It has been nice to follow this year's Space Access Conference in Phoenix through the corps of live bloggers at the event. One of the more interesting presentations has been from Armadillo Aerospace (blogged by Clark Lindsey and Rand Simberg). As some of you know, while I follow with interest Armadillo and the other suborbital teams, I don't really cover them in this blog in a systematic manner. This seems likely to change as John Carmack, head of the Armadillo team, has started to mention long-term plans for orbital launch. Carmack is becoming something of a kindred spirit with and disciple of Lutz Kayser, the "voice in the wilderness" for modular launch. Carmack proposes to create an array of Pixel-like modules in order to eventually make orbit in two or three stages.
One year ago tomorrow, on the occasion of the first Falcon 1 failure, Lutz Kayser commented on this blog about his skepticism on the way SpaceX and some others were doing business. I replied that John Carmack might be doing something along the lines that he advocated. Subsequently, in May last year, Kayser visited Armadillo and was said to be quite generous. For instance, he left some hardware with Armadillo from the OTRAG effort -- a project in the late '70s and early '80s which Kayser led.
I can claim no credit for getting the two together -- they probably spoke months before Kayser posted here, and in any event it is no big parlor trick to recognize the similarities between them and mention it. But it is interesting to see the conversation and interplay between the two similar approaches develop over the last year+.
From the little I understand of Kayser's thoughts, the benefits of the modular approach are closely tied to the engineering process necessary to make the approach work. There is no substitute for building, flying, and redesigning key parts in an iterative manner. If your vehicle is modular, you can focus on simplifying and making robust those few key parts. You can test low numbers of modules and then scale up once you're ready for something bigger. Low costs and high reliability at scale follow naturally.
I don't think the point of the exercise is to assign a definite per pound price for modular spaceflight -- it can't help but be lower than that available through the traditional vendors -- but some of the cost figures would be absurd, if they weren't true. Carmack says that so far, Armadillo has spent some $3 million in research and development funds and that the company could get to orbit on some million more. He believes that a team like his, 8 part-time volunteers and 1 paid employee working out of a garage or small warehouse, could produce an orbital launch vehicle. Of course, the marginal costs of the rocket depend on module manufacturing costs and any reusability in the system.
In addition to Space Access, we are being treated to a wealth of treasure on modular vehicles, like Carmack's February and March updates. These updates are where we can follow as Carmack backs up his words with flight-tested hardware. Also, in the next few days, Armadillo will release on its website the video it presented at Space Access. As I understand from the reports, the video is very good, like similar high quality videos presented by Armadillo in the past.