Hacker News Reader

Comment by johngalt | Hacker News Reader

The F-35 is becoming the definitive example of how scope creep destroys a project. A system designed to do everything for everyone. When it should be designed to do one thing better than everything else.

Divide the F-35 into three different attack aircraft. Single engine CAS with STOVL capability. Big/fast/stealthy twin engine bomber. Carrier based attack drone. All three would likely fulfill their roles better for lower total cost and be operational on a shorter timeframe.

The F-35 reminds me of this short story by A. C. Clarke

http://www.mayofamily.com/RLM/txt_Clarke_Superiority.html

And this scene from pentagon wars:

https://youtu.be/aXQ2lO3ieBA

View on HN

The F-35 is becoming the definitive example of how scope creep destroys a project

Or an example of how we don't learn from history, for the same exact thing happened with the '60s TFX, except the F-35 substitutes S/VTOL for air superiority. 5 models of it were planned, two for the Navy, all but what ended up as the F-111 were canceled, and we resumed building focused planes like the F-14 through F-16, and the A-10 I think. Read the relevant section in Possony, Pournelle, and Kane's classic The Strategy of Technology https://www.jerrypournelle.com/slowchange/Strat.html for lots more details.

ADDED: And I've now recalled that there was stiff criticism of the F-111 as it starting flying and was fielded in the hot war of Vietnam. By and large the bugs were worked out, and e.g. surviving pilots were convinced that they should not turn off the terrain following radar and auto pilot system so they wouldn't follow their predecessors in immediately flying it into a hill. Which didn't happen with the following similar in that way B-1, come to think of it.

That I remember mostly from contemporaneous reports, the frequently poor prosecution of the Vietnam war was a big issue in the heartland.

My dad was a fighter pilot in the USAF at the time the F-111 was supposed to come out. It was late, so he ended up flying the F-100.

McNamara was the genius pushing the F-111. Also the genius behind the Ford Edsel. When he came to visit the base one time, one of the NCOs who was near retirement had an Edsel painted in the camo pattern specified for the F-111. It was parked within clear view so that McNamara could see it when he disembarked from his plane. He was pretty steamed.

That's about what airmen thought of the F-111 and McNamara at the time.

From my reading of the excellent book Clashes: Air Combat over North Vietnam, 1965-1972 (http://www.amazon.com/Clashes-Combat-North-Vietnam-1965-1972...) that would imply your father was fighting over South Vietnam (or perhaps to the west), but, hmmm, it didn't mention F-111 missions at all. For the big operations up north, the Air Force used up its F-105 inventory, and had to switch to their own version of the Navy's (spit) F-4.

Ah, per Wikipedia the F-111 was employed, but I can see why it wasn't covered in book, it was ideally employed as a low level terrain following penetrator, so it would not be a part of the big daytime flocks of birds that the NVA were be able to follow and vector their Migs to for one pass attacks. So not relevant to a book focused on air combat per se.

Heh, love the anecdote about the Edsel. The Official Story, on Wikipedia at least, is that he saved Ford not by pushing it, but by killing it, but I have my doubts, everything he touched after that turned to shit, including his disastrous tenure at the World Bank where he destroyed untold agricultural systems in the Third World....

I haven't read anything, but certainly killing the Edsel could have been his predecessor's idea, but it was too late and the task was saved to give some shine to McNamara. That happens (or seems to) in exec pushouts all the time.

"It was late, so he ended up flying the F-100."

From Wikipedia (https://en.wikipedia.org/wiki/North_American_F-100_Super_Sab...):

Adapted as a fighter bomber, the F-100 was supplanted by the Mach two class F-105 Thunderchief for strike missions over North Vietnam. The F-100 flew extensively over South Vietnam as the air force's primary close air support jet until being replaced by the more efficient subsonic LTV A-7 Corsair II.

And this was mentioned in passing in Clashes.

>>Or an example of how we don't learn from history

Of course we learn from history. The thing you have to understand is that everything is working as intended. F-35, and programs like it, are a way to use public money to create private sector jobs. They aren't meant to be efficient or even effective means to develop and manufacture military hardware.

You can see a direct analogy of this in the space sector, where SpaceX came along and severely disrupted the incumbents by doing things efficiently.

The thing you have to understand is that your initial statement has literally no foundation to it. There isn't any solid proof that program failures like this are direct means to keeping workers employed (where it is implied that they otherwise wouldn't be). I am also not really sure how some crazy guy (I am being hyperbolic) betting hundreds of millions of dollars (some of it his own funds) somehow becomes analogous with a purposefully obtuse military industrial complex scheme to keep people employed.

There are certainly cases where the government soaks up funding, gives it to a company, for absolutely no other reason that we can tell other than to just funnel money along into their little corner of the country (and thus their constituents).

But I have a really hard time believing that is the case here.

It's hard to say with any certainty, to be sure. But I wouldn't be so skeptical -- Nixon did after all announce the Space Shuttle program in part to help secure his re-election. That too in California in a naked attempt to win SoCal.

Source: https://books.google.com/books?id=kje_BAAAQBAJ&lpg=PA12&ots=...

NASA was always a jobs program as well as what it ostensibly was for, it was officially tasked with industrially uplifting the South, which became an even bigger thing when LBJ became president. The area to the immediate northeast of MIT was still mostly a wasteland when I showed up in 1979 because NASA all of a sudden decided that Houston was the ideal place to put the Mission Control Center....

After Apollo, that became many times more what it was for.

You beat me to it. I was going to add something along lines of expanding existing corruption where military people authorizing it get cushy jobs in return over time. The "revolving door." So consistent that it should be our first hypothesis for every program like this.

It is an extremely inefficient jobs program, an extremely inefficient research program, and an extremely inefficient aircraft. As he said: do one thing, and do it well.

It /is/ a pretty good wealth distribution program, though. If you count dollars distributed, not number of recipients.

>>It is an extremely inefficient jobs program, an extremely inefficient research program, and an extremely inefficient aircraft.

Yep. Just like the aircraft: tries to do too many things, ends up sucking at all of them.

I agree, except the problem is that it's sold to the public as defense, not as a jobs program. Part of the sales job is involves having American pilots fly those airplanes into combat.

Decades ago there was the Boeing/Airbus fight over government spending in the aerospace industry.

Boeing was angry that Airbus got cheap credi backed by or from the government, Airbus was angry because Boeing got all the military money from the government.

RAND published a writeup of joint acquisition programs recently: http://www.rand.org/content/dam/rand/pubs/monographs/MG1200/...

Spoiler: they're a bad idea.

RAND didn't mention the commonality between the F/A-18A and the F-117A aircraft.

McDonnell-Douglas significantly upgraded the YF-17 mission systems and cockpit avionics for the F/A-18A/B and it was more advanced than the F-16A version at the time.

The F-117A basically used the same mission avionics and engines as the F/A-18 which was a few years ahead in development.

Having inter-service commonality in major systems and components sometimes a better option than a joint aircraft program. The USAF is upgrading the F-15C (& some export customers) with AN/APG-82 radar, which is basically a larger AESA "antenna" on a AN/APG-79 radar from a Super Hornet.

F-117 commonality was due to the "Skunk Works" model. Lockheed didn't have the time to pursue a custom set of avionics. IIRC, Lockheed also used existing forgings for landing gear for a similar reason.

There's great coverage of all that in: http://www.amazon.com/Skunk-Works-Personal-Memoir-Lockheed/d...

In particularly, they patched together their prototype in a way which all hackers will appreciate.

Avionics and engines can be treated in a modular fashion. Airframes cannot, and that's probably the biggest sin of the F-35.

I'd say that calling the lack of modularity an outright sin glosses over the engineering reality behind why the F-35 variants aren't as common with each other as initially planned.

https://www.youtube.com/watch?v=RQ5gmZ74YBY

Here's Jon Beesley, the pilot who did first flight in the F-35, along with quite a bit of the early air worthiness work. He is discussing the weight reduction efforts associated with what was called SWAT, STOVL Weight ATtack.

Originally the three F-35 variants were supposed to have about 75% of their structures as either common (identical) or "cousin" (largely the same but with some minor differences) parts. In other words, the F-35 was supposed to be modular. Unfortunately, this made all three variants overweight. This matters most for the F-35B, where extra airframe weight directly reduces bring-back weight and cuts into the performance margins of STOVL flight.

Starting at 2:15, he discusses the modularity issue directly (with a bonus discussion on scope creep!) "Holes are really heavy." For every hole, you must have a more robust structure around that hole to maintain overall structural integrity. A modular aircraft meant a heavier aircraft, so heavy that the F-35 was in great danger of being too heavy, missing a key performance parameter for the contract. Eliminating the modularity was the way to make the F-35 meet this performance parameter.

It's all about engineering tradeoffs.

I think it's more complicated than that. If you develop avionics and engines for a single specific craft, then you can be focused on the requirements at hand and a single working version. When you develop them for 3 or 4 crafts all at the same time, the feature creep gets pushed back into those modules as well.

I bet most of the avionics that get adapted for a new plane require some modification to get them working. This is, IMO, much easier if you start from a working version.

Consider the semi-disaster that is JTRS[1] (basically the f-35 of software defined radios). It was designed to be modular, and that was part of the failure. If they had picked a single form-factor that was useful in isolation, that would have allowed a single system to get up and running, and then it would make sense to grow it out from there, reusing where it made sense to.

Instead there was this byzantine connection of interdependencies and parallel development with few integration points that seemed likely to doom it.

1: https://en.wikipedia.org/wiki/Joint_Tactical_Radio_System#Pr...

I think your point is that commonality in the different bins in the bazaar is good; that being said using a cathedral to do everything is bad...

When designing for complex functions think orchestra not one man band.

Lots of standard interchangeable parts that can be individually replaced rather than one "do-it-all" pile of dependencies.

Dr Michael Gilmore, who wrote that report, apparently had not actually seen the aircraft up close until less than a year ago.

https://www.youtube.com/watch?v=9re9tJckTlk

The pilot in this video, LtCol Christine Mau is a F-15E combat veteran is now flying the F-35. The video description incorrectly states the director of the JSF is "General Michael Gilmore". LTG Chris Bogdan is the JSF Executive Officer and has seen the F-35 up close and flown chased flights from an F-16.

For the OT&E Office at the Pentagon, they do not actually fly the jet, or operate the jet, or even touch the jet.

There are a handful of DoD "civilian" pilots who evaluate the aircraft for the Office of Secretary of Defense. None of the DoD pilots have made public statements regarding the F-35 airframe having any major issues.

It is also worth considering that the reason given for merging three aircrafts into one was to save money. I know the F-35 has a lot of defenders, but even they have to admit it is unlikely that the F-35 will ever be seen as a cost saving measure.

I'd defend that statement even though I'm not an F-35 fan. For cost savings, think of what it would have cost to field two different designs (F-35 as F-16 replacement and F-XX as Harrier replacement). VTOL has added cost to F-35, but it's unlikely that it has been so much that you could fund an entirely different airframe as well.

Then you get into the situation where the second aircraft is delayed by one FY because nobody wants to put two of these programs into one budget. Then there's pressure to diverge on engines, avionics, etc. which winds up adding to the cost of the second program.

It is a respunding success according to the amount of porking it generated. Is the f-22 exported?

Edit: to clarify: they're not exporting the good one which could change an actual war.

No exports of the F-22. It's apparently got some technology they're not willing to transfer, even to allies.

The politics are a bit more complicated. The DoD would be willing to export to, eg, the UK & Canada, but any exports at all would trigger calls by lobbying organizations to send it to "allies" like Israel & Saudi Arabia, which then immediately finds its way to China / Pakistan / etc.

I'm half-way suspecting that the F-22 isn't available for export, because it would hurt the sales of the F-35.

The F-22 is not for export because we want to make sure we can shoot down the f35s we do export.

a software kill switch could be a solution to that problem

The French are known for doing that. I'm sure most militaries do in some way

http://spectrum.ieee.org/semiconductors/design/the-hunt-for-...

Good thing we have such competent developers on the project.

Plus if it's three different ones when the world changes and you need 85% carrier drones, 10% stealth bombers and 5% close air support you can build accordingly. Not to mention the fact that one of those can be available when it's done instead of not getting anything until they are all done.

Seriously – this is project management and design 101 stuff here. The points you make highlight the HUGE cost savings that could be accomplished by breaking it up into 3 different aircraft, at different price points, tailored to the different roles.

Sounds like a great idea.

It's not like an adversary could shoot down satellites, jam controller transmissions or hack a drones control plane.

We'd be better off buying aircraft from the French.

Unsurprisingly, it's already divided into three different aircraft - F-35A, F-35B, F-35C. They were supposed to be 70% common, but that's simply not achievable so the differences are quite a bit larger.

I always felt that F-35B at least should have been its own design. As for A & C, there have been successful examples of naval aircraft being used by air forces too - such as F-4, and F/A-18. But trying to have a design that is also STOVL capable?

It is my understanding that the biggest problem with the naval version is that it has only one engine. The F-4 and F/A-18 were both twin engine planes and if one engine failed or was damaged, the plane could still fly and the pilot could land on the aircraft carrier and save the plane. With the F-35C, if the engine is damaged far from the carrier, the pilot will probably have to eject and be rescued. In that case, even if cost is no object there is the inconvenience of going to rescue the pilot and being short one airframe. I also remember reading somewhere that the repair bays on an aircraft carrier can swap an engine in an F/A-18 in just a few hours.

Well, the Navy's been there fairly recently, the single engined A-4 Skyhawk, the F-8 Crusader, and the A-7 Corsair which was based on the latter all saw heavy service in Vietnam. And per Wikipedia, the latter replaced the A-1 Skyraider (to the extent a jet like that could) and the F-100.

But for the obvious reasons they vastly prefer two engined planes, plus I've read the F-35's single engine is so big their current logistics system will have trouble getting new ones to a fleet at see, I assume this means it can't fit in a C-2 or maybe V-22.

By some estimates the parts commonality is now about 25%.

> A. C. Clarke _Superiority_

This fits the Tiger Tank vs. T-33 in WWII. The T-33 had very good firepower and armor. The Tiger was superior in both those regards. However, the T-33 could be more easily produced, dealt better with Russia's muddy seasons and winters, and was far cheaper to maintain.

I think you meant the T34, the second most produced tank in history.

Doh. Off by one! Yes. T-34. There was a recent analysis by the DoD, finding that the logistic costs of modern MBTs scale by the fifth power of their weight. This is a big problem with state of the art modern tanks. Right now, reactive armor is good enough to stymie 105mm and 120mm guns in battle conditions. However, making tanks heavy enough to support 140mm and larger guns looks like it will make them prohibitively expensive.

>Carrier based attack drone

Drones were science fiction back in the early 90's when the F-35 program was started. Even today they can't even come close to outperforming a skilled human pilot in a combat scenario. (Although perhaps if it were Google working on the hardware and software it'd be doable).

> Even today they can't even come close to outperforming a skilled human pilot in a combat scenario.

The Navy's X-47B "Salty Dog" got pretty damn close (before the program was cancelled). It can take off, on-air refuel, and land back on a carrier with no human input.

http://dronecenter.bard.edu/chronicle-of-the-salty-dogs-x-47...

https://www.youtube.com/watch?v=twdCdlo3BtA

> Drones were science fiction back in the early 90's when the F-35 program was started.

AFAICT, the first armed drones deployed by the US military were ASW drone helicopters in the 1960s. Ground attack jet drones were successfully tested in the 1970s, and some preliminary and positive testing on air-to-air combat drones also happened in the 1970s. Drones certainly weren't science-fiction in the early 1990s, even if the military hadn't gotten around to actually fielding armed drones (despite the tests) beyond those old Navy ASW drones.

What if that skilled human pilot was flying a "drone"/unmanned craft rather than sitting in the cockpit? I imagine there are benefits to being "in-person" but the unmanned craft could theoretically endure far more extreme maneuvers that would leave a human pilot unconscious or dead.

Most of that benefit is that you still have something capable of mission decisions when an enemy engages in jamming or other information distortion tactics.

On the positive side for 'drones': they can be viewed as both expendable (with costs/limits) and capable of performing maneuvers that would kill a human.

I can see narrow beam transmissions on multiple carriers and extremely fault resistant communication protocols combining with remotely manned vehicles. The pilots might best be closer for speed of light delay reduction, but realistically you could have a relay in some kind of helicopter type vehicle a short distance from the actual fight, and the expert pilots stationed somewhere even more safe.

While you're at it, that helicopter/blimp/etc could also be your primary communication platform for a local region; it's already a high value target so there's little reason not to use it as an additional uplink.

Just making some small improvements in this process would save the country a ton of money. I cannot believe the cost of the F-35 program has exceeded 1 trillion.

Isn't the F-35's high price due to it already being 3 very different attack aircraft?

1. CAS with STOVL capability (Marine)

2. Light attack fighter (Air Force)

3. Carrier attack plane with bigger wings (Navy)

Creating 3 separate programs isn't going to change that. And, if not that, then the software they want is going raise the prices of the 3 different aircraft anyways. The Marines are probably going to want to develop the expensive Augmented Reality helmet anyways, for example.

If you create 3 separate programs, where will you save money?

Max, Amy, and Ned are driving old cars and want new ones. Rather than buying three cars, they decide to go in together and buy one car to share.

Max wants to replace his offroading jeep, Amy wants to replace her fast sports car, and Ned drives an amphicar because he lives on an island.

So Max Amy and Ned go to GM and say "Design us a floating, offroading, sports car please we are trying to save money."

By the end of the day GM has to convince them to take three variants of the same car, but the cost, time and design tradeoffs needed to reach that point were never worth it. When you could have designed three (possibly four) vehicles separately that each excel in their focus area.

Saying that 'well all three would need gauges and windshields anyway' isn't a compelling reason to combine them.

They don't have the kind of assembly commonality that the program aimed for initially, but I think it's a stretch to call the three variants "very different" aircraft. There's a pretty high amount of systems similarity. And pilots don't have to go through completely separate training programs for the three variants, just separate qualifications for things like STOVL or CV operations. If you can aviate in an F-35A, you can aviate in the other two variants.

Not sure why you single out the USMC for developing the helmet mounted display, because the HMD is common to all variants and is already a working system.

Because in real world costs tend to grow exponentially with complexity.

That scene from Pentagon Wars should be mandatory viewing for every student in systems engineering and project management.

It's the worst case of both scope creep and putting your eggs in one basket.

You could start a brand new Harrier programme right now, same basic design but new build airframes, avionics and engines using modern materials, and you would still beat the F35 not only into-service, but in a stand-up fight if it ever came to that.

Harrier design concept is limited by the massive frontal cross section that Pegasus type engines imposes on the craft - this limits aero efficiency, and especially the ability to go supersonic.

Harrier design is also completely not low observable. An F-35 could splash any harrier derivative from afar, and could probably beat out an Harrier in anything but the lowest altitude, lowest speed confrontations.

F-35 is full of problems, but there's no sense in bringing up ridiculous arguments. Harriers might be able to perform some CAS roles better than a F-35, but it would get completely demolished in any 'stand-up fight'.

Any attempt to turn something based on a Harrier airframe into something as LO and as fast as the F-35 runs most of the same design risks and problems as the F-35.

"Hunh, ComplicatedX is expensive and 20% of the project is causing 80% of the overruns. We should just do SimpleY, we get 80% of the value for 20% of the price!"

...Ten years later...

"SimpleY is no longer Simple, it turns out that although we got 80% of what we wanted for 20% of the price, the last 20% of what we wanted were responsible for 80% of the overruns!"

F-35 could splash any harrier derivative from afar

And an X-wing could take out an F-35, I mean, if we're talking complete hypotheticals.

The Harrier is battle-proven everywhere from the Falklands to Afghanistan. Marry it up with modern radar and air-to-air missiles and you will have a very capable platform - one that the F-35 - if it ever flies in combat - which it won't because no-one will risk those mind-bogglingly expensive airframes in a shooting war - won't see coming because its radar has rebooted again.

I am not convinced than an X-wing would beat an F-35. Do you notice that the cannons on the X-wing have a slow rate of fire, and the beams move slowly through the air. Not to mention the limited missile capabilities.

There's that but the X-wing's repulsorlift gives it superior V/STOL capability to the F-35 remember.

I love the Harrier and agree that a modern one would be a better buy than the F-35.

However, the argument you're making is silly. A Harrier won't beat an F-35 in combat. The F-35 would shoot down the Harrier before the Harrier had even detected it.

The argument you SHOULD be making is one of cost, performance at roles (e.g. CAS, bombing missions, etc), proven for naval operations, extremely flexible, and would likely be delivered sooner (even if they started today, if only one manufacturer was involved, it will likely ship before the naval variant of the F-35 in 2022 or whatever they're guesstimating now).

The core problem with the F-35 and F-22 is that they're the most advanced air to air combat aircraft ever built, so advanced in fact that they have nothing to fight, and worse still will arrive right before swarms of drones take over from manned aircraft (e.g. 10x drones Vs. 1x F-35, the F-35 would literally run out of munitions).

Please tell me what a modern Harrier would look like. Here is an (admittedly contrived) list of some design goals:

* Greater STOVL performance, without the limitations imposed by needing to carry water for thrust augmentation

* Low observability (certainly lower than a Harrier, which has a big RCS)

* Greater sensor capabilities. Built-in electro-optical targeting so that radar can be used less, or not at all

* Reduce pilot workload by filtering information presented in an intelligent way

* Reduce STOVL cognitive errors created by differing control schemes between STOVL flight and conventional flight

This sounds an awful lot like an F-35B.

My point is, saying you want a "modern Harrier" isn't enough. Because in many important ways the F-35B is a "modern Harrier". You have to figure out how whatever it is you're going to build will be different, if you hope to avoid the same problems.

>> The Harrier is battle-proven everywhere from the Falklands to Afghanistan

This doesn't translate into a tactical capability today: Falklands: Harriers fighting Mirage III, which were 20 years old, 34 years ago. Afghanistan: dropping dumb bombs on adversaries who didn't have a SAM capability

Isn't that the point, though? Who are these jets going to fight -- pretty much any realistic hypothesis of conflict today would consist of asymmetrical warfare, especially if waged by the US or US allies. Asymmetrical warfare with existing weaponry, let alone next-generation planes. The only scenario where the US would face significant resistance is vs China or Russia, where planes wouldn't even get the chance to take off before Midnight is upon us.

Given the situation, it would make more sense to invest in something cheaper that will do the job just as well.

Harrier design concept is limited by the massive frontal cross section

The same criticism has been levied on the F-35, due to the S/VTOL airframe support requirement.

....I assume your design would be just as fast as a F-35 and also just as stealthy?

True. At this point they might as well build a Space Shuttle with guns.

Reply on news.ycombinator.com