Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
NASA Software

Officials Raise Concerns About Software for the Most Powerful Rocket Ever Flown (msn.com) 71

Posted by EditorDavid from the mission-control dept.
NASA's 322-foot rocket "Space Launch System" rocket "would be the most powerful rocket ever flown, eclipsing both the Saturn V that flew astronauts to the moon and SpaceX's Falcon Heavy," reports the Washington Post.

But "it's not the rocket's engines that concern officials but the software that will control everything the rocket does, from setting its trajectory to opening individual valves to open and close." Computing power has become as critical to rockets as the brute force that lifts them out of Earth's atmosphere, especially rockets like the SLS, which is really an amalgamation of parts built by a variety of manufacturers: Boeing builds the rocket's "core stage," the main part of the vehicle. Lockheed Martin builds the Orion spacecraft. Aerojet Rocketdyne and Northrop Grumman are responsible for the RS-25 engines and the side boosters, respectively. And the United Launch Alliance handles the upper stage. All of those components need to work together for a mission to be successful. But NASA's Aerospace Safety Advisory Panel recently said it was concerned about the disjointed way the complicated system was being developed and tested...

Also troubling to the safety panel was that NASA and its contractors appeared not to have taken "advantage of the lessons learned" from the botched flight last year of Boeing's Starliner spacecraft, which suffered a pair of software errors that prevented it from docking with the International Space Station as planned and forced controllers to cut the mission short. NASA has since said that it did a poor job of overseeing Boeing on the Starliner program, and has since vowed to have more rigorous reviews of its work, especially its software testing...

NASA pushed back on the safety panel's findings, saying in a statement that "all software, hardware, and combination for every phase of the Artemis I mission is thoroughly tested and evaluated to ensure that it meets NASA's strict safety requirements and is fully qualified for human spaceflight." The agency and its contractors are "conducting integrated end-to-end testing for the software, hardware, avionics and integrated systems needed to fly Artemis missions," it said.
This discussion has been archived. No new comments can be posted.

Officials Raise Concerns About Software for the Most Powerful Rocket Ever Flown More

Officials Raise Concerns About Software for the Most Powerful Rocket Ever Flown

Comments Filter:

  • Still less than the Energia (Score:5, Informative)

    by Sique ( 173459 ) on Sunday November 01, 2020 @12:58PM (#60672218) Homepage
    With 32,000 kN total thrust in a vacuum, the SLS would about as powerful as the Russian Energia of 30 years ago, which had about 32,000 to 35,000 kN thrust. Otherwise, Energia was so expensive, that it just made two starts in total.

    • Re:Still less than the Energia (Score:5, Interesting)

      by Ecuador ( 740021 ) on Sunday November 01, 2020 @01:37PM (#60672350) Homepage

      With 32,000 kN total thrust in a vacuum, the SLS would about as powerful as the Russian Energia of 30 years ago, which had about 32,000 to 35,000 kN thrust. Otherwise, Energia was so expensive, that it just made two starts in total.

      It was not an issue of the cost of Energia specifically, but the Soviet Union fell so the Buran project (main reason for developing Energia) was abandoned. Energia's boosters or derivatives of still power current rockets at both the East (Zenit) and the West (Atlas V). So the technology was probably not "expensive" compared to other systems, especially those made by various cost plus contractors in the West...

      • Oh, but it was very expensive. Bleeding edge manufacturing, especially for the Soviets. This is why the RD-170 derivatives have not replaced the WW2 technology Soyuz engines despite being so much better.

        • Re: (Score:2)

          by Ecuador ( 740021 )

          Yes, it was expensive compared to the other projects the Soviets were running but it was not nearly the same scope. I was responding to a post where it was compared to the SLS, my point is that it would definitely be much, much cheaper than the SLS. The poster did not make it clear if he meant it was expensive for what the Soviets could afford at the time (true) and not expensive for what it was (not sure, but definitely would come ahead of the SLS at value/money)

        • Uh...RD-170 derivatives (specifically RD-193) will have replaced Soyuz engines in future units of Soyuz-2.1v, once NK-33s run out.

          • Maybe. But probably not.
            Russia had many projects over the years to use the RD-170 technology - Soyuz-5, Rus-M, Angara. Only Angara ever flew. Just two times, years ago. These engines are unfortunately so expensive to manufacture that Russia only has been able to sell them way below cost.

  • Re: (Score:2, Funny)

    by account_deleted ( 4530225 )
    Comment removed based on user account deletion

  • These sorts of problems are fundamental (Score:4, Interesting)

    by Rei ( 128717 ) on Sunday November 01, 2020 @01:08PM (#60672248) Homepage

    ... to the analysis-paralysis design approach, where everything has to be perfect, every last bolt qualified to its odds of failure, before you fly unmanned test vehicles.
      Rather than just iterating rapidly and cheaply, seeing what breaks, and flying so many times that you've proven the rocket as a whole, rather than every last bolt.

    Rapid iteration is a good thing. And when humans aren't involved, failure should be an option. We learn through our failures.

    • its called "normalization of deviance" where people stop caring about things being perfect because of human nature to stop caring when things keep going right.

      at thiks point, i realize a significant portion of the population doesnt value human life and actively goes to huge rallies and complains about wearing masks, so they dont care if astronauts die either....

      but it really does destroy momentum for public support of a program when it repeatedly kills astronauts. so even people who dont care if the

      • Re:thats how NASA killed 14 astronauts (Score:5, Interesting)

        by Richard_at_work ( 517087 ) on Sunday November 01, 2020 @03:04PM (#60672632)

        The Space Shuttle was a death trap from day one, theres your "normalisation of deviance" - its the only crewed launch vehicle without a survivable escape option during 100% of the launch (astronauts had something like a 30% window during the entire launch where they could potentially survive and even then only by doing untested manoeuvres to turn the craft around etc - the failure conditions which resulted in death with no option of escape made up the remaining window, plus the Space Shuttle was the only crewed vehicle to date where the heat shield is not protected during launch, resulting in unsurvivable damage being possible on ascent).

        Thats the Nasa we are talking about here - things might have changed with retirements etc, but overall the goals havent, and the budget restrictions and timescale pressures are still there.

        • An iteration of my code just wastes a few cycles of CPU time. An iteration of a rocket costs $$$$. So getting it right by design rather than trial and error is critical.

          I wonder how modern software compares to the Apollo mission. They used very few lines of (assembler) code as the rope memory was very limited.

          Is today's software 100 times as complex? 1000 times? 100,000 times? Or are they using the latest JavaScript frameworks? :(.

          • I actually agree with you. It's easier to test and verify 1K-2K of assembler vs 1M lines of java/c#/go/koitlin/python or whatever else they decide to you.

        • Re: (Score:1)

          by Tablizer ( 95088 )

          The Space Shuttle was a death trap from day one

          True, but both accidents were probably preventable with the shuttle as-is.

          • > but both accidents were probably preventable with the shuttle as-is.

            Probably, but only by design and implementation changes beforehand. And remember that anything you change is itself a source of new flaws.

            Remember also that the causes of accidents can occur faster than the human decision process - or even the twitch-response time. So prevention indeed, rather than escape-response.

            • Re: (Score:1)

              by Tablizer ( 95088 )

              No. In both cases the commission said there were reasonable actions that could have been taken to reduce the risk of or prevent death.

        • How does the shuttle compare to other manned launchers as far as demonstrated safety? 2 losses out of 135 launches.

          Soyuz has had one fatal accident out of 140 launches and several close calls.

          Not a statistically significant difference. Most other launch vehicles haven't done enough missions to have sufficient statistics yet.

          • Take a look into the actual safety record of the Space Shuttle and you will see that there are several missions where they scraped through by the skin of their teeth - but they rarely seem to be counted, even when people include “near misses” for Soyuz...

            Theres also the argument that the Shuttle was unsafe because if it was carrying cargo, it had to carry a crew even if the crew is 100% superfluous to the cargos operation - when you take that into account, Soyuz is much better as it has an unman

            • The original shuttle design was for a fully reusable vehicle, but the usual budget cuts reduced it to what was built - with plans for heavy lift unmanned variants that look very much like SLS (except this was in the last 70s!).

              Once the US gave up on manned deep space, NASA never was able to regain a long term focus.

              That said, the shuttle put a lot of people and hardware into orbit and if I ever had the chance to fly on one I would have taken it without hesitation.

    • ... to the analysis-paralysis design approach, where everything has to be perfect

      One might argue that the very idea of SLS (the requirement to reuse STS hardware at any cost) was much less than perfect, so there was definitely no analysis paralysis at the very least in *that* department.

  • From the news snippet:
    "from setting its trajectory to opening individual valves to open and close."

    Well, I am going to patent my own system to _close_ valves to open and close. Or valve opens to open and close. Or to open and valve.
    Makes sense, doesn't it ?

  • If it is it might crash due to Microsoft forcing an update/restart in the middle of the launch sequence -- or maybe it just experiences a BSoD for no apparent reason.

    In all seriousness: points of failure are always directly proportional to the complexity of the machine in question. This subject is far from being an exception.
    • I can see it now. "T minus 10, 9, 8, 7, 6..." *BSOD* *Shoot to external image of rocket launching and careening into the nearby structures.* Bill Gates: "Oops. Did I do that?"

      • Re: (Score:2)

        by PPH ( 736903 )

        "T minus 10, 9, 8 ...."

        "Please enter your Product License Key. It can be found on a sticker affixed to the bottom of the first stage."

        • "This program has encountered a critical error and must be terminated. A useless error message may be found in the system log."
      • Comment removed based on user account deletion

  • Safety is not a luxury that can be downplayed (Score:3)

    by ITRambo ( 1467509 ) on Sunday November 01, 2020 @01:12PM (#60672270)
    There must be lobbyist pressure to keep NASA's top engineers away from the software. This is very shortsighted. Should the software fail the companies that fought to keep NASA out of time consuming safety testing should be forced into bankruptcy. Boeing has already shown that they cannot be trusted to code optimally.

    • Re:Safety is not a luxury that can be downplayed (Score:4, Informative)

      by PPH ( 736903 ) on Sunday November 01, 2020 @08:24PM (#60673394)

      There must be lobbyist pressure to keep NASA's top engineers away from the software.

      Away from the software development process. When I worked therein engineering, there was a push to package up all the requirements and throw them over the wall to a software development group. Which reported to an independent chain of management. And then when I went to work for that software development organization, our goal was to throw those requirements over the wall to an outside contractor.

      It's consultants all the way down.

  • "Boeing builds the rocket's "core stage.". 'Nuf said.

  • Talk to the hand -- with the money. (Score:4, Insightful)

    by fahrbot-bot ( 874524 ) on Sunday November 01, 2020 @01:17PM (#60672296)

    Boeing builds the rocket's "core stage," ... Lockheed Martin builds the Orion spacecraft. Aerojet Rocketdyne and Northrop Grumman are responsible for the RS-25 engines and the side boosters, respectively. And the United Launch Alliance handles the upper stage.

    But NASA's Aerospace Safety Advisory Panel recently said it was concerned about the disjointed way the complicated system was being developed and tested...

    Well ... NASA should talk with the people who authorized the Senate Launch System [competitivespace.org] and why *they* wanted things distributed -- across their constituencies.

    • It's not NASA's fault Congress demands that the funding be spread among as many districts as possible, and as many campaign-contributing contractors as possible to approve stuff like this. (Then don't outright "demand" it but they will be much quicker to approve such projects and the administrators get the message loud and clear.) Congress then has the gall to turn around and ask why its so inefficient. "I'm shocked, shocked I say..." type posturing in committee hearings. Look in the dang mirror (or at
  • Uh, no. The safety panel found issues. The correct response, "We will look into these findings immediately and report our corrective actions."

  • NASA sent human beings to the Moon and brought them safely home on less computing power than a half-assed cell phone.

    Now, it looks like it's only a matter of time before Boeing or one of the other aerospace corporations run the body count up again because they're more interested in cutting in on the competition or cutting corners to do the job they're being very well paid to do.

    The problem here is basically political. Regardless of whether or not it makes sense in terms of a successful mission, every hyena

  • Time for a new ( please ) president to work bipartisanly to end the dreadful waste of money that SLS has become, and allow companies to bid on a next-generation, fully-reusable launcher for NASA to use in their missions.
    We know who will probably win that, but only fair to give Boeing and co a shot at admitting defeat on Shuttle 2, and helping to develop something MODERN ( and cost-effective ).

    If not, events are going to get make it steadily more embarassing for NASA, Boeing and all.

  • Perhaps they can launch the JWST with the SLS, too. You know... when that pork-barrel project is ready.

  • Bad headline (Score:4, Interesting)

    by Zarquon ( 1778 ) on Sunday November 01, 2020 @02:08PM (#60672434)

    The Most Powerful Rocket Ever Flown was the Saturn V (N1 was bigger but didn't fly successfully). SLS has not flown yet, and may be beaten to first launch by the much larger Superheavy/Starship at the rate things are slipping.

  • Non reusable is a wasye (Score:4, Informative)

    by backslashdot ( 95548 ) on Sunday November 01, 2020 @02:52PM (#60672594)

    Why aren't we putting funds towards developing re-usable rockets? This rocket isn't as powerful as the one SpaceX is building, but even worse it's not reusable. Nasa should insist on re-usable rockets, or we should just give Blue Origin or SpaceX funding for their R&D instead of this maximum wasteof moneu

    • This rocket isn't as powerful as the one SpaceX is building, but even worse it's not reusable.

      Balderdash! The main engines are perfectly reusable. It just happens that they're throwing them all away at each launch.

      • The engines might be reusable, but you have to replace most of the part in order to do so. And the cost of inspecting those that remain is as high as the cost of making new ones.

        There is huge stress on a rocket engine. Every piece of metal is pushed to within an inch of its life. It is pretty amazing that you can burn hydrogen and oxygen with a flame temperature much, much higher than the melting point of the rocket metals. It is amazing that it works at all, even once.

        • I wonder if there is a solid business plan in making an engine designed to be easily torn-down completely back to its bare components each time. But also design so that a minimum number of those parts experience wear and tear over a single launch. Such that a vast majority of the parts are expected to pass inspection. Then reassemble them. A happy medium between MTBF of the whole assembly being > 20 launches, and rebuilding and screening each new part each time. Obviously the cost/complexity/reliab

          • Yeah, that still runs into the tens of millions in terms of cost. There is a reason the space shuttle cost a whopping $450 million per launch .. way more expensive than less re-usable rockets.

    • Re: (Score:1)

      by Tablizer ( 95088 )

      Re-usable rockets are still being road-tested. If NASA waits for them to be human-ready, it may end up never happening. Scheduling based on cutting-edge is too big a gamble.

      They should launch the parts separately and rendezvous, using re-usable rockets for the non-crew parts and regular rockets for the crew.

      • Re-usable rockets are still being road-tested. If NASA waits for them to be human-ready, it may end up never happening.

        What are you talking about? SpaceX has already done a manned launch, back in August.

        • Re-usable rockets are still being road-tested. If NASA waits for them to be human-ready, it may end up never happening.

          What are you talking about? SpaceX has already done a manned launch, back in August.

          And then reused that first stage booster that previously lifted the astronauts. Twice. B1058 was the first stage for Crew Dragon, ANASIS-II, and a Starlink launch. It's slated to be reused a fourth time in December for a Space Station resupply cargo flight.

        • Re: (Score:1)

          by Tablizer ( 95088 )

          One launch doesn't really tell us how reliable it is, other than suggesting it's at least around 50%.

          • Good thing then they've already launched a hundred Falcon 9s... Also, I don't remember making any claims about reliability.

  • Say what you will about SpaceX, but other than the payload they built it...

  • Just want to verify (Score:3)

    by LeeLynx ( 6219816 ) on Sunday November 01, 2020 @09:45PM (#60673580)

    NASA's 322-foot rocket "Space Launch System" rocket "would be the most powerful rocket ever flown

    So, this is a rocket, right?

  • The only thing trying to launch SLS is Richard Shelby, dickhead Alabama (backwater America) congressman.

    Everyone else is either playing along or accepting payments, which is pretty much the same thing.

    This puppy won't ever fly.

    You've read it other places first. Now you've read it here.

    E

  • "the software that will control everything the rocket does" , does it happen to be called systemd?

Slashdot Top Deals

Keep up the good work! But please don't ask me to help.

Close