[FoRK] Losing my religion
jbone at place.org
Sat Feb 13 15:44:22 PST 2010
Look, this only takes common sense to figure out. NASA has created a garden industry of highly-specialized contractor firms that exist to design, build, maintain, and service specific subsystems or provide highly specialized services. IT has always been the central planner and coordinator. The money --- government dollars allocated to private-sector solutions to the problem --- was always out there, low-hanging fruit, for a *high degree of specialization.* No incentives existed for any one firm or firms to individually acquire all the competencies necessary to do the whole enchilada. Even the private sector launch companies you mention don't do the whole enchilada themselves: the rely heavily on the services of NASA, the military, and others to get it done. By limiting the profitable scope of what can be done privately, by providing barely-adequate services at below cost, and by making high specialization the most-profitable activity, NASA's mere existence has *inevitably* supported the traditional aerospace players while discouraging efforts like e.g. Musk's.
Until NASA started looking *really* flaky, that is. At which point, things like Musk's effort started to look reasonable.
Getting to orbit *is* expensive; even moreso with the waste and inefficiency of the existing system. (Aside: the thing that annoyed me the most about the X-Prize was just that: if it doesn't achieve orbit, it's a freakin' airplane. The techniques necessary to "win" the X-Prize are virtually useless for any real commercial use of space. Some of the newer challenges are much more reasonable. And in fact, I think the prize model is a reasonable way for e.g. government to use public funds to encourage significant R&D and technical achievement. Been done before. Works well. Encourages massive parallelism, especially compared to singular government pursuit of the same ends; maximum dollar leverage, and guarantees payout corresponds with success.)
But it's also profitable: even putting satellites in place is big business. Obviously, otherwise smart guys like Elon and company wouldn't be chasing it. So the "it hasn't happened because it wouldn't be profitable" argument is a half-truth; the business overall has been a profitable opportunity for a long time --- but the existing system has distorted the incentives such that actually attempting to do the whole thing has been less profitable than e.g. the chasing of McGuffins by the traditional players.
Anyway, there's definitely a Long Bet possible in this. There's all kinds of contradictions wrapped into your thesis vis-a-vis the present burst of new private sector activity on this stuff. Clearly either there's something flawed in somebody's understanding of difficulties and opportunities.
>> and all the while our know-how has bled away.
> And this is provable, how? I mean, besides a trite throwaway phrase, that is.
Well, sorry to be "trite." I don't find it trite at all, I find it tragic.
One data point, just one: the buddy I mentioned is (was, I assume this is up in the air now) working on flight control systems for future lunar landing craft. Pointing out the lunar lander trainer was something he specifically wanted to do to make the point. He's had to redo certain basic research in various areas in order to come up with a viable solution. It's just gone, Bill.
Re: the Atlas bit, ask anybody on the inside. Read even the popular literature. It's been stated repeatedly, in some cases by NASA officials in testimony before Congress, in justifying Orion etc. The data, the designs, the engineering, the test results --- misplaced, missing, destroyed. The tool and die capability gone. The generation of experts that designed and built them, gone, retired or dead. Continuity of know-how destroyed by lack of continuity in the program these three decades past, therefore no passing-down of the craft and science of it. The companies that built the parts, gone / consolidated / moved on. The capability is *just gone.*
Disagree if you want, prove me wrong. Prove my NASA friends wrong. I'd be thrilled to be wrong about this. But I happen to have both a little inside info and a lot of reading on and long-term interest in the topic. *Even the project management systems and tools that were employed are, in some cases, lost --- aside from the work product thereof, which as mentioned is piled in drifts and scattered across the floors of various NASA facility buildings.* But don't take my word on it; go find your own answer. If it's very different from mine, after doing the homework, I'll be shocked.
> And your suggested alternative is, what?
Here's a couple (ok, three) right off-the-cuff suggestions for improving the ability of elected officials to focus on big-picture projects rather than "whatever it takes to get my district to send me back in 4 years." (1) Term limits, (2) at-large representation, and (3) replacement of voter-selected representation with a public service lottery. Unfortunately, those are all rather large hacks of the operating system. But we may be reaching a point where large hacks or a complete rewrite are the only things that will get us to the next quantum of scale.
That's a whole other topic, of course.
More information about the FoRK