Rapid Re-engineering is a Key to SpaceX and Tesla Success

by

Rapidly iterating on versions rockets and rocket engines is a huge advantage for SpaceX. Elon said he expects that the fifth version of the Raptor engine should be an A+ version of that engine. The second version of the Raptor is testing soon and should be about B+.

SpaceX is not only make more versions they are going to far higher production volumes for rockets and engines.

This is also happening with the Tesla cars. Tesla is making improvements and adding them while the production lines keep running. They make running changes. They made 13 changes in some areas of their Model 3s.

They make small and big changes.

The big changes are things like new alloys so that they can cast the front and back end of their vehicles without heat treating and other work.

SOURCES- elon Musk, Everyday Astronaut, Sandy Munro
Written By Brian Wang, Nextbigfuture.com (Brian has shares of Tesla)

46 thoughts on “Rapid Re-engineering is a Key to SpaceX and Tesla Success”

  1. "Thus, my orig observation, that Musk having more tests does not prove he is closer to crew flights." Yep.

  2. I agree in general. It depends on the cost of the tests. The two methods become the same as they are imperfect. As Musk first guess is better (did too much work), more waterfall like, and as Bezos crew design is flawed, he has to explore. So, does Musk having more tests mean he is ahead in race to produce crew rated engine? I'll give Musk credit even if he has to wait for SS to use it, just NASA ok good enuf. Now, Musk/Falcon cargo super reuse booster is nothing but goodness.

    However, as design tech gets better, expect more waterfall. It does not have the hidden factors that iterative design can develop as changes are tried. The cost of waterfall is both the expensive design and the missed problems tests would have revealed. Better design tech *solves* both, or at least improves them. Sims are this. Iterative costs are test costs and AI like hidden factors as the test design changes have far less big picture focus. When I do code, if I change something I try to find all the effects. Far easier if the final design is being thot of, from the beginning.

  3. True for blue Origin. But not true for ULA. Most of the parts used By ULA are purchased from other companies. The purchase the solid rocket busters and liquid dueled engines from other companies. About the only things ULA developed and makes are the fuel tanks, guidance and control electronics and the launch towers and platforms.

  5. There's a reason why basically no companies use the Waterfall software development model. Same reason why SpaceX doesn't make rockets the same way BO "makes" rockets.

    You can see the reason why old style code applied to rocketry fails if you look at the irrelevancy of Arianespace. BO is also a similar failure but kept alive by Bezos' billions.

  6. They have a different model for development. They do not quickly iterate by design. They are slowly working up to launching a fully working rocket that will meet all of the requirements they identified years ago.

    And when it launches it will be obsolete if Starship succeeds at being reusable.

  7. And the engineers run the company. A very important point.

    Engineering companies should be run by engineers.
    Soda companies should be run by marketers.
    Gas stations should be run by MBAs.

  9. Supposedly they have tests at Stennis in Sept sched. There is no sched until you have the first one done. There are only projections. It is true, ahem, rocket science.

  11. no one appreciate their self-esteem, culture or individual, tweaked – tis the Millenial/ Gen Z way.

  16. ouff.
    I am envisioning 'affordability': a huge hive of tiny micro-space condos with exorbitant monthly fees for minimal amenities and non-existent parking… here we come… wasn't there a novel/ movie that de-glamorized space travel by showing the required orbital slum just so we could have a regular 'presence'?

  17. Well. we could put some rare aboriginal entities (tardigrades and such) to conquer, some rare resources to abuse and exploit, and allow for there to be an unfettered, anything goes approach to orbital colonization — maybe that would attract colonialism – it did a few hundred years ago.

  18. I would agree with that analogy if north america was more like the arctic, inhospitable and minimally self-sustaining – even if the resources are immense.
    It is that first toe hold and how easy it is to provide for itself locally that may influence the speed and size of future waves.

  19. Well, we were talking NASA in the early 2000s. NASA people had the incentive to build rockets from scratch. NASA is not building rockets of any practical significance anymore (sorry, SLS). So these people need too move on, not wait around. Plenty of work! I think Rutan's basic large point was that NASA had no vision, and was just building repetitive rockets. Shuttle was not going to be affordable. ISS was not a planet. Boring. Musk won't lose focus.

  20. What % of Europeans actually travelled across the atlantic in the first 20 years after Columbus? Even the first century?

  21. Even there, who had better career prospects in 1940? Airplane designers, or airship designers?

    The airship guys were out of luck because their projects never delivered viable products. The airplane designers had successful products, so they got the next job with the next project.

  22. the upper middle class – top 10% will always be the goal of most tourist companies… sub-LEO, etc.

  23. but how big of a market really exists? if only 1% of 1% of 1% make it on some LEO trip in the next 20 years, how can you make an industry out of that? A couple of academic programs get some launch access. A few dozen companies get rockets, craft, and satelites up there… then…?

  25. to a degree… with increasing success, comes increasing exposure to malignant forces — witness Amazon, Apple, Google — monopoly investigations, union attacks, patent troll invasions…

  26. and no unions… and increasingly permissive regulatory environment… and active competition… and increased public interest and acceptance… it appears. Golden age.

  27. A Xerox PARC of space?
    PARC could hide its failures in a tightly secured lab (and put out its curated successes) — but SpaceX; this is for all the world to see.

  29. tis but a fine line — easy to describe but difficult in practice. SpaceX is an impressive inventor's paradise… only to be accomplished by amazing people. A few less key players and we have either many, many failures — or slower successes.

  30. I did not call Musk a moron. The story is about raptors, not old style kerosine engines, which many have flown.

  33. What surprises me is that neither ULA nor Blue Origin seem to be able to rapidly finish their first iteration.

  34. Exactly. Design like old style code is written, the testing is pretty straightforward, some, not "no", just the logic. For AI, millions of *tests* are run, and the machine gradually learns by passes or fails, and we have our solution. The Bezos tests are not for training, as Musk's ARE, to learn, but to verify the intentional design to be crew rated at the start of use. Thus, my orig observation, that Musk having more tests does not prove he is closer to crew flights.

  35. "Beware of the AI nature of his solutions, they work but it may not be clear exactly why, theoretically. "

    This is rich, coming from the guy who obsessively talks about Bezos, who literally believes in doing no real world testing at all until the inaugural flight of New Glenn.

  36. It may have been his only big talk, certainly one of the few. Big crowd in the LBJ Library. Great optimism.

  39. That is true overall, but the "designs rockets" part makes for a special case. The *not* starting over from scratch option can easily leave these people out from the start of the project at hand. Re order, not re design. Things seem to be picking up for NASA tho, Rutan was talking before (2004) the Commercial project was going, no real plans. Dark days. Now, NASA designs payloads and buys rockets, so they actually do want the re order option if it works. Micr0g ISS stuff is about to explode. Electrophoresis, mass spectroscopy, in drug mfg alone will be bigger than Mars program.

  40. Every single employee that designs rockets has the financial incentive to start over from scratch.

    Except for that whole "project cancelled because we were 10 years and $10B over budget. Everyone collect their redundancy letter on the way out."

    ONce you take that into account, having the project succeed is actually better for your career.

  41. Every single employee that designs rockets has the financial incentive to start over from scratch. When Burt Rutan gave a talk after the X Prize, he had a chart of all the similar rockets NASA and the military had designed. He pronounced NASA as "nay-say". Not all people go by their selfish incentives, btw.

  42. This, to me, is from his computer code background. When coding, the tests are constant and cheap. Cheap is very important! Musk has decided that the product is cheap enuf to plow ahead. Good for him! Beware of the AI nature of his solutions, they work but it may not be clear exactly why, theoretically. This is made worse by the constant upgrades. The explosion of the Dragon on a very late test run was this sort of thing. Not landing propulsively means the designed check valves were not routinely used, after a change!, and should have been blow outs or, as BB had already discovered, doubled. Could have been very bad. But, Apollo waterfall do it all one time, do it right, launch it is even more frightening. Big advance if simulations take over, tests are really cheap then.

  43. Elon Musk has a unique all rounded design philosophy. On one hand he is using parts that are in the market as much as possible off shelve, on the other, he is avoiding subcontracting other companies to design unique parts and have them made in his companies as much as possible at a lower price and design time cycle. In both cases money and time is saved in his companies' designs. His companies initial designs are intentionally simple made for easy testing, and they are followed soon by more complex and improved ones. That is definitely true for the payload capacity of the Falcon 9. Still we need to understand that like everything he does his approach is going to be emulated soon by other companies.

  46. A welcome sight in an industry rife with dreams aborted too soon, just because of a single misstep and the eagerness of politicians to start from scratch.

    Given the only leader SpaceX really has to please is Elon Musk, the rocket will be whatever he wants it to be and the gods of engineering allow.

Comments are closed.