There is no spoon: A look into Will Roper’s new digital acquisition reality

As background, we’re all aware U.S. military aviation has been collapsing for nearly 40 years. In the 1950s, over a dozen companies built spiraling supersonic fighters every 2-4 years, called the “Century Series,” during a fiercely competitive period with the Soviet Union. In the 1970s, we still enjoyed 10 companies building fighters every 4 years. Today, we’re down to 3 primes with innovation cycles now decades apart. A vicious circle, comprised of multiple locked-arm acquisition Agents, still drives this alarming trend (as shown previously)…

That was Will Roper’s new paper, There Is No Spoon. Read the whole thing. I think this paper is intended to supersede the Red Pill piece since it covers the same ground and adds some to it.

One thing to note is how Roper’s view of the “digital trinity” has changed over time. Here’s Roper back in 2019:

That trinity is: agile software development—no surprises, there—modular, open-systems architecture—because we want to be able to change out components quickly and seamlessly—and, finally, digital engineering, which is the new element.

And here he is in There Is No Spoon:

(1) own, share, and furnish the tech stack and its reference architecture; (2) warp from the tech stack to edge effortlessly; and (3) eCreate before you Activate.

I think the big thing to note is Roper’s new emphasis on enterprise infrastructure and tooling, as evidenced by CloudOne and PlatformOne. In some ways, this enforces a modular open systems architecture and so is consistent with what he said before. But the emphasis here in the new trinity is toward government ownership of IaaS and PaaS, allowing vendors to focus on SaaS. It’s not clear to me how much farther than that Roper really means when he says “own the tech stack.”

The agile part of the trinity has dropped out. But “warp to the edge” using Kubernetes implies DevSecOps, and that is consistent with agile development. I take that all as wrapped up in the new phraseology. But Roper is again talking about a particular technical solution — Kubernetes as a requirement.

Digital engineering is a common thread, but the emphasis is now on it being a requirement to enter prototyping. I think that’s perhaps a dangerous move for a general policy, but maybe workable for major systems more-or-less within the “state of the art.”

Here’s an explanation of the NGAD strategy:

Firstly, this notional eSeries awards lots of 87 aircraft every 8 years, as well as three ePlane design contracts for each future lot. This increases the rate of competition, lowers the need for industry buy-in by maintaining steady design revenue; lowers the need for industry lock-ins by replacing, instead of upgrading, aircraft; and provides a much healthier Competition Index. It also cuts the average age of the fleet in half.

 

Secondly, it increases RDT&E and production costs by 17 and 28 percent, respectively, while annihilating modernization and sustainment costs by 75 and 48 percent, but keeping total ownership cost equal to the previous (30,30) example. Assumptions obviously matter for these notional percentages.