Wednesday, March 26, 2014

F-35 News. A note from a reader...

I got an interesting note from a reader of this blog and I thought I would share it....
Hi Sol,

I read an interesting article today on the F-35 software problems here: https://medium.com/war-is-boring/b6ae0dd4d0ee

For me, this was the first article that really touches on the main problem for the F-35. It isn't the hardware, it's the software. In fact, in the 21st century, most engineering problems are software related and not much to do with hardware.

The article made a misquotes the number of lines of code (LOC) for the F-35 at 8 million, but actually that's just for the battle management software. The total lines of code in the F-35 is approximately: 23 million.

To provide some quick related reference:
Space shuttle: 400K LOC
F-22 Raptor: 1.2 M LOC
Curiosity Rover: 5 M LOC
Boeing 787: 5.5 M LOC
F-35 Lightning II: 23 M LOC
Healthcare.gov: 400M LOC

Why is the F-35 software so much more complex than the F-22's? Mostly it's because of two features: Networking & Collaboration. But it's also due to poor code documentation, bureaucratic management, and multiple vendors. Ignoring whether those two features are useful, whenever these two words appear in a software development project, re-designs and delays occur. Go ask any software project managers with experience in government work and they will tell you the same thing.

But I think everybody knows which aircraft reigns supreme: the F-22. Simplicity really is the ultimate sophistication. The F-35 was born out of politics, its problems were compounded by politics, and its demise will be due to politics. Talk about a vicious cycle.

It was a mistake to close the F-22 assembly line. The correct choice should've been to produce more F-22s with cheaper materials (steel or aluminum vs composites) and sticking with the F/A-18 for carrier aviation until the F-22 could be naval-ized.



Volleyballer out
If the Volleyballer is even halfway right then my suspicion that we might never see this thing work as planned is spot on.

I can't wait for the "supporters" to attempt to refute this info.

15 comments :

  1. Salomão:

    Desde que vi o perfil do F-35 uns anos atrás fiquei imaginando o que significava o "F".
    De pronto imaginei que era "fat" (gordo).
    hehe.
    :-)

    ReplyDelete
  2. The more there is code lines, the more bugs there are...
    F22 has less computing power than a 10 years old PC. (and chips were sold higher than gold price altougth being very deprecated ) so the f22 is unable to handle more code.
    The main problem is that they never thinked about upgradable code. They coded the whole thing has a block, deactivated much think until test.
    at each thing they reactivate, they have to recheck the whole code (otherwise it could have domino effect).
    You see problems coming : the igher is theprogram standard, the longer is code check.
    they will not modify it to make it modulable, upgradable like european fighters, because it s too late now.
    Too many people and firms worked on the code, resulting in a mess in the code. Many of them were certainly contracted because of the ponzy sheme.
    To conclude, has programmer and fighter enthusiast, i think there will be intensive usage of martin baker technology ine the future, if not disabled by ALIS.

    ReplyDelete
  3. According to my knowledge the Saturn V spacecraft had less computing power than an Apple ][. This was sufficient because the hardware was OK. The hardware was OK because the mission was clearly specified: Take a man to the moon and back. Today we more computing power is used to do PowerPoint presentations about (A)/F-35's battlefield awarness...

    ReplyDelete
  4. Lets face it, the DoD and government as a whole have had a terrible track record with software ever since the early '90s. I can't explain the reasons for this as I am not a software expert but the problem isn't exclusive to the F-35. Lets not forget that the F-22 had a lot of software problems earlier in its career. Supposedly bringing some F-22s (Block 10 and earlier) up to the standards of later Blocks has been a very difficult process. There are several ways such problems were supposed to avoided with the F-35. It is supposed to be much easier to upgrade and effort has been made to avoid the computer hardware obsolescence issues that were a source of trouble for the F-22. This is why you have some "technology refreshes" occurring before the aircraft has even reached IOC.

    Yet for whatever reasons a lot of software related things have been screwed up resulting in delays and cost increases. Lockheed also vastly underestimated the amount of code that would have to be written for this project. Yet is the solution to panic, cancel the F-35 and start over? How is that wise? What's to say the next aircraft won't encounter similar problems?

    Start using heavier materials on the F-22 and you start losing the performance advantage it brings. You also cannot afford such an aircraft in anything near the numbers needed to replace our F-16 fleet. The USAF has known this since the F-15.

    Navalizing the F-22? That will require a lot of work and the resulting aircraft would be different enough to warrant its own designation. Take a look at Lockheed's NATF concept art before you go thinking it would be an easy task. The Navy will also want a true multi-role capability similar to what the F/A-18 can do. The F-22's current ability to drop some JDAMs doesn't cut it. Adding these capabilities is going to require larger weapons bays and a lot of new hardware and software. That means the potential for software development problems similar to those encountered with the F-35.

    I loved Northrop's F-23 design and the YF-23 was said to have good low speed handling, yet before somebody suggests a navalized F-23, take a look at Northrop's NATF proposal. That ugly thing has extremely little in common with the sleek F-23.

    Regarding sensor networking and collaboration, that is a very important feature for all sorts of reasons. he same reasons why assets like the AWACS and JSTARS have been so useful to us in the past several wars. This is a natural development of those concepts and improves upon the capabilities of datalinks like Link 16.

    ReplyDelete
    Replies
    1. its not working! the program manager himself has outlined so many risks with this program that i'm seriously wondering how its continuing! the USMC had the plug pulled on the EFV and it was much MUCH closer to success than the F-35 can even dream of.

      quite honestly this 2015 IOC is just a stunt. its pure bullshit and the program manager is all but saying so.

      the craziest thing about all this is that without the "sensor fusion" its just a F-117. it can attack preplanned targets but little else.

      it will be slower, be less maneuverable and not as affordable as planes in production now. i know you like the plane but i'm getting tired of the old...what else are we gonna do line.

      what we can do is extend the life of the Harriers, buy F-18's for the Navy and Marines, F-16's for the USAF and start this shit all over again but this time get it right.

      Lockheed Martin made a ton of money.

      the taxpayer got the big LM dick.

      just let them walk away and call it a day.

      Delete
    2. ASH F-18s and latest block F-16's purchased in modest numbers would be a prudent way to go in ensuring that both Naval tac air and AF tac air have the assets to be capable.

      Delete
  5. The F-22 can be upgrated to share data with other aircrafts.

    www.aviationweek.com/Article.aspx?id=/article-xml/awx_02_25_2014_p0-666721.xml&p=1

    In fact last year re-targed a tomahawk launched from a submarine.

    ReplyDelete
  6. The French Rafale have 2M LOC.
    39 Gripen have 1.5M LOC with fighter link. (Unclear if the number is for C/D or E version.) The Gripen E will have layer and partition separation between flight critical and mission related software.

    ReplyDelete
    Replies
    1. do you happen to know the code number for the Typhoon?
      In the last years, I read so much about the Gripen NG, that I start to fanzy this plane. It just has the best price/performance ratio. And in the end, this seems to be what really counts in the western air forces

      Delete
    2. Their Captor radar have 500k lines of ADA code.
      Typhoon tranche 1 had a total of 1.5M LOC.
      Those are official numbers.

      I've seen a figure of 5M for Tranche 3 but that number is not from BAE.

      Gripen NG might actually be in the sweet spot for a lot of countries though it will not fit all of them. But if you have to choose between a very capable multi role fighter you can afford and sustain over time or a supremacy fighter you can't afford, some say the choice is simple.

      Delete
  7. Both the F-22 and the F-35 both suck. The degrees to which they suck just varies.

    ReplyDelete
  8. F35 Computer cooling use the Fuel of the aircraft, as cooling liquid... I hope the pilot like Sauna when he will serve in Saudia ^^
    A major issue in F22 was the cooling of chips, i don't remember numbers but it was pretty heavy, over 500kg... And now one Iphone is better.
    French Rafale has 2M LOC, with link 16, nuclear link and SPECTRA. Rafale hasn't ALIS, it's a module included in the 2M LOC. It's almost plug and play ^^
    Last year, on the USS Truman, 7 mecanicians changed one engine of the Rafale and relaunched it, in 3 Hours.
    It's possible to make the F35 as efficient, if some people get fired and other managed...

    ReplyDelete
  9. I can understand the frustration Solomon, but lets agree to disagree about the need for the F-35.

    Fabsther using fuel as coolant has been done in the past to different extents. The most notable example I can think of is the F-15. One problem in the early years of F-15 service was that systems were overheating when the F-15s were on the tarmac, engines running, waiting for takeoff. Normally they could get airborne before this became a problem but at a busy airbase there can be quite a wait sometimes. I forget the solution but they changed something that significantly reduced the rate of heat increase.

    I do think it is possible to correct these software issues with the F-35 and make maintenance procedures comparable to the Rafale in most respects, although being a heavier and somewhat larger aircraft it may never quite match that quick turn-around time. If we can fix whatever is wrong with our process that leads to messes like the F-35's software delays and problems, I think we can fix the F-35. If we can't I don't see why our next fighter program wouldn't suffer similar issues.

    Cost overruns aside (we really ought be able to plan a program better than this by now), the two greatest things that could be done to improve the F-35 are correcting the software situation and reducing the weight of the aircraft. If we could bring the weight down we could wouldn't have to reduce those performance specifications, range would be increased, and we could sort out the F-35B bulkhead issue right away. Of course saying this is one thing, doing it is the real challenge.

    ReplyDelete
  10. I have personally seen the F-22 fly several sometimes as I live near an air reserve base.
    I have never seen an F-35 flying or sitting around like some brand new hangar queen.
    That says it all.

    ReplyDelete

Note: Only a member of this blog may post a comment.