

    Now, because I have a wicked poor memory, could somebody validate this?

    Anyhow, the point is this: God help them all if they're upgrading and still running Windows.

    -Waldo
  • Sir, you have given me quite a fright. I originally parsed "ISS" as "IIS". *shudder*...
  • Well I am also of the understanding that all nasa computers that go into space are powered byc 286's. Apparently they're the only things that don't get screwed up by cosmic rays. S If this is the case then I assume hardrives have similar problems. Although this does pose a problem if Slashdot ever aqquires a space shuttle. Commander Taco would be forced to become a minix guru.
  • They're retrofiting these things because they're trying to run a business.

    Retrofiting accomplishes all of the following tasks:
    1. More press coverage.
    2. Provides a back-up if the new vehicles don't work right at first.
    3. Provides additional payload capacity.

    Besides, what are we gonna use after we wreck the X-33s blowing up a giant asteroid on a collision course with Earth.

    --
  • And of course, the obligatory: Wow, when are they gonna run Linux?!?! Let's get a whole bunch of them up there and form a huge Beowolf cluster!!! How do they communicate? BlueTooth 2001!
  • I would say, let's hope they only use the metric system...
  • The station is called NTV (yea I know, it is kinda lame, but they wanted to be cool and all). Currently the main computers on board are not Win computers, neither is the glass cockpit. They do have suplamentary computers and computers to run the experements that are Windows, usually laptops because they are completely unclosed in a nice SMALL package and they do not weigh much. The glass cockpit is really nice, I saw a report on NTV about, really cool stuff.

    They are not dead yet, so whatever they are using must be working.

  • The flat panels kinda remind me of star trek >=TNG type panels... now all they need are touchscreens and a star trek theme, and they could change the name from Atlantis to Enterprise. No warp drive though... gotta wait till 2060s for Zephram Cochran (sp?) (OK, I've been watching first contact too much).
  • The space shuttles are 20 years old, do you have any idea how little code they use? They have less than 100k of RAM if I remember correctly.
  • This page [nasa.gov] has all kinds of crazy info on the original stuff that the Shuttle ran on. It gives a good idea of what it takes to run a shuttle- something like 6 processing units, each with about 400 KB of memory. Lots of stuff in there- details on the custom OS and language which they use, and on the extreme levels of redundancy (software written in two separate locations, 4 copies running in parallel, and an emergency landing program constantly loaded on a separate processor just in case.) Cool stuff...
    ~luge
  • The reason the 286s are in there is because the shuttles were built in the mid-70s. It was cutting edge stuff back then, and they've served perfectly well ever since. There is little reason to spend the millions of dollars to upgrade (since mission control, both in Fla and Tx, must also be upgraded) unless safety becomes an issue. The new glass cockpits are part of that. Besides, the new stuff is 70 lbs lighter, which cuts launch costs by $700,000.

  • Re:Why retrofit these things? (Score:3)

    by Detritus ( 11846 ) on Saturday April 22, 2000 @04:29PM (#1116321) Homepage
    The main computers in the Shuttle are IBM AP-101S computers. The AP-101S is a member of the IBM 360/370 family of computers.

    The problem with hard drives is their fragility. They are too sensitive to vibration and hostile environmental conditions. Tape drives have been successfully flying on spacecraft for decades.

    The Space Station was using radiation hardened 80386 chips the last time I checked.

  • a buggy OS used for mission critical systems...

    Think Not!

    Ever notice the life support clauses in most comercial compilers for non-realtime OS's?
    the usual line in most microsoft end user agreements :

    The SOFTWARE PRODUCT is not fault-tolerant and is not designed, manufactured, or intended for use or resale as on-line control equipment in hazardous environments requiring fail-safe performance, such as in the operation of nuclear facilities, aircraft navigation or communication systems, air traffic control, direct life support machines, or weapons systems, in which the failure of the SOFTWARE PRODUCT could lead directly to death, personal injury, or severe physical or environmental damage.

    Just do a search of the MS site for "life support". That one was ripped from the IE 4 end user agreement [microsoft.com].

    Sorry no wince powered porn sites for NASA...

    TastesLikeHerringFlavoredChicken
  • Those screens looked pretty blue to me... which means it must be windows pulling another one of it's wonderfull faults. Seriously though my bet would be on a unix derivitive (bsd, linux, etc). NASA knows what its doing (well, not really but lets just say they do for the sake of argument) so the chances of using anything M$ is nill.

    But then again, I also think the space program hasn't acomplished anything big enough to justify the billions spent on it. And why does everyone assume that all of us hackers/nerds are into sci-fi (or anything to do with space for that matter)? Ugh.

  • Yeah, and a heckuva lot of people now know Z80 ASM. Why? Because the TI-82,83,83+,85, and 86 graphing calculators -- err, portable monochrome game consoles that you can sometimes use in math class -- have Z80s at their cores. See ticalc.org [ticalc.org] for some info.

  • Isn't there some benefit to having analogue instruments?

    Given radiation and other space born anomolies, don't these upgrades create another occasion for something to go wrong?

    If I recall correctly (IIRC), nearly half episodes on Star Trek: The Next Generation were based on problems that could have been avoided with analogue equipment. So there.


    Soldier(R)

  • This article "They Write the Right Stuff [fastcompany.com]" appeared in the December 1996 issue of Fast Company. It paints an interesting picture of the requirements and culture of NASA's on-board shuttle programming team (circa 1996.)

    The article's main trope pits the discipline and insanely controlled, methodical management of the team and its product against the stereotype of the undisciplined, unruly geeks pounding mountain dews and hammering out bug-riddled code at all hours.

    It's still worth a look, though.

  • The MEDS upgrade includes an Analog/Digital converter which allows the video signals sent to the old CRTs to be displayed on the LCD units. All new computing takes place within the LCD units themselves (for switching between available displays). As far as the old flight computers are concerned, they are still hooked up to the old analog TV tubes that they were designed for.
  • This biggest problem with mission critical code is it needs to be designed arround the belief that every thing is going wrong other wise you end up being unable to handle any problems (aka : the mars polar lander).

    Or am I the only geek that read more than the summary of the reports [nasa.gov] released on the loss...

    TastesLikeHerringFlavoredChicken
  • Sure the tech is old but its been proven to work lots of times. Every new system has bugs that gets past the fault simulators.


    The last thing NASA needs right now is another goof-up. The money involved in deploying systems like this plus the training for the crew must have hurt on their shrinking budget.

  • What OS? (Score:5)

    by darkbabbit ( 172002 ) on Saturday April 22, 2000 @04:47PM (#1116330)
    I don't work on the shuttle, I work in the Avionic biz, which shares many similarities to the shuttle project.

    The avionics biz is very conservative when it comes to items that relate to safety. The primary and secondary displays have to be so safe, bug free, and have such a small memory space that they don't use an operating system such as windows, *nix, or DOS. It is strictly bare metal programming. In the boxes that I work on, having 500k of ram to work with is a luxury item.

    The graphics are usually handled by seperate chip with a dedicated graphics engine embedded into it. The main processor and graphics chip usually communicate via shared memory locations and the commands don't get any more complicated than "draw blue circle at location x,y with radius r" and many of the items come predrawn.

  • Actually, the Shuttle airframes are theoretically rated for 100 flights each. Since the shuttle fleet has just passed the 100th mission mark, the fleet is approximately one quarter of the way through their lifespan.

    The STS was originally intended to be a 'space truck' each orbiter doing maybe 10+ missions a year. When it became clear that was an absolutely unrealistic number, the story was switched that the Shuttle was in fact a test-bed for the technologies that would make the second-generation shuttle the real space-truck.

    The VentureStar, the vehicle the X-33 is a technology demonstrator for, will be built and owned commercially, not by NASA. NASA will purchase cargo-space on the new vehicle, but they will actually be owned by Lockmart (Lockheed-Martin) or whoever else decides to build them.

    The MEDS upgrade is a logical next step in maintaining the shuttle until a replacement launcher is available. It will reduce maintenance costs, increase performance, increase safety, and make the shuttle fleet better vehicles overall until it is retired.

  • Shuttle GPC Operating Systems (Score:5)

    by Detritus ( 11846 ) on Saturday April 22, 2000 @04:52PM (#1116332) Homepage
    The GPCs (General Purpose Computer) in the Shuttle use two software packages. Four of the computers run PASS (Primary Avionic Software System), which was originally written by the IBM Federal Systems Division. The fifth computer runs BFS (Backup Flight System), which was originally written by North American-Rockwell. The "operating system" is unique to the Shuttle, it isn't a port of a commercial product. PASS is the primary system, BFS is there as a backup in case of a common mode software failure during ascent or entry. The Shuttle is a fly-by-wire spacecraft. All of the control surfaces, and many other critical functions, are controlled by the computers. Without an operational computer, you crash and burn. Shuttle software is written in a language called HAL/S (High-Order Aerospace Language Shuttle), which was developed by Intermetrics. The Shuttle's operating system is a hard real-time operating system based on cyclic scheduling. A task is guaranteed to get N cycles of CPU time every X milliseconds. The tasks are managed by three executives, the HFE (High Frequency Executive), MFE (Medium Frequency Executive) and LFE (Low Frequency Executive). A task that issues commands to control surfaces is going to run at a high frequency. A task that checks tire pressure (really!) can run at a low frequency.

  • Re:Why retrofit these things? (Score:1)

    by Anonymous Coward
    There seems to be this misconception that NASA actually produces most of the hardware that goes into space. In reality NASA does next to nothing in house. They subcontract jobs out to Boeing (the entire ISS program), Hamilton Sundstrandt(Space suit/EMU), Lockheed, etc, who in turn subcontracts to hundreds of other companies. Each contract for a program pretty much does his own thing, the result is that you can't make a statement like, all nasa computers that go into space are 286s. If you took a poll today, you would probably see that the rad6000 is the most common chip going into space today and there are probably more Z80s floating around then all the others combined. Bottom line is that for the most part, each sub contractor decides what to put into space. There really aren't all that many choices tho.
  • I believe I heard/read/saw this long ago (don't know why the story would just come out now). Why would they need an operating system though? To play quake during the journey or something?
  • C'mon, this is space. They'll be playing Descent.

  • the weight -- and more links/pics (Score:3)

    by turg ( 19864 ) <turg.winston@org> on Saturday April 22, 2000 @05:01PM (#1116337) Journal
    This CNN story [cnn.com] is more informative and has better pictures.

    The part I found amazing was that, in this conversion, the shuttle cockpit lost 25 pounds of weight. Now, CRT's aren't the lightest thing in the world, so those old dials and guages must have been pretty heavy

    ======
    Webmasters: get a Free Palm Pilot [jackpot.com] for referring 25 signups (Web-based games).

    ========

  • Hidden back doors on Space Stations could ruin your whole day :)
  • Do you think it would be possible for open source code to be implemented in this way: [fastcompany.com] or are these software writing processes already used in OSS? Somehow I doublt that GNU/Linux community could be utilized for such serious work.
    Go ahead, prove me wrong...
  • There was no problem with the mars lander. It was the /. black UFO. Didn't you know, the comspiracy about the /. compound on %^%%^$$#@ connection terminated.
  • NASA Shuttle Avionics System

    One of your applications has violated system integrity.
    Please shut down all systems and restart Shuttle.

    [OK][Help]

    (Original message source: Shoei, RoI, 1994)
  • This is the type of programming I am talking about [fastcompany.com]
    Do you believe it would be possible to have an open source software project that could be successfully and without any problems or risks utilized for a space station?
  • That's DOS, not DoS.

    DOS is stable, and well known. You design these displays with either IR touch or resistive touch.

    The reason for IR or resistive is it can handle gloved hands and the calibration is not effected by temperature or vibartions as much as accoustic or capacitance.

    The reason for DOS, is it may have bugs, but everyone knows the bugs and all the tools for it is well developed and stable.

    Using touch screens allow you to reduce the area needed for controls.

  • they could change the name from Atlantis to Enterprise.

    Why would NASA want a shuttle named 'Enterprise' when they already *have* one?

    NASA's first Shuttle was the Enterprise (OV-101). It was an experimental orbiter, and was used in various atmospheric flight tests in the 1970's. Unfortunately, it was retired before it ever flew in space but it did provide valuable information during it's lifetime. (The first shuttle that made it into space was Columbia, in 1981. Enterprise last flew on October 26, 1977.)

    You can find a really nice picture of the Enterprise here [nasa.gov].

    -- Give him Head? Be a Beacon?

  • and you know this how? Any references?

    I spent years writing telemetry processor software for NASA ground stations, including software to support the Shuttle. You have to have a basic knowledge of Shuttle computer systems to understand how the telemetry is structured.

  • No. Let's make sure we don't use standard AMERICAN measurements. If we do everything in metric at all points in the line, you won't have that problem arise.

  • The reason for installing this new 'glass cockpit' is not merely so that shuttles can look all high-tech and cool. There have been constant complaints from shuttle pilots that the old control systems tended towards information overload. There was too much irrelevant data, and readings that were relevant were not prominent enough on the control panel. The new system will presumably have been designed to address this issue: the most relevant data to a given situation should be presented to the pilots on the display panels (presumably 'relevance' of data will be something which is decided by computer after a lot of simulations and consultations with shuttle pilots), while at the same time if they do need other data it should also be relatively quick and easy for them to access from the panels. Remember, by definition when they built the first shuttle nobody had flown it before. Now they have 20 years worth of shuttle pilots to consult in building a system that suits their needs precisely. Control systems can evolve, and should, just like everything else.
  • The CNN article you linked to says, in the second paragraph, that they are replacing CRTs with flat-panel displays.

  • not a huge suprise (Score:3)

    by jetpack ( 22743 ) on Saturday April 22, 2000 @05:43PM (#1116349) Homepage
    This isn't exactly a huge suprise to me. I had a TA in university that had previously worked on the shuttle code, and showed us the design documents and the coding procedures for some of the shuttle control code. The attention to detail was almost unreal. Even the requirements doc was triple-intense, not to mention the coding procedures themselves. The shuttle folks are coding-for-keeps, and well they should; peoples lives are *directly* on the line.

    Too bad the industry-at-large refuses to be as thorough as NASA.
  • Something, just SOMETHING tells me that no publicly-available OS is going to be used for an extremely specialized, critical application such as the space shuttle. Do you have ANY idea how complex the systems on that thing are? I sure as hell don't, but I remember reading somewhere that they are very, very complex. Very.
    Nothing short of a custum, proprietary OS is going to run all that equipment. Proprietary, embedded, real-time OS. AtlantOS or something, or maybe DOScovery.
    Gee, if they WERE using Linux, they'd have to release their code modifications, for the rest of us to install on our space shuttles.
    Keep in mind that the new control systems have probably been in design for at least 5 years. What I'm getting at is that they're not going to design a bunch of custom hardware, then apply the latest kernel patch from kernel.org and then launch it. The OS was likely designed the same time as the hardware (only FOR that hardware, strangely enough) and probably has been rigorously tested and debugged for several years without any major changes. Just guessing, but it sounds similar to logic.
  • If the industry-at-large was this anal s/w would cost 10 to 100 times. Notice in the article there was no mention on $ / LOC.

  • Re: Why retrofit these things? (Score:5)

    by mperrin ( 41687 ) on Saturday April 22, 2000 @05:56PM (#1116352) Homepage
    There's a very good reason for not upgrading the shuttle's computers: They don't need it.

    Think about it. How many other computer systems can you think of that have been running for the last twenty years? How many other systems have had the tens of thousands of hours of testing and, even more importantly, have worked flawlessly every time? The shuttle computers work. They have never failed in flight, period. Some new system? Hah. We all know how solid most applications are today. The above jokes about running CE on this shuttle show that for sure. ;-) The shuttle today does fundamentally the same thing it did twenty years ago; the calculations needed for launch and landing haven't changed. So why replace the most tested piece of code in the world with something new? Would you like to ride in the first flight with brand new software? Thought not.

    Next question: So then why upgrade the displays? Partly it's economic - those bulky old dials weigh a lot, and it's still ten kilobucks a pound to LEO. Beyond that, the interface can be made far better and more adaptable than anything you can do with gauges and switches in hardware. Anything that can be done to make the pilot's life easier is a net win. They way they did it, they basically put in a new system which takes the data and runs the displays, leaving the main computers pretty much untouched, running the same ol' rock-solid code.

    Another point about the backup system. As Detritus posted elsewhere, the four main machines all run a program called PASS, the Primary Avionic Software System , while the fifth computer runs BFS, the Backup Flight System. These two programs were written by completely independent groups of programmers. To this day, no one who has worked on one of them is allowed to ever see the code for the other. They're completely indepentent. The idea behind this is called "diverse design". The more dissimilar two systems are, the more improbably it becomes that both will fail at the same time. They originally wanted to have totally different hardware for the backup too, but that was nixed to save on costs. Lastly, the backup system most certainly doesn't kick in automatically, under any circumstances. It's the mission commander's call, whether to hit the big red button on the control stick or not. (And yes, it actually is a big red button.) The designers judged that it was better to keep a human in charge than blindly trust the software.

  • Computers in Spaceflight: The NASA Experience (Score:3)

    by Detritus ( 11846 ) on Saturday April 22, 2000 @06:04PM (#1116353) Homepage
    This is an interesting book on the history of computers in space, it is available on the web here [nasa.gov]. It has a chapter on the Shuttle's computer systems.

    Also see CACM Volume 27, Issue 9 (September 1984) for an interesting article on Shuttle software.

  • Most definitely. This is something that has mystefied me for some time now regarding the failed Mars lander: Why in the HELL would they do anything at all using English measurements? Isn't this something you learn in high school chemistry? If it's science related, USE METRIC. That's all there is to it.

  • Re:Common sense time. (Score:1)

    by Anonymous Coward
    While I can't speak for the primary system computers, the embedded are probably running vxworks, by Wind Rivers. Although they may not be fancy, its the embedded systems that control everything from the positioning thrusters to the oxygen regeneration systems to the space toilet. Theres really no reason for a super secret os, vxworks is time proven as an embedded real time os. In general, NASA and its subcontractors use old, time proven technology. In the aerospace industry, you are generally given a job, and a deadline. You don't have time to try and invent a whole bunch of new wheels, you look at the ones you have available and use those. The result is new applications that do amazing things using antiquated parts.

  • Houston, we got a problem (Score:3)

    by ContinuousPark ( 92960 ) on Saturday April 22, 2000 @06:19PM (#1116356)
    I'm not sure if this [wired.com] is what you're looking for but it's pretty similar. It's from Wired News; they were watching NTV too and listened to this conversation where one of the astronauts had a problem with, guess what?, Microsoft Outlook. =) and called Houston for help. I suppose it was running on one of the Thinkpads that they now carry with them on the shuttle.

  • Aviation in general is a conservative business. If it works, and it's safe, there's no reason to change it. ILS hasn't really changed since it was first developed (mid 30's or so?) Aviation is almost never on the cutting edge. We're still clinging on to NDB's....

  • darpa? (Score:1)

    by craw ( 6958 )
    I believe that DARPA has in the past funded research into high-tech display screens. While it is nice that the space shuttle is upgrading their displays, it also note-worthy to consider what may implemented in the future. One such technology is head-mounted displays (HMD).

    News for Nerds, right? So if like high-tech, go check out what DARPA [darpa.mil] is currently funding. Remember, high risk, but the potential for high paybacks. For HMD resolutions check out this project [darpa.mil].

    Hmmm, just had an interesting thought. DARPA played a major role in the development of the internet. DARPA obviously also funds other seriously high-tech projects. Why isn't there more DARPA stories posted here? Perhaps it is time to change to the more politically correct name, ARPA? Note to some ppl, DARPA and ARPA are the same, only the name has been changed to appease the political gods of funding.

    Side note, I don't know why the DARPA web site wants to give me cookies some of the time.

  • If it were that simple, I'd agree. However, software maintainance costs a significant amount of cash. The only real advantage, as far as I can tell, in being lax with requirements/design/coding is being first to market. Unfortunately you wind up paying for it in maintenance and bug fixes later. If you got it right the first time, you might be a bit behind the competitors, but you'd look good in the eyes of perspective clients because your software didnt suck. And you wouldnt have to pay your coders to fix things. You could pay them to add interesting and useful features.

    So, yes, in the end, it's all about cash. I'm just convinced that in the general case cash is being spent improperly.

  • Re:Good luck. (Score:3)

    by Detritus ( 11846 ) on Saturday April 22, 2000 @07:05PM (#1116360) Homepage
    Doing something new in metric units is easy. Switching from English/Imperial to metric is a bitch when you have invested huge amounts of money and training in the old units.

    Spacecraft navigation is the descendant of aircraft navigation which descended from the navigation of wooden ships. That is why they still use knots, feet and nautical miles.

    I've wondered how we could switch to metric units for the operation and navigation of aircraft without killing thousands of people in the process. Everyone is used to "500 knots at 35,000 feet", and the air traffic control system is built around those units. How do we seamlessly change that to metric units? Plus the pilots have the old units deeply embedded in their brains. They know that the stall speed of their plane is 100 knots, the fuel consumption of the engine is 10 gallons/hour and many other important facts in the old units.

  • I'd rather not have glass... (Score:3)

    by DeepDarkSky ( 111382 ) on Saturday April 22, 2000 @07:22PM (#1116361)
    glass CRTs just seems a little too vulnerable to failures to me. I mean, for god's sake, it's glass! We are talking about space vehicles which may be subject to several G's of acceleration. Granted, that may not cause any problems, but if one of the CRT cracks or breaks, it will be useless. In many ways, I think that analog gauges would be more reliable. You certainly should not put several gauges together such that there's a vulnerable single point of failure. I suppose they could make it with something like bullet proof or shatter resistant glass (I hope) to remedy this proble.

    Then there's the electrical system that's needed to drive such as system. What if the electrical system for the gauges were to lose power? No gauges? There should at least be backup analog gauges for some vital system gauges.

    What they should also try to develop that would be better than CRTs though not as good as analogs would be some kind of flexible durable flat panel display. Something like the surface of those plastic mouse pads. You could bend it, press on it, and it would not cause significant damage. Even better still, make it so that the display is not a general purpose display, but pre-configured display that changes depending on electrical current/voltage like the analog gauges. But it would have advantage over the analog gauges because it would not have any moving parts. To get an idea of what I'm talking about, imagine those Duracell battery testers used as gauges. I mean, wouldn't that fairly resistant to all kinds of adverse conditions and abuse? And if you make them like "skins" that you could just snap in place, then you could easily carry redundant gauges because they would be very light!

    Come on, they could do better than that! Or am I just completely wrong on this?!

  • OK, I know this is off topic, but can someone answer a question for me that was spurred by this post? What the heck is a "real time" OS? And how does it differ froma Linux or Windows? just curious, Chris

  • I don't see why these processes couldn't be implemented in OSS work. They already are, in part. The "many eyes" principle ensures rigourous code checking.

    The reason open software still tends to be buggy (at least compared to the space shuttle software) is partially the lack of a coherent plan for the software before writing begins. Most OSS projects start life as a small chunk of sloppy code that is interesting enough to attract the attention of other developers. Consider what would happen if, instead, a project began life as a high-level spec, and the specs were filled out through an open-source process before coding began. This could produce very sophisticated code quite quickly. It requires pretty good organization, though.

    Interestingly, most large OSS projects use CVS repositories and bug-tracking software, which fill the role of the code and error databases mentioned in the article. So it would seem that the only thing standing between OSS and this kind of sophistication is the need for a coherent and detailed design for the software before coding starts.

    That said, I doubt that this process would ever generate anything much more creative than a hardware driver. The Shuttle Group programmers themselves admit that the all-important process stifles creativity. If you think about it, considering the level of specification they've been given, their end-product is really little more than a device driver for the entire space shuttle. You would be hard-pressed to write something requiring real innovation this way. If you tried, you'd probably find that you spend the same amount of effort making the spec book for the software, which would end up with subtle bugs in the spec due to the evolutionary nature of the project, which would then get translated directly into software bugs in the code. Oh well...

  • The shuttle can't land by itself (Score:1)

    by Anonymous Coward
    The landing gear can't be deployed except by a human pilot flipping the right switch. There must be a live pilot on board for else it would make a big hole in the runway.
  • As I understand it the only reason they still use ILS and not GPS is that the gov't messes up the GPS signal so it is not accurate enough. They are investigating ways to fix it up becuase GPS would be a lot safer and more accurate than ILS, at least in theory. But you are right that aviation is generally a conservative business, which is as it should be. If I get on a plane to go somewhere I want it to work. I don't want to be put in the role of test pilot for some new, unproven technology someone thought had a high "gee-whiz" factor...

  • Re:The shuttle can't land by itself (Score:4)

    by CharBoy ( 100343 ) on Saturday April 22, 2000 @08:23PM (#1116366) Homepage
    This is actually true, the Shuttle cannot deploy its landing gear in the computer. The previous users post shouldn't have been moderated down, he's absolutely correct and it is on-topic.

    The astronauts objected to the computer being able to deploy the gear automatically, and this was given to them by the software people as a meaningless victory in their fight against the full automation of the shuttle.

    The concern the astronauts claimed was that if a computer glitch caused the gear to deploy while in orbit, the result would be a loss-of-vehicle scenario.

    The software is capable of handling every aspect of a shuttle landing except for the landing gear itself.

    An interesting side note, if the gear aren't deploying within a half a second or so of the deploy switch being pressed, there are pyrotechnic charges that deploy the gear by force. As far as I know, they haven't been needed yet, but I imagine it'd be quite a sensation...
  • Don't know about the shuttle, but the glass mission control center in Houston uses Alphas running Digital Unix. The old dial-and-gauge mission control center used during the Apollo missions is now preserved as a histroical monument.

    ----
  • What if they branched out and picked up some bucks on the side creating and selling bug-free software? How cool would it be to run NASA-WARE ?
  • I guess in a few weeks all of us will be able to run our own experiments on the ISS. "if you know what i mean of course"

    insert "buhaha...buhahah"

    B
  • How many other computer systems can you think of that have been running for the last twenty years? How many other systems have had the tens of thousands of hours of testing and, even more importantly, have worked flawlessly every time?

    Believe it or not, but the PDP/LSI-11 series of computer is still in use in sheet metal and other factory floors even today. The damn thing is a workhorse which lives in production even after the death of Digital -- long after DEC stopped supporting the hardware.

    Not that this diminishes the value of your point. :-)
  • First off, I'm about as far from being an expert in such matters as you can get.

    It occurs to me that CRT's would be far BETTER for high stress environments, simply due to the removal of moving parts.

    Doug

  • Well, I've seen a few 'comical' posts about this story, so I just have to weigh in. I'm an Aerospace Engineer, and I have friends who are 'Aviation Computer Scientists'. That means that they program the computers that run airplanes.

    Observation 1. The 'glass cockpit' displays in airplanes (and, presumably the Space Shuttle) do not run off the shelf operating systems. They are custom, proprietary operating systems written specifically for what they do. They have no need to use a general purpose OS, when what they need above all else is rock solid stability for just a few functions. One of my old college buddies is even working on the code for the 'operating system' of the new space station.

    Myth: Glass displays are fragile. This is simply not true. The space shuttle has had glass displays since the first one, but there just weren't as many, and as prevalent as they are after this refurbishing. Glass displays can handle the 4 G's experienced by the shuttle just fine, as well as the vibration. For one, jet fighters have glass displays in them, and regularly handle worse G forces than the shuttle ever does.

    Hopefully this clears it up a little more. Any questions, just post.

  • I'm glad to hear that ... our company has quite high standards too, but that is really only because they are smart enough to employ exceptionally good developers and testers. The process itself sucks rock, but we do ok because of the folks involved in the process. I'm a developer for these guys [zeh.com] (turn on your javascript :P) who do you work for?
  • The year that SIGCHI was held in Austin (88? 89?) one of the keynote speakers was a shuttle astronaut. He said that since three systems in the cockpit had been desinged by three different companies with no overall design spec, the shuttle "programmers" had to learn three different obscure languages simly to "fly the bird:" a rudimentary BASIC-like command language, octal, and hex! One of the earliest missions nearly landed in the ocean because of a hex-octal mistake. It was this guy's opinion that the shuttle was as much an ergonomic lab as a zero=gravity, low=space lab. Also, the designers had bolted shut one of the access panels because they were sure nobody would ever need to get into it. One mission was saved only because one of the astronauts had smuggled up a Vise-grip, which was verboten.

  • Re:Why retrofit these things? (Score:3)

    by kzinti ( 9651 ) on Sunday April 23, 2000 @03:45AM (#1116391) Homepage Journal
    The tapes aren't big enough to hold everything, so there are seperate tapes for take-off, landing, orbit, etc. The tapes need to be changed by hand on the first four machines.

    This isn't correct. There are two tape drives called Mass Memory Units (MMUs); each has an identical copy of the flight software, both PASS and BFS. The drives are physically separate units from the GPCs; for the sake of redundancy, they communicate across separate data buses. All five GPCs share the same two MMUs.

    The tapes in the MMUs are large enough to hold all the flight software for the different phases of the mission. These separate programs are called "Ops modes". There is an ops mode for each mission phase: OPS 1 for Ascent, OPS 2 for Orbit, OPS 3 for Entry/Landing. There are some other modes, but these are the significant ones.

    The tapes do not have to be changed during the mission, but an crew member must manually enter the command for the ops mode transitions. Within an ops mode, the flight software can make its own transitions between program phases -- these are called "major mode transitions" and don't require an access to the MMU.

    I'm not a FSW expert, but I've spent the last two+ years writing GPC emulators, so I've become familiar with some of these concepts.

    --Jim

  • Re: GPS for aviation (Score:3)

    by Ledge Kindred ( 82988 ) on Sunday April 23, 2000 @05:32AM (#1116402)
    If you have a military GPS receiver, you can decode the encrypted military GPS signals that are on a different frequency than the civilian signals which will allow you to find your GPS coordinates down to a few inches.

    But there are a lot of other reasons why GPS isn't very good for aviation use:

    GPS does not handle altitude very accurately.

    GPS does not handle very high speeds all that well.

    I imagine the GPS system would break down entirely when you reach a good percentage of the altitude of the GPS satellites themselves. Not so good for the shuttle. (Seeing as this is what the article is about...)

    Here's a great URL for lots of info on how GPS works: http://www.trimble.com/gps/index.htm [trimble.com] (It uses Shockwave, but you can still view it without.) Here's another with lots of information about GPS units: http://joe.mehaffey.com/ [mehaffey.com]

    -=-=-=-=-

  • Good lord, if things are THAT complex they're DEFINITELY not going to want to write everything from scratch on their own! Do you have any idea how many bugs they'd have to work out on their own - bugs that have ALREADY been worked out for large & complicated systems which are put under severe stress every day.

    More than likely, they picked something which was as simple in construction as possible, but still can get the job done, and then had their own people crawl over the architecture & source code with a fine-tooth comb. You'd better believe that they're not going to blindly trust a software company just because they SAY their code is mission-critical ready - they're going to want to look at the source code themselves.
  • Perhaps if you had a detailed enough simulation of the software's working environment (with the caveat that you REALLY have to trust your simulator...), then you could have hundreds (or thousands) of people trying to write code, and then the simulation tries to break everything.

    You could have a dedicated bunch of people who deliberately try and break any proposed system (including setting up hardware glitches, like cosmic rays or something).

    Over time, the software would probably evolve into something quite robust.
  • Yes, it would take a group of rocket scientists to design an OS from scratch for the space shuttle.

    (....pondering the previous statement)

    Whatever they use, it's surely been in testing for years. I'm sure that the "booster O-ring bug" taught a hard lesson on cutting corners.

    Linux is great, but it can't do everything. I still have to wipe my own ass.
  • I'm curious -- does anyone know the error rate for Airbus software?

    Airbus has taken the position that the computer, rather than the pilots, can fly the airplane better. In fact, the pilots cannot override the computer.

    As a pilot and software engineer, I find the Airbus philosophy fundamentally flawed. I can think of a few incidents on Boeing aircraft that would have turned into an "everyone dead" senario on a fly-by-wire Airbus (A320/330/340). You won't catch me dead flying on or acting as Pilot In Command of an aircraft exhibiting the Airbus philosophy.

    OK, French Persons, flame away. :-)

  • Glass displays can handle the 4 G's experienced by the shuttle just fine, as well as the vibration.

    Yes, but they're bulky, heavy, use lots of power and generate lots of heat. Basically the worst possible thing to use on a spacecraft besides uninsulated O-rings... Why didn't they go with flat-panel (plasma or LCD) displays?

    Your Working Boy,
  • In fact, the new Airbus (A3xx) is supposed to have all of their flight instruments connected via an Ethernet interface.

    I'd hate to see what happens when all those controls attempt to operate simultaneously.. Gives 'packet collision' a new image, eh? (and judging how cheap some of these airlines get you KNOW like Tower Air is gonna go for the daisy-chained Linksys hubs...)

    (Token Ring and FDDI are far more suitable for realtime communications than CSMA/CD systems, particularly the interference-resistance of optical connections from FDDI... And hell, the French love AIX/IBM...)

    Your Working Boy,

Slashdot Top Deals

Do you suffer painful illumination? -- Isaac Newton, "Optics"

Close