F-35: Death spiral is closing in

Yes before you just say “happened with every fighter, they’ll get there”, read this article:

It makes very very clear that the main delay in the F-35 at this point is actually the software. Now I don’t know anything about fighter development, but I do know something about software development, and I can recognise the symptoms of a software project in a death spiral. The airforce’s and marine’s declaration of initial operational capability is a joke, the F-35 will not actually be usable in combat until the delivery of the Block3F software. That’s currently seven years behind schedule and not expected to be delivered until 2018 at earliest (thats optimistic in my opinion).

This is not the same as previous fighters were the kinks were worked out eventually, because no other plane is dependent to anywhere near such a degree on extremely complex and never before achieved software. Yes the F-22 has some sensor fusion but not anywhere near to the level they are aiming for in the F-35. It’s entirely possible that the features that are needed in BlockF are simply not possible on the current embedded computers, they underestimated the spec and need considerably faster more powerful computers to achieve what they have promised. On a plane you can’t just pop in a faster processors and some extra ram, more powerful computers means more space, heat, power and weight, all of which are already constrained on the F-35. Plus every chip used for aerospace needs to go through an approval process that takes years or even decades. Maybe they’ve already had to respec the embedded computers, that would explain the incredibly long delays which otherwise seem hard to understand.

Whats more, the article above also mentions they don’t have the verification simulator to check fixes on yet and won’t have it until at least 2018. That means every fix has to instead be tested on an actual F-35 in the air, slowing down the process even more.

Lastly and most worrying, it makes clear that Lockheed Martin and Pratt and Whitney are asking for a larger batch of 456 F-35’s to ramp up from initial production to full rate but Congress so far won’t approve such a large order until they see better results, but they’re not approving the funding needed to accelerate operational testing either. That’s how the death spiral starts. It’s behind in schedule, so order numbers are cut and production ramp up is delayed which then delays the schedule even more. Rinse and repeat.

You can say it will all be fixed, and yes that would be true eventually if money and time were unlimited and politics wasn’t involved. In reality at a certain point overseas purchasers will pull out and it will limp along for years more because no one is willing to pull the plug. Block 3F will probably eventually be delivered (around 2021 at this rate) but it will be achieved by fudging the requirements during operational testing like they have done so far with the current milestones at every stage.

Whatever happens by the time the F-35 is actually operational, it will be largely obsolete, due to advances in drones and radar. Chuck enough cheap drone’s into the air in a F-35’s path, once they get close enough they can detect the F-35 despite it’s stealth and the F-35 doesn’t have anywhere near enough ammo to shoot them all down.

Assuming the costs are as you describe them, you have actually made no argument against continuing to absorb those costs. This is because you show yourself to be so unbelievably ignorant concerning the capabilities of drones versus the F-35. Further, you ignored the drone-killing drones which has been the topic of discussion from people talking out their ass for months now.

You haven’t read the article, read it and and then read this:

That’s the DOT&E’s own report on the current capabilities. Then read the first link again. The argument for not continuing to absorb those costs is what I have laid out:

  1. You might never get there, there is no guarantee the goals of Block3F can actually be achieved, no matter how much money you throw at it. Read the descriptions of the current problems with the sensor fusion systems in the first article.
  2. By the time you do achieve Block3F it will be partly or completely obsolete, as stated above.

Then read the memo again, it confirms they are reducing staffing on testing at exactly the time they need to be increasing it, because congress won’t approve ramping up to full production. That’s a death spiral.

While the facts raised by the OP (e.g., delay in software, impacts on capabilities and cost, etc) are generally correct, the conclusions drawn from them make it clear that the OP is completely out of touch with reality.

Since it isn’t even worth going point by point to refute the conclusions – since they are as poorly thought out as someone suggesting that Microsoft has already invented Skynet so humanity is at risk – let me confine my remarks to just a couple points.

  1. There is absolutely no chance that the United States is going to substantially reduce its F-35 buys. The Navy may slow F-35 procurement over the next decade or so, but it has no choice but to keep buying them. F-18s are not a long-term solution, and the F/A-XX (google it, OP) is going to take another 20 years before it arrives on flight decks.

  2. Along with 1, Congress has actually been ADDING F-35s to recent budgets, not cutting them. This is the exact friggin’ opposite of a death spiral. US quantities are going to continue to grow, but at a much slower pace than what Lockheed would prefer.

  3. Foreign sales are not as clear-cut as the US commitment to the program, but on balance, the popularity of the F-35 is growing. Sure, Canada has cold feet because the Prime Minister has about as much clue about tactical aircraft as the OP, but Israel, South Korea, and Japan are joining the buys.

  4. Software is a problem, period. But as aviation fanboys fall all over themselves to declare the F-35 a piece of junk and the F-22 the best thing ever, the lack of substantive knowledge of these programs indicates that most people have no clue how much capability was left out of the F-22, and deferred to future software upgrades.

Here is what is going to happen: work on Block 3F is actually quite mature, but as the jets enter OT&E there’s going to be a lot of problems discovered. Production of the F-35 is going to continue to grow at a slow rate for several years, until either the Pentagon gets a big budget bump (who knows when?) or the cost of maintaining 4th gen fighters as they go through depot starts to become overwhelming (which is exactly what is happening with the Marine Corps legacy Hornet fleet at this exact moment). The F-35 may never meet 100% of the promised capabilities, but all other options – for the US and foreign partners – is to meet only 60% of the desired capabilities by buying more fourth gen fighters.

The idea that drones are going to replace tactical aircraft like the F-35 in the next twenty years is ignorant nonsense. Might as well say that Special Forces are going to start shooting laser beams out of their eyes.

You don’t address the crucial points I’m making and why it’s a death spiral.

  1. congress will not yet approve ramping up to full production because the F-35 is massively behind schedule and has delivered poor results.
  2. Because of this they are shedding testing staff at exactly the time they need faster testing (this is in the memo, it’s a fact). This further delays the schedule
  3. The verification simulator is so far behind schedule that it’s been taken away from it’s contractor and given to a Navy lab, who never planned to make the thing. This further delays the schedule.
  4. congress won’t approve ramping up production because it’s behind schedule and has delivered poor results… repeat… schedule slips again… death spiral. Your claim that congress has to approve ramping up in future regardless of the capabilities demonstrated is nothing but an article of faith.

So i’d like a cite please that work on Block3F is quite mature, that is directly contradicted by the DOT&E memo I linked above.

You’re missing important facts and context. The only significant importance of a full rate production decision – which has not been proposed yet – is to allow for multi-year procurement contracts that generally cut the cost of production items by roughly 10%. That’s pretty much it.

You are laboring under the perception that, say, the Air Force would have a significant increase in production rate following a full-rate production decision. YOU ARE WRONG. The Air Force’s five year budget has them buying ~45 F-35s a year under a full-rate production decision, which is the rate that they are currently buying them.

I’m very serious about this: I will bet you any amount of money – five dollars, a hundred dollars, ten thousand dollars – that Congress will not cut planned F-35 production quantities in the next two years. I will additionally bet you any amount of money at 2:1 odds that Congress ADDS F-35s over the next two years (if they don’t, I pay you $1, if they do, you pay me $2, etc).

Want to bet? You name the stakes. Seriously, I’ll bet my entire bank account on this.

It is a fact that test staff are being rolled off, as SDD (also called EMD) is concluding and OTE ramps up. I totally agree this is not wise. But testing should NEVER be accelerated; this is just a stupid idea. Time is more valuable right now than trying to squeeze in more test points on particular flights.

Once again, name your bet. I’m willing to put tens of thousands of dollars on this, or just a few bucks to prove my point. Put your money where your mouth is.

The development is mature, the testing is not. “Mission Systems Block 3F software development is 98 percent complete.” As I said before, there are going to be problems found in 3F, there is no question at all, especially as OTE ramps up. But the Marine Corps is planning to install 3F on its Bs in the next year, and there will be bugs to be worked out, no question. You just seem to be trying to make a point that 3F is vaporware. That’s wrong.

Your source is Lockheed Martin’s claims? Seriously? When the project is 10 years behind schedule and massively over budget? The memo I linked from the government’s independent assessment office indicates clearly that Lockheed Martin has repeatedly over estimated their current progress, eg they’ve lied. Why should we believe them this time?

Even if they are not lying about their progress it’s meaningless to say only 10,000 lines of code remain to be written, because getting those last 10,000 correct might require rewriting another 3 million and take 5 or 10 years, that’s how software works, the last 5% takes more than 70 percent of the time and effort. See page 12 of the linked memo for the current deficiencies in Block3F. And they are still discovering deficiencies in Block 3i, also on page 12.

So yeah, the capabilities of the F-35 right now is the capabilities that Block3i has, incapable of being deployed in combat. The claims of the capabilities that will be delivered for the F-35 are at the moment, just that, claims by a company that has repeatedly overestimated it’s capability to deliver.

Now I never said it’s impossible for the F-35 to recover from it’s current death spiral, that’s possible, but the odds are against it and dropping every time the schedule slips again.

Not willing to back up your predictions with even a dollar, eh?

Actual F-35 test pilot response:

Yeah, I was going to link to that. Same guy I think was linked to in the last OMG, FG35 SUXES!! thread. The OP should really click the link to get some perspective. I think the key points are:

Nice article but it still doesn’t address my point. The definition of a death spiral in software is a project that is slipping further and further behind schedule, and blowing out in cost so much that resources begin to be diverted away from it, and necessary testing and full scale deployment is delayed more and more, making the problem worse, repeat.

The F-35 certainly meets that definition, so it is currently in a death spiral. That’s just fact.

What it has to do to pull out of the death spiral:

  1. get the verification simulator working
  2. get sufficient funding to do the testing they need (which they don’t currently have, see the memo)
  3. demonstrate enough capability that the needed resources are fully allocated to the project.
  4. ramp up to full production

(1) will certainly happen at some point, but until it does the F-35 will continue to slip further behind schedule, eg the death spiral will continue.
(2) and (3) may or not ever happen, and if they don’t then (4) will never happen.

And answer this very good question, if Lockheed Martin is so certain they can deliver Block3F by 2018 why are they not funding the necessary testing themselves? They have ample cash reserves and resources to do that but they are not? Why?

Again, how much money do you want to put on your prediction here? I’ve offered you quite generous terms, I think.

There is no death spiral here. Congress is going to keep funding the program, and I’m willing to put up tens of thousands of dollars to back up my prediction. Are you?

Dude, for like the third time, the current rate of production is basically the same as what it will be under full rate production in a few years. Lockheed is building roughly 120 jets a year today (including international partners) and under full rate production, they will add maybe 5, 10, at most 15 jets per year.

The only real significance of full rate production is the ability of the U.S. government to sign a multi-year production contract that by law needs to show roughly a 10 percent decrease in unit price. You continue to imply that full rate production will mean a significant increase in production, and that’s just not what is going to happen in the near- to medium-term.

Because that isn’t the deal signed with the government. It’s a really lame question. My plumber has “ample cash reserves,” but that doesn’t mean that they come fix my toilet for free.

ETA: besides, 3F is going to “deliver” in the next year or so, but again – testing is going to show problems that are going to need correction. Meanwhile, work on Block 4 is ramping up. Are you aware of that?

That the F-35 is unflyable without complex software in control tells me one thing: that any kind of powerful EMP is going to be their Achilles’ heel. Yes, yes, super-hardened avionics and fiber connectivity and all that brilliant design.

And one port seal gone bad, one maintenance hatch not properly secured, one stray leakage path from any cause, and an EMP pulse will turn it into a $300M rock.

I guess all other aircraft benefit from not having computers in them. Touche.

Most of them remain flyable when the computers fuck up. It sounds like you can’t adjust the seat headrest in a 35 without an onboard supercomputer cranking.

This is, of course, the pattern for all inter-war development: years and decades of increasingly complicated and touchy development that goes south/to shit/to the boneyard the minute things get hot, to be replaced by far more sensibly engineered weapons that actually keep working. The F-35 has managed to surpass the B-1 and the Abrams tank as an example of something that any fool watching the parade can see is going to have about a one-week lifespan in anything but the very controlled wars of the last forty years.

Most? So, which US combat aircraft remain operational through an EMP blast?

Not the F-35, which is the topic here. It’s just another and yet-more-extreme case of “if we engineer and test the mofo’in’ shit out of it, it has to work in the rule-free chaotic madness of war.”

Should have just called it the Maginot and been done with it.

I see. And your knowledge of general DoD HEMP protection standards exceeds our esteemed Stranger on a Train in which specific ways?

And your thoughts on the plan for the F-35 be a nuclear capable aircraft – I guess nobody thought about the airplane dropping nuclear bombs when they were designing the electronics?

It’s relevant whether a vulnerability portrayed as fatal to one a/c applies to other a/c, cannot be dismissed with ‘but we’re talking about this a/c’. No modern a/c is combat effective with its computers knocked out, and some a/c of the previous generation are also unflyable in that case, for example the most widely used Western fighter of the previous generation, the F-16.

The issues about gaining full software capability (Block 3F and future 4) are more relevant for a targeted discussion of one a/c, though even that discussion has bear some reference to what else one would do, and not hand waving like ‘drones’.

I don’t need to put up money to post my opinion on this board, facts speak for themselves. Your plumber analogy is really lame, actually if your plumber messes up and fails to deliver a project on time, then yes they put the necessary hours in for free, or you sue them. Why shouldn’t Lockheed Martin take responsibility for their failure to deliver?

There is clearly a death spiral, because the sources I have quoted state quite clearly that there is no verification simulator, that there is no time scale for it’s delivery and the lack of that is going to keep delaying delivery further and further. Second, that necessary resources are being cut from the project, quoted on page 12 of the memo.

Do you deny either of these?

If you accept these, then that’s the definition of a death spiral, and whatever one test pilot thinks is completely irrelevant, and whatever you think is completely irrelevant. The F-35 is in a software death spiral which it may recover from, but so far it’s not meeting the goals it needs to do that.