02 September 2009

The Cult of FIST

One of the main rebuttals to our FIST concept is that it is incapable of producing the kinds of capabilities seen in weapon systems like the F-22 or an aircraft carrier. The argument is that these kinds of capabilities can only be produced by programs that are methodical, deliberate, systematic and massive. We often hear that FIST programs are cute and may be fine to fill stop gaps, but aren't sufficient to do the heavy lifting of developing and fielding "real" weapon systems.

But that's our point exactly. Weapon systems with tons of capabilities and features are overrated, if not useless, in the current state of warfare. We postulate that what the warfighter's really want and need is something that is Good Enough. Turns out defense isn't the only industry where this is true. WIRED just published an amazing article which could easily be our manifesto for FIST, The Good Enough Revolution: When Cheap and Simple is Just Fine. An excerpt:

Military aircraft are experiencing their own version of the MP3 effect.

Why, if manned planes are so superior, is the Predator saturating the combat market?
[Their] ability to maintain a constant presence in the air. That's because the drones are relatively cheap to build, can fly for more than 20 hours straight, and don't require pilots who need sleep, food, and bathroom breaks (and who might die if the plane is shot down).

Piloted aircraft are still valuable......but because the Predator can linger, it has enabled a new type of strategy—remotely guided surgical strikes with fewer troops and armaments. It's a lesson that surprised the Air Force and other services, Mathewson says, but one that has been learned definitively.


RhetTbull said...

I'm a huge fan of Good Enough and think the FIST approach is dead on for many systems. But I do worry that the Cult of FIST preaches that FIST is the only viable solution for all systems in all cases. There are military capabilities for which no FISTy solution exists (at least not with today's technology). That shouldn't mean that we don't try to build those solutions. We should limit the requirements set and use FISTy acquisition principles wherever we can to minimize schedule and cost and maximize delivery, but we shouldn't fear complexity if the only solution we can come up happens to be complex. Complexity may not be desirable, but it's not evil.

And can we please stop comparing manned fighters to UAVs? They don't do the same thing. Sure, UAVs are used today to do a lot of things that fighters used to be used for and that is goodness--we can use fewer fighters and also use those fighters to do other things (that UAVs can't yet do). But UAVs are only useful where they can fly without getting shot down. In many parts of the world, no UAV would last 30 seconds in the air until the fighters killed all the things that are trying to kill the UAVs. I fear that the UAV advocates have taken air superiority for granted and when we have to re-learn that lesson, it will come at great cost in blood & treasure.

The Dan Ward said...

You're dead on, Rhet. There are some nuances & caveats to the FIST thing that sometimes aren't as front-and-center as they should be.

I try to always point out that there's "no guarantee, no monopoly" to this approach. That is, FIST projects can definitely fail, and they're not the only way to succeed.

And while elegant simplicity is optimal & desirable, complexity is not necessarily bad. It may simply indicate that there's still some maturity needed in the organization / technology / process. The key in applying the Simplicity Cycle is to assess the project's vector, not just its position on the complexity scale.

And you're probably right about the UAV's vs Manned Fighters. However, I think the manned fighter paradigm's days are numbered, but not because a single UAV can outperform or outsurivve an F-22. It's because a swarm of 1000 UAV's would out perform 1 F-22, and we can probably get 1000 UAV's for the cost of one Raptor.

That's the thing about FISTy systems - they're inexpensive enough to be more expendable than our precious expensive systems. Sure, a bunch would get shot down, but we'd have the element of Mass on our side, and with no person in the cockpit, losses would be fairly easy to accept.

RhetTbull said...

Dan, good thoughts! My comment about not comparing the F-22 and UAVs was not meant to imply that manned fighters are here forever. I agree that their days are numbered. But almost all the comparisons running rampant in the blogosphere compare today's UAVS to today's fighters and that's not a fair comparison. I think the F-35 will be the last manned fighter we ever make. But we'll probably still be flying it in 20 years. Why? Because the technology to replace it won't be mature enough until then. (Personally, I'd cancel the F-35 and buy a bunch of F-18s equipped with low observable (LO) and electronic warfare (EW) capabilities to tide us over until then since they would be cheaper and we buy more of them)

Your vision of a swarm of cheap, disposable FISTy UAVs to kill an enemy IADS is great and I've seen several proposals for this kind of thing. But it's not as simple as at sounds and there are still a lot of complex problems to solve before we can implement it. Let's take a quick look at what's involved to bring the idea to life against today's surface to air missile systems (SAMs) (not an in-depth analysis--brainstorming as write):

-Lots of UAVs means problems of discrimination for the SAM (referred to as the shooter from now on)...if you give him too many radar tracks, he doesn't know how to prioritize them or what to shoot first
-If you can fly the UAVs close enough, you can get inside his track algorithm so he can't tell there are 2 or more tracks...requires really good flight controls and very precise position, navigation and timing (PNT), also requires the UAVs to be networked
--What if the shooter jams your PNT?
-You can use jammers to make the mini-UAVs look much more menacing than they are...holy cow is that 1000 F-35s coming at me!? (ref MALD-J)
--Shooter can counter this with better discrimination algorithms
-What makes the manned fighter (and today's UAVs!) so good today is the man (or woman!)...mainly his/her brain and training...it's better than any flight control and artificial algorithm (AI) we've been able to come up with...today's UAVs are not unmanned, they're remotely-manned
--The mini-UAVs need to be autonomous not remotely-manned, we can't rely on having comm links...too much bandwidth required and too easy to jam
--To make the UAVs that good will require quantum leaps in AI technology, not FISTy yet
--Some really cool research being done into hive intelligence looking at bees, how birds and fish maneuver so close without hitting each other, I saw a story this week about a nano-UAV that can withstand bumping into things--designed by studying flies bumping into windows, etc -- I think the AI tech will come from copying nature but not humans...let's copy simpler systems like bees, flies, birds--but the tech is not FISTy yet
-We have to get the delivery platform close enough to deliver the mini UAVs (modern SAMs have incredible ranges)...how do we do that? What's the delivery platform? (airplane? sub? stealthy ship? missile?) How close do we need to get?
-Are the mini-UAVs there just to distract the shooter, to kill the shooter (how?), to make the shooter use all his missiles?
-Don't forget the shooter gets a vote too...he can invest in better algorithms, better radars, more missile batteries and all sorts of FISTy things himself

I think it's a great idea and we should be funding FISTy ways to do what big, expensive, complex airplanes do today. But until then, I'm glad we have those big, expensive, complex airplanes. As an AF acquirer, it's my job to make sure the President can project force when and where he wants to. There are a lot of places that, with today's technology, would require an F-22 or something similar in order to do that. That won't always be the case but don't forget that that is the reality today. We built it, we have it, let's use it if we need to. And let's think about what to build next...I'm pretty sure it won't look like an F-22.

The Dan Ward said...

It's always fun to discuss this stuff with you, Rhet. Nicely done, and thanks for the summary of what's involved to do this UAV swarm thing. Definitely a future (way-future) capability, and plenty of work still to do there.

I think the only point where we diverge is on the value of the current big, expensive, complex airplanes (I'm thinking of the F-22 & F-35 in particular). I wish we had tried "FISTy ways to do what big, expensive, complex airplanes do today," such as the idea of buying a bunch of F-18's with LO & EW that you mentioned.

The thing is, we can develop the necessary capabilities and completely dominate the battlefield without spending so much time and money, and without relying on such high levels of complexity. Sure, spending $65B and 20+ years is one way to deliver an F-22... but like you said, maybe what we really need is an upgraded F-18 or F-15, ready in 3 years... and then do it again 3 years from now, etc.

What I'm fighting against is the belief in the inevitability of complexity and high costs. That road may get us there, but there are better roads to take.