(CNN) Senior US general warns China’s military progress is ‘stunning’ as US is hampered by ‘brutal’ bureaucracy. Several considerations:
- Artisanal space hardware
- Secret Equations
Yes, it is brutal. In one small company, there were four workers,two principals, and one person whose specialty, with arduous academic training, was “managing” a DoD project. That’s 43% overhead. Yet they couldn’t figure out how to buy a $30 book I needed, so I bought it myself.
DoD is hard on big companies too. In the same epoch, salaried management at Lockheed-Martin were required to attend interminable after-hours meetings with the brass. At these meetings it was understood that even if their presence wasn’t topical, they had to show face — at the cost of a decent life, family, and sleep.
(CNN)Member of CIA chief’s team reported Havana syndrome symptoms on recent trip to India describes a $5 instrument that, had it been deployed, would have made a big contribution to solving the riddle. Someone close to the problem may have thought of it, deterred by the project overhead of something so intrinsically cheap.
Some programs under DoD management nevertheless get away with murder. The Littoral combat ship program is marked by instances of what some might call fraud, such as fixes to the propulsion systems without telling the Navy what those fixes were. DoD management requirements drowned management in paper, inviting fraud instead of preventing it. Six ships, with an average age of only 8 years, LCS-1, LCS-2, LCS-3, LCS-4, LCS-7, and LCS-9 have been or will be decommissioned as unserviceable. This is not William Proxmire’s $600 toilet seat. What did all that onerous bureaucracy buy the taxpayer?
So DoD bureaucracy cuts both ways. To address General Hyten’s concern, we need a compact goal. So here’s a proposition:
- Establish a class of project that can be managed by a competent manager who has not received a special education in DoD projects.
- Insulate the class from DoD intrusion of the onerous kind. Replace short review intervals with longer ones, and trust (see below.)
- It will probably involve research and prototyping, not large scale production.
- Depending upon the outcome, that manager receives a “trust rating” that facilitates further management opportunities.
- Skunk Works may provide some inspiration.
Artisanal Space Hardware
Hyten pointed to the development of hypersonic weapons to highlight the stark difference in approaches by the US and China. He said the US has carried out nine hypersonic tests in around the last five years while the “Chinese have done hundreds.”
It is concerning, though the blow softens when we consider that the U.S. approach to hardware has always been artisanal, relying on finely crafted rockets, sparse, slow rate test shots, and detailed telemetry to get as much data as possible from those shots. In contrast, at least some of the numerous China shots appeared to have very primitive telemetry.
If hundreds of tests are the way to go, a cheap, non-artisanal booster is needed. U.S. rockets are artisanal for a reason: to get the biggest boost with reliability appropriate to the payload, in a process of extremely violent, yet controlled combustion. Though printed parts present new possibilities, a non-artisanal booster that is safe to be around is not a small job.
Although secret equations are usually found in grade B movies, they may actually exist. The public version has a big name, Navier-Stokes. And they have a big problem. Other than a few special cases, nobody knows if they have solutions. And they are critical to design of hypersonic bodies. The Clay Institute offers a $1M prize if you can prove they do or they don’t.
Paradox: Even though nobody knows if they have solutions, they are used in computer simulation of hypersonic flight. The sims give solutions, the equations do not, so what gives? It turns out that the sims blow up, go to infinity, unless the numbers are massaged. This is quicksand.
A while back, a Chinese graduate student in Beijing managed to simulate Navier-Stokes on a primitive AMD parallel processing platform called Close to Metal. This was very impressive, since successful use of CTM, which had no debugging facilities, was very rare.
Navier-Stokes brings the most powerful supercomputers to their knees, so slow they seem like last year’s smartphone. This is why test shots are so important. There has been some progress; some people in the Netherlands figured out how to make the sims more efficient.
What if the Chinese have solved the Clay Institute Millennium Challenge? The solution could enable simulations that run more efficiently and accurately. As with the U.S. nuclear capability, it could eventually result in the ability to design a hypersonic body with minimal testing.
Good hunting, gentlemen.