3 ms·

Just to enumerate some recent missions: * NASA Curiosity: "2 years" -> 9.5 years, and counting. * NASA Spirit: "90 days" -> Just over 6 years, ending in 20

Just to enumerate some recent missions: * NASA Curiosity: "2 years" -> 9.5 years, and counting. * NASA Spirit: "90 days" -> Just over 6 years, ending in 2010. * NASA Opportunity: "90 days" -> Over 14 years, ending in 2018. NASA has had some failures too, of course. These stats are cherry-picked but still impressive.

wolverine876 5y ago
My point is that it's not impressive, it's just an old sales technique. Curiosity could have been designed for 20 years, but all NASA needs to do is say '2 years' and the project sounds like a resounding success if it fails in 10. (It is extremely impressive to have rovers and helicopters traversing Mars.)
- Rebelgecko 5y ago
  It's pretty much expected, these sorts of complex systems follow a bathtub curve. Failures either happen very early or after a very long time. If you design something so that it has a 99% of lasting 1 year, there's pretty good odds that it'll make it to 5 years since the riskiest time is the first weeks/months.
- quesera 5y ago
  That's a cynical perspective. though. In each of the above cases, the mission end date was extended because the hardware was capable of doing more science (or for more time) than it was designed for (and budgeted for). Like JWST, some of the procedures required for a successful mission had never been done before, and could not be tested fully on Earth. Many of those procedures have implications on mission longevity. Very smart and dedicated people make their most educated guesses, build in what they know, try to handle the unknowns, and hope and pray -- and more often than not, are proven correct. That's pretty close to my personal definition of "impressive engineering". There's a political angle too, of course. We don't know what NASA's "estimate fudge factor" is. I'd be amazed to learn that it is 5x, 24x, or 56x.
  - wolverine876 5y ago
    > the mission end date was extended because the hardware was capable of doing more science (or for more time) than it was designed for (and budgeted for) That is circular, it assumes that the announced end date isn't just underpromising what they know will be a later date. Neither of us is there when these dates are chosen, so we don't know what they are thinking. Arguably, NASA should be better able to estimate project lifetimes by now. NASA is doing amazing things; I'm not so cynical.
    - frosted-flakes 5y ago
      The announced end date is the goal. It just so happens that making the system reliable enough to last x days means it also has a high, but not sure, chance that it will last 10x days.
      - wolverine876 5y ago
        That is a possible description of their thinking, but what is it based on?
      - loonster 5y ago
        Easier to just say it's NASA's high percentage confidence interval. Maybe it's the 95% mark. They are 95% certain it will last 5 years. They likely also have lower confidence intervals with longer times. (If anyone knows what percentage the confidence interval actually is, I would be curious to know)
- garaetjjte 5y ago
  I guess it's that way for budgeting reasons. Probably easier to get funding for mission extension for rover that's already on Mars than budgeting 15 year mission at the beginning. And when it does actually fail earlier there's no hassle with reallocating funds.

wolverine876 5y ago

My point is that it's not impressive, it's just an old sales technique. Curiosity could have been designed for 20 years, but all NASA needs to do is say '2 years' and the project sounds like a resounding success if it fails in 10. (It is extremely impressive to have rovers and helicopters traversing Mars.)

Rebelgecko 5y ago
It's pretty much expected, these sorts of complex systems follow a bathtub curve. Failures either happen very early or after a very long time. If you design something so that it has a 99% of lasting 1 year, there's pretty good odds that it'll make it to 5 years since the riskiest time is the first weeks/months.
quesera 5y ago
That's a cynical perspective. though. In each of the above cases, the mission end date was extended because the hardware was capable of doing more science (or for more time) than it was designed for (and budgeted for). Like JWST, some of the procedures required for a successful mission had never been done before, and could not be tested fully on Earth. Many of those procedures have implications on mission longevity. Very smart and dedicated people make their most educated guesses, build in what they know, try to handle the unknowns, and hope and pray -- and more often than not, are proven correct. That's pretty close to my personal definition of "impressive engineering". There's a political angle too, of course. We don't know what NASA's "estimate fudge factor" is. I'd be amazed to learn that it is 5x, 24x, or 56x.
- wolverine876 5y ago
  > the mission end date was extended because the hardware was capable of doing more science (or for more time) than it was designed for (and budgeted for) That is circular, it assumes that the announced end date isn't just underpromising what they know will be a later date. Neither of us is there when these dates are chosen, so we don't know what they are thinking. Arguably, NASA should be better able to estimate project lifetimes by now. NASA is doing amazing things; I'm not so cynical.
  - frosted-flakes 5y ago
    The announced end date is the goal. It just so happens that making the system reliable enough to last x days means it also has a high, but not sure, chance that it will last 10x days.
    - wolverine876 5y ago
      That is a possible description of their thinking, but what is it based on?
    - loonster 5y ago
      Easier to just say it's NASA's high percentage confidence interval. Maybe it's the 95% mark. They are 95% certain it will last 5 years. They likely also have lower confidence intervals with longer times. (If anyone knows what percentage the confidence interval actually is, I would be curious to know)
garaetjjte 5y ago
I guess it's that way for budgeting reasons. Probably easier to get funding for mission extension for rover that's already on Mars than budgeting 15 year mission at the beginning. And when it does actually fail earlier there's no hassle with reallocating funds.