Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Science

Hubble Repairs Declared "Complete Success" 67

Posted by Roblimo from the upgrades-that-work dept.
Matt G writes "The Hubble Telescope's brain transplant seems to have been a perfect succss - British-born Michael Foale and Swiss Claude Nicollier carried out the delicate operation of installing a new computer as they flew over Australia at an altitude of about 600km (360 miles) on Thursday. The full story is posted at The BBC News site here. "
This discussion has been archived. No new comments can be posted.

Hubble Repairs Declared "Complete Success" More

Hubble Repairs Declared "Complete Success"

Comments Filter:
  • Does anybody know what operating system (if any) runs the new 'brain' ? Is it VxWorks like pathfinder, or maybe just a very thin blanket over the hardware (just enough to get C code working or something) ?

    Actually, why was it needed to upgrade the cpus? I mean, I thought hubble's cpu doesn't do more than controlling movements and handling communication protocols so it can send pictures. The new 'brain' will be 20 times faster, but what's the point? Does the telescope perform calculations on-site ? Just wondering.

  • Best Wishes to NASA (Score:5)

    by Effugas ( 2378 ) on Sunday December 26, 1999 @03:10AM (#1444649) Homepage
    The Hubble Repair mission should remind us of what, sadly, has been somewhat forgotten as of late:

    These guys know their stuff.

    When I sysadmin a machine I'm standing next to...I'm standing on something. I'm not floating in nothingness, hoping my toolkits don't float away into the emptiness of space, trying not to bend a couple hundred gold pins while wearing massive mittens and a spacesuit that I have to continually check for tears.

    I also don't generally do it for eight hours straight without so much as a water break.

    Similarly, when I'm admining a system remotely, I'm not piggybacking on top of a defense network that I can lose access to at any moment, nor am I trying to fit modern computational systems into a space-hardened antiquated piece of hardware. These are some crazy skilled coders, and they deserve much more respect than the budget-forced unit conversion fiasco implied. (We should be ashamed for the reaction! These )

    I'm proud of NASA, and I'm proud of the engineer-athlete-scientists who made the Hubble space telescope possible. Thank you. Your work is appreciated.

    Yours Truly,

    Dan Kaminsky
    DoxPara Research
    http://www.doxpara.com
  • Well, I'd imagine that since the Hubble is rotating around the earth, as well as a large number of other objects, lots of calculations have to be done to determine what direction to be facing and when. It probably gives NASA scientists an easier time to let Hubble figure some of this out for itself...

    It's probably specially-developed embedded code, too. Specialized real-time operating systems seem to run rampant in projects like these ;-)

    Now what NASA needs is to get a few satellites running Linux up... (do they have some already? who knows...)

    Maybe slashdot readers could contribute enough money to put Rob up in space in a giant hamster-wheel like device for a few days. Then he could go visit the moon (which people haven't done in a long time), and discover some deserted moon base. Then they could ship him out to Jupiter, where he would misteriously disappear into a large black hole at least several thousand miles deep on a small moon that is only a few hundred miles in diameter...

    Oh wait, that was the plot from 2001.


    progress, n.:
    Medieval man thought disease was caused by invisible demons
    invading the body and taking possession of it.

    Modern man knows disease is caused by microscopic bacteria
    and viruses invading the body and causing it to malfunction.

  • I would like to be the first to congratulate the people at NASA for doing a fine, fine job.

    The hubble telescope is a big focus for the public's attention. With the recent, uhm, mistakes at NASA, I'm sure that alot of managers were sweating about the PR disaster if this would have gone anything but perfectly.

    I can see the headlines now:

    "NASA fails again!"
    "Hubble goes back to sleep"
    "Public faith in NASA shaken"
    "NASA's funding cut for continued blunders"
    "NASA shuts down for restructure"
    20 years later
    "Anybody remember when we went to the moon?"

    Perhaps I'm being a tad melodramatic here, but who else had that queasy feeling that hubble wasn't going to go back online.. ever?

    Once again, congrats to all involved, and good luck in all future endeavours.

    Rami James
    Israel

    --
  • Actually, why was it needed to upgrade the cpus? I mean, I thought hubble's cpu doesn't do more than controlling movements and handling communication protocols so it can send pictures. The new 'brain' will be 20 times faster, but what's the point? Does the telescope perform calculations on-site ? Just wondering

    Don't know for a fact but based on reading some of the other Hubble related threads it's likely that it was mainly just 'preventative maintenance'.. the old CPUs probably already had some damage from the length of time they'd been up there and they needed to be replaced anyway, and possibly the 486s now cost the same as the original CPUs did when they were built. (ie. if your P133 dies now you're likely to not get another P133 CPU but some faster Socket 7 CPU like a K6-2 because it costs the same). Yeah. That's my theory ;)

  • You really got the plot wrong though.

  • HST maintenance (Score:4)

    by Captain Zion ( 33522 ) on Sunday December 26, 1999 @03:46AM (#1444654)
    it's likely that it was mainly just 'preventative maintenance'.. the old CPUs probably already had some damage from the length of time they'd been up there and they needed to be replaced anyway
    From the Space Telescope Science Institute [stsci.edu]:

    When originally planned in 1979, the Large Space Telescope program called for return to Earth, refurbishment, and relaunch every 5 years, with on-orbit servicing every 2.5 years. Hardware lifetime and reliability requirements were based on that 2.5-year interval between servicing missions. In 1985, contamination and structural loading concerns associated with return to Earth aboard the shuttle eliminated the concept of ground return from the program. NASA decided that on-orbit servicing might be adequate to maintain HST for its 15- year design life. A three year cycle of on-orbit servicing was adopted. The two HST servicing missions in December 1993 and February 1997 were enormous successes. Future servicing missions are tentatively planned for mid-1999 and mid-2002. Contingency flights could still be added to the shuttle manifest to perform specific tasks that cannot wait for the next regularly scheduled servicing mission (and/or required tasks that were not completed on a given servicing mission).

  • British-born Michael Foale and Swiss Claude Nicollier carried out the delicate operation of installing a new computer as they flew over Australia at an altitude of about 600km (360 miles) on Thursday.

    I thought AIR was required to fly! :->

    Orbited is the right word - nothing ruins a good science story than bad reporting. Some of these science writers need to learn how to pay attention to detail...
    --
  • And does anyone know what kind of code they run? C? Assembler?
    --
  • It's good to see NASA having success after the loss of those two Mars probes.

    Notice how the media's not making a big deal out of this. I suppose you could argue that successes just don't sell as many newspapers as failures. I think, however, that the mass media likes to manufacture issues by hyping up failures like the Mars mission and then making another story out of the public stir they create. Think about it; "NASA Mars Probe Lost" is just one story, but "Second NASA Failure This Year Causes Public Doubt About Future of Space Agency" can be a snowballing event -- shorts, 'talkback' segments, polls, comments from pundits, etc.

    After all, when you're on 24 hours a day, you can't be expected to fill the time by just reporting what really happened when, where and why. That'd require too much actual reporting.

    ----

  • Congratulations to all involved, it's nice to see something going right for nasa. It's interesting how the british media always say "British born Michael Foale" as if that somehow means that we are involved in this. It would be nice if Britain actualy had some real contribution to something like this, but do they really need to say this every time the story is mentioned.
  • Just wondering: Y2K?

    I can't see any compelling reason for changing the cpu, if the old one was working. If the old cpu had failed, then why not replace by a new identical chip? Why risk the chance of finding a bug in the new one? Unless they want to try some new algorithm, why upgrade?

    Of course, pointing the HST isn't "just" that simple. At the precision level required, some of the calculations are incredibly complex. There are also other problems, like battery management, for instance. The HST crosses the shadow of the earth several times a day, and this imposes stringent requirements on the batteries, not to mention thermal stresses.

    Just to give you an idea of the precision needed for spacecraft control, I work at a commercial satellite control center. Some years ago, we did a spin rate measurement on a satellite, while moving fuel from one tank to the other. This measurement had to be done at a 0.000001 rpm precision level, for a nominal 30 rpm rotation, meaning 0.000003% error. The result? A few extra million $ worth in the satellite value, due to a more accurate estimate of the remaining fuel.

    To do a measurement to this precision one has to consider even the smallest details. For instance, the satellite is in a 24 hour period orbit, and each spin rate measurement took about 10 minutes. In this time, the satellite moved along its orbit, changing slightly the angle of the antenna (which was pointed to the Earth) with relation to the sun. This caused a variation of a fraction of a degree in the temperature of the satellite structure, which was enough to cause a marked difference (much greater than the required precision) in the spin rate, and had to be discounted from the measurement.

    So, I guess NASA improved the pointing algorithms in HST to a degree that made necessary a more powerful cpu.
  • Yah, I know, but I was trying to make it sound like I was making it up as I went along...
    2001 was _the_ most realistic sci-fi movie ever done, *please* be defensive, my fault for mutilating it ;-)

    Hrmm.... the accuracy is claimed to have been increased... would this imply that they had 16-bit etc. processors before that were incapable of higher-precision arithmetic?

    In that case, they probably could've used a high-precision arithmetic function library, but it would've been slow - a good reason for the upgrade, I guess.
  • Considering that this was just a routine maintenance mission that shouldn't really be that interesting in itself it seems to have got quite a lot of coverage in the uk.

    Routine servicing mission doesn't fail is never going to grab the headlines, but it did seem to get quite a lot of coverage here.,

  • Now that the HST is back online, perhaps NASA might use it to examine the recent failed probe. I hope SOMEONE has thought to at least consider using it to look for the lost probe. If nothing else to see if it was destroyed on impact. I know it is normally used for deep space observing, but hey, I know I for one am at least curious as to the condition of the failed probe.
  • "Notice how the media's not making a big deal out of this".

    Notice also how the media completely disregars all successfull airplane landings and focuses on those that end up in fireball killing 200 people. Some nerve they have..

    It's very easy to critize the BBC and forget that there is reason why some stories get less atention. The reason is us, we don't care. Counter what you might think, newspapers (and by newspapers I mean newspapers, not tabloids etc) actually do try to serve the public. The thing is that the public is not served by running stories nobody cares about. Stories like "NASA succesfuly launches a satellite for the 748th time" intrest very few people so they are left for publications the specialize on astronomy.

    Success is not news if it happens often or is in some other way obvious (neither is failure for that matter). That is why you'd expect to see headlines like "Kid not hit by a car while crossing road", "Actor James Coburn still alive", or "cold fusion experiments prove unsuccesfull" in The Onion, not The Times.

    "I think, however, that the mass media likes to manufacture issues by hyping up failures like the Mars mission and then making another story out of the public stir they create".

    The Mars missions were largely hyped even before they turned out to be failures. Remember the Pathfinder mission and how it created a lot of media attention even though it was a complete success? A lot of people were expecting something similar, and when that didn't happend, it became a big news. Very logical, IMHO.

  • thought AIR was required to fly!

    Nope. The Dictionary gives as definition 1h of the verb "Of an aircraft or spacecraft: to travel through the air or through space."

    My bad. I was thinking of birds and airplanes, with no dictionary handy.

    You gotta love the English language. To paraphrase President Clinton: "that depends what 'is' is..."
    --
  • There's no way the HST has enough resolving power to find something so small. Especially with Mars now well past opposition. A better idea is to use the Mars Orbiter to try to find it, which I believe they are doing. But even that is a long shot.
  • If you privatize NASA, technological and procedural breakthrough's will all be constrained under patent and licensing constraints that will ultimately hinder progress.

    For NASA to *cooperate* with private firms is good, but NASA must remain ultimately free of market constraints'
  • From what I read, they used to use assembler on the ancient 1970s CPU they had in there. One of the reasons on the NASA site for the CPU upgrade to 486s is that it is cheaper and easier to maintain code in C (well some "high level language" than assembly.
  • Interesting. I had no idea the precision needed to be so great. Kinda makes you wonder what kind of code their using. Numerical Recipes in C?

    But I don't think the upgrade had anything to do with Y2K. The idea that NASA would wait until the 11th hour to attempt a fix on their most visible -- and really, one of the only recent -- success stories seems pretty unlikely to me.

    John
  • That is why you'd expect to see headlines like "Kid not hit by a car while crossing road", "Actor James Coburn still alive", or "cold fusion experiments prove unsuccesfull" in The Onion, not The Times.

    Cute, but off the mark. A successful and difficult space mission repairing a costly piece of taxpayer-funded hardware is a cause for celebration and should be lauded as such. Modern media only celebrates Tradgedy and Celebrity.

    Notice also how the media completely disregars all successfull airplane landings and focuses on those that end up in fireball killing 200 people.

    What bugs me is how the airplane/fireball incidents are blown out of all proportion. At least 20,000 if not 200,000 people are killed every year by stalkers, 'loved-ones' and spousal abuse, but this is not considered 'tragic' because it happens over a long period of time. Knowledge of this would help to prevent similar abuses in the future; knowledge of 200 people dying in a fireball is interesting but otherwise useless information, serving only to raise fears about which nothing can be done and encourage further consumption of sensational media. Modern media, BBC or otherwise, serves to further it's own existence long before it serves the public.

  • Good for NASA! Now they need to release a few pictures to the media to convince the world that their streak of bad luck is over.

    I'm just glad I don't need to do a spacewalk to upgrade the machines I admin!

  • Re:Technical details? (Score:3)

    by breser ( 16790 ) on Sunday December 26, 1999 @09:51AM (#1444677) Homepage
    According to the mission website [nasa.gov] the new computer has advanced radition protection. The radiation in space will eventually destroy the computer. They replaced it with a newer computer that was designed to have a longer life span. Additionally the new computer has six times as much memory and three times as much processor speed. The new processor speed and memory will be used to give them more accuracy in the pointing of the spacecraft.
  • Congratulations to the team members for accomplishing this difficult mission so perfectly!

    We Swiss people of course are proud of Claude Nicollier beeing part of the team.
  • The hubbles computer runs off a 486 Intel CPU. I am aware that the CPU had to meet certian standards like radiation shielding and emp resistant. But aren't there better CPU's out there that meet those standards? It seems to me that NASA (which can spend billions on failed mars missions) might have the money to pay for a better type of CPU (should one exist).

  • I'm sure they are not using Numerical Recipes. It's a fine book, excellent for general information, but shallow - for each program there you can find several whole books if you search. I usually copy from NR the first prototype in each project I do, then I start digging and develop something more advanced. My guess is they are using in-house developed code, written in Fortran 30 years ago and never changed until now. If it's working, don't fix it, that's the spacecraft engineer's motto. They are extremely conservative and never risk introducing bugs into the system if they can avoid it. For the precision, do your own calculations: a *small* communications satellite starts life with 150kg of fuel, nominal lifetime is 12 years, earns US$2 million / month. How much each kg of fuel is worth? It literally *pays* to be precise. Of course, I work for a commercial operator, NASA has other worries.
  • Just wondering: Y2K?

    Remember that this was a contigency mission to replace the gyroscopes, not a normal servicing mission. The new computer was simply an added bonus that was ready to go up now. This mission was originally scheduled for mid-2001. That completely kills Y2K as any reason....

  • As I understand it (IANA NASA guy), there is a mandatory technology freeze approximately 5 years before launch. Every component must go through rigorous tests, and this is the only thing holding them back really.
  • Yeah... just wondering...

    Why would they fly on Christmas 1999, coming back just before 2000/01/01? After Challenger they had stopped flying during your winter - too much risk of freezing those rubber o-rings.

    Spacecraft are very sensitive to date issues. The Earth's axis wobbles slightly and in 1900 the North Pole was pointing towards a direction that is significantly different from the 2000 direction. This difference alone is enough to lose a mission, even if the North Star seems to be exactly at the same place to the naked eye.
  • More wallpaper for me!
    Seriously, even if you don't appreciate the scientific level of what NASA is doing. You can get some really awsome images [stsci.edu] from them. Check out NASA's awsome collection of images, which have aver generous copying policy.
  • I highly doubt the cost of the CPU was an issue in this case. I mean, a 486 CPU is worth about $10 now. If they were looking for speed, they could have used an Athlon for about $500, so apparently they weren't.
  • "Cute, but off the mark. A successful and difficult space mission repairing a costly piece of taxpayer-funded hardware is a cause for celebration and should be lauded as such. Modern media only celebrates Tradgedy and Celebrity"

    While you may appreciate the difficulty and the true nature of this mission, to someone who doesn't know it just looks like a routine service mission. Sure it seems difficult because it is in space but it isn't something new and exciting. It's ok for people to not be excited about this.
  • Also keep in mind that a good-sized chunk of NASA's technological advances come in the realm of metallurgy, plastics, life support, etc. Using 5 year old technology in the complex systems they design makes it even easier to implement them in the "real world".... Which is better, a system that does the job superlatively which costs $500 to build, or one that costs $5000 and can play Quake3 while doing its job?

    Admittedly, if I was using it on my desktop, I'd like to be able to do other stuff at the same time; for a dedicated system, I'd rather have it be simple and capable of doing its job... There's just less to go wrong that way...

  • There is a design freeze built into almost every program, but it is not, repeat not, based on anything as inane (or prescient) as a five-years-before-launch timeline. For example, the faster-better-cheaper probes don't necessarily even take that long from project approval to launch! The Confirmation Design Review of MPL was barely two years prior to launch. (Read into that what you will.)

    The reason for using a 486 instead of a Pentium can be as simple as power and cooling requirements, or as complex as the issue of running a custom RTOS written in C versus running an off-the-shelf Windoe Manager written in C++ and something else godawful. (Or maybe that's the simple decision.)
    ----
  • The computer on Hubble is programmed from the ground. The new computer simply upgrades the processing capability of the existing software, perhaps allowing new routines. (Early on, with the original 286 board, they discovered that the anti-jitter routines they devised wouldn't fit -- let alone anything else like data transfer. Presumably 1993's 386 co-px solved that.)

    The post-Challenger restriction on flights was NOT on an artificial window like "winter", but on specific temperatures reached at the Cape. It's very, very rare for even a January freeze there, but it happened in 1986. No, they should not launch the shuttle, even with redesigned O-rings and boosters, if it gets too cold. But just because it's December does not mean the same thing as being too cold.
    ----
  • I can't see any compelling reason for changing the cpu, if the old one was working. If the old cpu had failed, then why not replace by a new identical chip? Why risk the chance of finding a bug in the new one? Unless they want to try some new algorithm, why upgrade?

    The Hubble was designed from the beginning to be periodically upgraded, not only by swapping out instruments and installing different ones, but in improving the instruments in place. The computer upgrade has been in the works for several years; a prototype was tested aboard the John Glenn shuttle mission last year. This isn't an idle "hey, let's toss a new box in" procedure.

    You are correct about satellite precision; in fact, pointing precision has always been one of the sticking points regarding the computer. The new computer gives them more memory to run more complex routines, probably something the controllers have been clamoring for since 1993!
    ----
  • how does NASA spell success after two complete failures with the mars probes?

    The Hubble team is an entirely separate team from the Mars probes. One team had a success. Two Mars teams had failures.

    Anyway it's not like they haven't messed up on the hubble project before.

    True dat. But the embarrassment over the Hubble optics could have been handled much better; they dug their own grave there.

    We need to hurry up and privatize NASA befor ethey thow away any more tax dollars.

    So what business, exactly, is going to spend money on a Mars probe? Just asking.

    The MPL mission was about as privatized as a government program gets: the whole probe was designed and built by outside contractors to a NASA spec. Really, this is the way things should be done. MPL was a failure within aceptable risk. MCO was a horrible avoidable failure, but it's possible that the govt-contractor relationship was partly to blame. This isn't easy.
    ----
  • Yeah, I mean, two failures in a year? It's not like you have to be a rocket scientist to run something like NASA...

    There's a reason for that phrase, BTW.

    Cheers to NASA for having the skill to take into account as much as they do and still come out with a comparatively very high success rate. I'm reasonably secure in the notion that, granted all of their money and all of their experimental data, I could probably get a rocket to ... fall over.
  • Now that Hubble is repaired and according to earlier comments can see near the beginning or first "light", what do we expect Hubble to "see"? Is there an "edge" or "limit" to the extent that Hubble can "see" ... whether x-ray or infared or visible spectrum ... what to see ... is this the first "light"? How about a topic on "extended" Hubble for some comments on how far can we see?

Slashdot Top Deals

"The whole problem with the world is that fools and fanatics are always so certain of themselves, but wiser people so full of doubts." -- Bertrand Russell

Close