Armadillo Aerospace Update Thread

[Patchouli]

http://www.theregister.co.uk/2013/08/02/armadillo_aerospace_hibernation/

Maybe he just needs to take time out to improve his business model

Definitely sad news as they were a very interesting and off beat company that did things differently until recently.

Hopefully he'll be able to rethink his business model and come up with some money.
You need to either have deep pockets or be making money at it.

[FutureSpaceTourist]

I'm saddened too. I think a key issue is that John Carmack has a day job and AA just can't make enough progress without a leader with enough time to focus and push things on.

[mwfair]

AA is a reflection of Carmack, and it seemed to me like Carmack mostly just had a desire to build engines and rocket systems, for the fun of it (that probably describes all the crew).
The business model was secondary, and was either  A) 'build it and they will come' (on the off chance the a finished rocket is produced), and/or B) build a reputation for creative, quality solutions to limited problems that would result in contracting work with clients who needed rocket help.
 The product part turned out not to happen, IMHO because he was mostly motivated by the fun of it, not by a concrete, viable goal with a realistic customer base. 
As it turned out though, B) did happen.  JSC hired him to build exactly the kind of tinkering system that he was so successful in cranking out.
IMO, the above is most admirable and reasonable.  It contributed to the industry, and has resulted in a body of work that could be used to go all the way, i.e. A) above could be resurrected.
But Carmarck succeeded in doing what, in my estimation, he set out to do.  Have fun.  Either he can change his goals and restart the thing, or could sell it to be reformed with a viable business plan.
In my experience as an engineer, the hardest aspect of things is the product planning and marketing, i.e. 'business plan'.
Carmack succeeded in this with Quake due to both programming and business savvy.  But the rocket 'product equation' didn't quite solve.

[zaitcev]

The most important lesson is that as soon as newly constituted employees turned to the "real rocket" mode, they failed, because they blew the budget. This is what the "rockets are hard" crowd fail to understand again and again: the legacy way is not just the expensive way, it's ludicrously expensive way.

Some of the carping in this thread was simply obscene, and honestly it's a big pity that John didn't succeed. On the other hand, they do ignore every time Elon talks about ridiculous quotes from his suppliers for things like valves that are 30 times cheaper when made by SpaceX in-house. So even if John has succeeded, they would still say it.

All that wealth is taken from the customers, eventually, and floats into the pockets of people who are making a comfortable living by preaching "rockets are hard". This has got to stop for humanity to make meaningful progress in space.

[mr. mark]

Armadillo was just way too out front and did not have a good business plan going forth. They could have concentrated on eventual private research landers for the Moon and eventually Mars. Instead, they funneled money on pet projects with little or no focus on the end result. The company is still a good buy for someone who has a strong focus and can sift through the mud and focus the program. I'd kill off the Stig program. UP Aerospace already has that nitch. As I said focus the technology already developed toward a near term lunar lander for private research. This seems well within their sphere.

[baldusi]

Looking at the history of their postings, I'm finding that because rockets are hard, you can't keep the tinkerer's attitude if you actually want to have a solid product. From things as little as not having a written launch sequence, and after having it not having validated it, to the problems they had with the navigation, telemetry and recovery systems. Most of those failures had "lack of ground testing" all over them.
There's a reason rockets can spend a year in the pad for the inaugural launch (think Delta IV Heavy at Vandemberg or Falcon 9 Flt 1). There's a lot of tests to make sure everything works. Of course you can do the old Russian way and launch ten or so prototypes until you have all the procedures set. But in the end that depends on extremely cheap labor.
My sense is that he assumed that the step from 60km to recoverable 200km was linear, when in fact is more like exponential. I'm pretty sure that he could do it a lot cheaper than if you asked LM or Astrium to develop such a system. But it was probably more like 5M and five years than 2M and two years.
In the end, I still believe that purely commercial companies can do it for 50% to 20% of a legacy system. But not by 2% to 5%. Even in the above stated case of SpaceX, one thing is the manufacturing and certification cost and another the design and qualification cost. SpaceX has one huge factory and lots of tooling and employees working on space quality production. I'm pretty sure they have one of the most efficient plants around and it's very difficult to compete on the marginal price with them. Any third party would have to dedicate a significant effort to keep all the certified processes and quality control just for a few parts. Thus, for SpaceX in particular, it's quite possible that they are doing the most economically sound decision by doing almost everything in house.
But they can do that because the have billions in contracts. AA wanted to do a business where they would have contracts in the 100k's range. That's two or three orders of magnitude less than SpaceX. At those revenue levels you simply can't afford your own factoy. Look at the trouble they had finding a bath big enough to age their new tanks. And that's something that's a century old process. There's no secret on how to do it. It's just that it's simply too expensive for AA. SpaceX could afford it no problem. AA couldn't for their business model.

[mr. mark]

Looking at the history of their postings, I'm finding that because rockets are hard, you can't keep the tinkerer's attitude if you actually want to have a solid product. From things as little as not having a written launch sequence, and after having it not having validated it, to the problems they had with the navigation, telemetry and recovery systems. Most of those failures had "lack of ground testing" all over them.
There's a reason rockets can spend a year in the pad for the inaugural launch (think Delta IV Heavy at Vandemberg or Falcon 9 Flt 1). There's a lot of tests to make sure everything works. Of course you can do the old Russian way and launch ten or so prototypes until you have all the procedures set. But in the end that depends on extremely cheap labor.
My sense is that he assumed that the step from 60km to recoverable 200km was linear, when in fact is more like exponential. I'm pretty sure that he could do it a lot cheaper than if you asked LM or Astrium to develop such a system. But it was probably more like 5M and five years than 2M and two years.
In the end, I still believe that purely commercial companies can do it for 50% to 20% of a legacy system. But not by 2% to 5%. Even in the above stated case of SpaceX, one thing is the manufacturing and certification cost and another the design and qualification cost. SpaceX has one huge factory and lots of tooling and employees working on space quality production. I'm pretty sure they have one of the most efficient plants around and it's very difficult to compete on the marginal price with them. Any third party would have to dedicate a significant effort to keep all the certified processes and quality control just for a few parts. Thus, for SpaceX in particular, it's quite possible that they are doing the most economically sound decision by doing almost everything in house.
But they can do that because the have billions in contracts. AA wanted to do a business where they would have contracts in the 100k's range. That's two or three orders of magnitude less than SpaceX. At those revenue levels you simply can't afford your own factoy. Look at the trouble they had finding a bath big enough to age their new tanks. And that's something that's a century old process. There's no secret on how to do it. It's just that it's simply too expensive for AA. SpaceX could afford it no problem. AA couldn't for their business model.

When you are a small company the last thing that you want to do is enter a market that has already been cornered. You need to find a nitch market for yourself. Lunar landers for private and scholastic research would have been a better option with the product they had on hand. Suborbital research would have been a hard nut to crack from a business perspective. Too many small companies already out there. At the end of the day, suborbital is just that suborbital. The end result is the same whether you have reuse or not. Research gets done. A lot of companies fighting over a small piece of pie.

[bad_astra]

I'm sad to see this happen to Armadillo. As others have said, they were one of my favorites.

I did think that many of their failures were perhaps a bit too public. It was refreshing to see build-a-little-test-a-little actually demonstrated out in the open, but it probably scared off investors who did not want their name attached to a lot of video clips of vehicles augering in.

They also seemed to have been all over the place on development plans. H202 monoprop w/jet vanes, then differential thrusters, then quads, then stigs, etc. If they had focused part of their energy on some lower hanging fruit and procuded consistant, albeit smaller, successes, could they have been more akin to where XCOR is at? Hard to say. If RRL had been a success, we might well be talking about how well they are doing right now, but many a good idea is hinged on the success of another.

I do hope the situation improves and someone with the means can save the Armadillo.

[A_M_Swallow]

One bit of good news for Armadillo they are no longer competing with the Falcon 1 to launch cubesats.

[Lars_J]

One bit of good news for Armadillo they are no longer competing with the Falcon 1 to launch cubesats.

Huh?

[Cinder]

Carmack has added another work priority (Oculus Rift) above id software, which was itself above AA.  He's gushing pretty good over OR, unlike about id, nevermind AA.

[beancounter]

One bit of good news for Armadillo they are no longer competing with the Falcon 1 to launch cubesats.

Huh?

Exactly, F1 no longer exists.

[savuporo]

, waiting for either Carmack to get free time and money or "someone with a few million dollars who wants to build rockets".

Oh, but some very wise people here were saying that Carmack had "all the money he wanted".

I dont think a few million dollars here would do the trick, if their previous run rate was a $1m a year - which is really cheap if you factor in only the labor of a few good top talent engineers - to keep the team motivated and on a path of sustained, interesting progress you would probably need to be running at $5M a year or so.

Many people with software backgrounds get into hardware projects and lose momentum in the similar way - for your software team building kick ass tech, you provide them with a few laptops and a few expensive tech toys as a downpayment, but from there on its pretty much salary, reasonable working environment but most of all results in a way of built, demonstrable tech that keep them engaged.
For hardware guys - you have to burn money on an ongoing basis to keep the metal moving, whether its chips, pcbs, pistons and exhaust parts, rocket motors or something else. Your guys WILL get bored looking at the same rocket for years - and Armadillo definitely slowed down in that regard towards the end. Any serious rapid development HW shop HAS to give enough freedom to engineers to burn out H-bridges, melt combustion chambers and annoy the hell out of neighbors, and you have to budget for that and make sure you can spin out usable products from that destructive cycle.

[savuporo]

Remember that one of the causes Carmack pointed to, himself, was that expensive and difficult-to-work materials were creeping into their designs.

Right - and at this point you would have to make a clear decision -
- multiply your cash burn rate accordingly
- slow down your hw progress
- go with the low tech, cheap materials approach at the expense of less elegant designs.

first usually involves finding other investors and a lot of people dont want do do that to retain full control
slowing down has the risks that AA ran into
third approach engineers just generally dont like in a lot of cases - every engineer wants to build elegant designs. However, effectiveness is often fundamentally at odds with elegance.

Akin's 36: Any run-of-the-mill engineer can design something which is elegant. A good engineer designs systems to be efficient. A great engineer designs them to be effective.

[Lars_J]

One bit of good news for Armadillo they are no longer competing with the Falcon 1 to launch cubesats.

Huh?

Exactly, F1 no longer exists.

Exactly, and Armadillos orbital plans were so vague and decades off that "falcon 1 competition" statement makes even less sense... Hence my "Huh?".

[JazzFan]

Sarcasm.........., got it.

[Danderman]

Oculus Rift hires Doom co-creator John Carmack as Chief Technology Officer

http://www.engadget.com/2013/08/07/oculus-rift-john-carmack-cto/

Carmack tweeted a bit of clarity to his new role at Oculus among his other jobs, saying, "My time division is now Oculus over Id over Armadillo. Busy busy busy!"

[QuantumG]

, waiting for either Carmack to get free time and money or "someone with a few million dollars who wants to build rockets".

Oh, but some very wise people here were saying that Carmack had "all the money he wanted".

and I stand by it. That's why he's not out raising money. Carmack don't work that way. If it can't be done with his out-of-pocket he simply doesn't do it. AA is being slow walked now because Carmack has come to the startling conclusion that this rocket stuff *really is hard*.

[R7]

Why won't they throttle down to work on STIG with the earlier volunteer work mode and correct the errors mentioned? Hibernation sounds like nothing will happen.

[DanielW]

I seem to remember him saying (in quake con keynote QA?) that he might have to wait for another liquidity event. Zenimax going public? or some other. I don't think the terms of the id deal were ever disclosed. But it stands to reason that he may not really have the liquid assets to pursue rocketry right now regardless of net worth.