Just before Christmas, the Center for a New American Security (CNAS) put out an interesting report Process over platforms: a paradigm shift in acquisition through advanced manufacturing (PDF) by Aaron Martin and (expat Australian) Ben FitzGerald. It looks at the potential impact on the acquisition of military capability of new and emerging technologies such as 3D printing (additive manufacturing) and robotics.

The authors recommend a move towards systems that are more mission specific (less multi-role) and which have a shorter overall lifetime, to allow for more rapid technological refresh of fielded capabilities. In this they echo the views in one of my favourite papers on defence acquisition (PDF) which argues for more R2D2 type systems and fewer Death Stars:

After watching the climactic battle scene in Return of the Jedi for the first time, my 8-year-old daughter said, 'They shouldn’t build those Death Stars anymore. They keep getting blown up'. She may be a little short for a stormtrooper, but the kid’s got a point.

Yes, the Empire should stop building Death Stars. It turns out the DoD shouldn’t build them either, metaphorically speaking. What sort of system fits into this category? I’ll resist the urge to give specific examples and instead will simply point out that any enormous project that is brain-meltingly complex, ravenously consumes resources, and aims to deliver an Undefeatable Ultimate Weapon is well on its way to becoming a Death Star, and that’s not a good thing.

The CNAS report points out that existing advanced platforms (such as the F-35 Joint Strike Fighter) are extremely resource intensive to develop and produce. This not only makes them expensive to procure, but it means that it can take years for a single item to move from initial parts fabrication to final assembly, which limits the ability of the United States to rapidly boost its numbers in the field.  It also has the effect of making any sort of 'war of attrition' deeply unappealing.

We've had a number of posts here on the Strategist about the decline in the size and the ageing of the USAF inventory. Both of those issues have important implications for the future of western air power. That'll be especially true as the technological advantages on which it's currently based erode—a process that's already underway and which can only be exacerbated by keeping platforms in inventory longer.

But the problem also exists for sea power. Analysts have started to worry about the impact of the higher than expected (though not by everyone) cost of the new Ford class carriers on America's ability to maintain the size of its overall fleet. Aircraft carriers have proven to be a tremendously powerful and versatile tool of hard power for more than half a century (even if other CNAS researchers worry that they might be looking a bit more like Death Stars these days). But if they become unsustainably expensive to acquire and operate, the net benefit won't be there.

That said, a recent statistical analysis of 422 US acquisition projects from 1970 to today suggests that the problem is worse for aircraft than for other systems. The table below shows the statistically significant trends over that period. (Blank cells had no measurable trends.)

	Army (n= 97)	Navy (n=146)	Air Force (n=179)
Total cost growth			upward
Work content growth			upward
Costs-over-target	downward
Schedule growth		downward
Duration			upward

Those trends are hurting the ability of the US to acquire combat aircraft. Perhaps the most graphic example of the effect of 'silver bullet' solutions for aerial platforms is shown in the chart below from the new CNAS study. It compares the acquisition parameters (development time and cost, production time and unit costs) with the numbers acquired for two USAF fighter types—the F-4 Phantom, which was the backbone of American air power for most of the 1970s and played a more than useful role for years after that, and the F-22 Raptor, generally agreed to be the best air-to-air platform in the world today. (All prices expressed in 2013 USD.)

* USAF acquisition only. Total F-4 production was 5,195. (Figure reproduced with kind permission of CNAS.)

The stark difference in numbers reflect a number of factors, including wartime imperatives in the late 1960s and a Cold War willingness to spend more on defence. But it also reflects the opportunity cost of each platform to other aspects of defence and national spending.

There's no gainsaying that the F-22 has vastly increased unit capability over the F-4. But in air combat, numbers really do matter. I argued last year that 'success in any future air campaign will be about numbers, persistence and technical quality—probably in that order'. Being able to sustain numbers in the field will remain important, which is where a fleet of long endurance drones makes sense. While there are still good reasons to think that a robotic revolution in air combat is still a little way off, the trend seems clear. The current approach of relying on increasingly complex, expensive and therefore relatively few platforms doesn't look to be a winning strategy into the 21st century. A 'more R2D2s' approach based on clever new technologies looks to be a better bet.

Andrew Davies is senior analyst for defence capability at ASPI and executive editor of The Strategist.