4 ms·

>I feel it's a lot easier to make SpaceX's rockets safer than ULA's rockets cheaper and reusable. The thing is that safety is cultural, not technical. It means

>I feel it's a lot easier to make SpaceX's rockets safer than ULA's rockets cheaper and reusable. The thing is that safety is cultural, not technical. It means that everyone realizes their limits, and has a conservative streak in them (if it ain't broken, don't fix it, or so least test it like crazy before putting a customers multi-million dollar payload on it). SpaceX, on the other hand, is all about moving fast and iterating fast.

ethbro 10y ago
To be fair, the conservative streak cuts against safety too. I believe a large number of US launch accidents were ultimately attributed to "component or measurement out of spec, but it's worked before so we'll expect it to work again."
- vanattab 10y ago
  Wouldn't the "conservative" choice in your example be to replace the part or correct the measurement to spec even though it has worked in the past.
  - Tempest1981 10y ago
    I think so. Conservative could mean 1) preferring gradual change over abrupt change, or 2) resistant to any change (stubbornly conservative?)
  - ethbro 10y ago
    Let's label them inertial conservatism and objective conservatism. Where the former is the sense of "we should do things the way we've previously done them" and the latter is "we should do things the most conservative way out of all of our options, even if we've never done it that way before." NASA during Challenger would have been an example of the former [1]. In a way that I think many overly bureaucratic organizations are. The other sense is somewhat oxymoronic given that conservatism is generally resistance to change. So when safety situations arise where the safest option requires acting novelly, who is there to lead the charge? I suppose in that case you'd want a radical safety department. [1] From the NYT article on the findings of the Challenger commission, "As Commissioner Feynman observed, the decision-making was 'a kind of Russian roulette. . . . (The shuttle) flies (with O-ring erosion) and nothing happens. Then it is suggested, therefore, that the risk is no longer so high for the next flights. We can lower our standards a little bit because we got away with it last time. . . . You got away with it, but it shouldn't be done over and over again like that.'" http://www.nytimes.com/1986/06/10/science/shuttle-findings-long-series-failures-key-portions-commission-report-challenger.html?pagewanted=all http://www.nytimes.com/1986/06/10/science/shuttle-findings-l...
greglindahl 10y ago
How can you tell what SpaceX's culture is like? Note that ULA does change things from time to time, but never does a test launch. Meanwhile, SpaceX is on the 3rd or 4th revision of F9, but their accidents have not been related to changes to the rocket.

ethbro 10y ago

To be fair, the conservative streak cuts against safety too. I believe a large number of US launch accidents were ultimately attributed to "component or measurement out of spec, but it's worked before so we'll expect it to work again."

vanattab 10y ago
Wouldn't the "conservative" choice in your example be to replace the part or correct the measurement to spec even though it has worked in the past.
- Tempest1981 10y ago
  I think so. Conservative could mean 1) preferring gradual change over abrupt change, or 2) resistant to any change (stubbornly conservative?)
- ethbro 10y ago
  Let's label them inertial conservatism and objective conservatism. Where the former is the sense of "we should do things the way we've previously done them" and the latter is "we should do things the most conservative way out of all of our options, even if we've never done it that way before." NASA during Challenger would have been an example of the former [1]. In a way that I think many overly bureaucratic organizations are. The other sense is somewhat oxymoronic given that conservatism is generally resistance to change. So when safety situations arise where the safest option requires acting novelly, who is there to lead the charge? I suppose in that case you'd want a radical safety department. [1] From the NYT article on the findings of the Challenger commission, "As Commissioner Feynman observed, the decision-making was 'a kind of Russian roulette. . . . (The shuttle) flies (with O-ring erosion) and nothing happens. Then it is suggested, therefore, that the risk is no longer so high for the next flights. We can lower our standards a little bit because we got away with it last time. . . . You got away with it, but it shouldn't be done over and over again like that.'" http://www.nytimes.com/1986/06/10/science/shuttle-findings-long-series-failures-key-portions-commission-report-challenger.html?pagewanted=all http://www.nytimes.com/1986/06/10/science/shuttle-findings-l...

greglindahl 10y ago

How can you tell what SpaceX's culture is like? Note that ULA does change things from time to time, but never does a test launch. Meanwhile, SpaceX is on the 3rd or 4th revision of F9, but their accidents have not been related to changes to the rocket.