3 ms·

what puzzles me all the time is why not build two instead. it looks like the R&D portion of the budget might be greater than the actual manufacturing cost, so

what puzzles me all the time is why not build two instead. it looks like the R&D portion of the budget might be greater than the actual manufacturing cost, so building two would be less than 2x the price. if one launch fails, we've got a backup. if both launch, we have two instruments deployed. why don't we do it??

icegreentea2 7y ago
The NASA-Congress decision making complex is kind of a monster. Often makes lot of penny wise, pound foolish decisions. (I should say ${X-Body}-Congress decision making complex is like that).
mhb 7y ago
Is there any economy of scale? Why not wait until the first one fails to build another one?
- java-man 7y ago
  I would expect the R&D costs to dominate. One does not even need to consider economy of scale. For example, if the R&D is $1 million, and manufacturing one mirror costs $100k, the total price for two mirrors is $1.2 million vs. $1.1 million for one. Plus, consider a non-zero probability of a launch failure.
- TeMPOraL 7y ago
  Because that requires keeping all the specialized machinery and knowledge alive for the duration of the whole project. Given the length of this project, I wouldn't be surprised if at this point NASA couldn't order a second one built even if they wanted to. The unfortunate thing about advanced technology is that if you don't use it, you lose it. More specifically, you lose the research and machining setups and people with knowledge about them in their head; the main problem here being that documentation doesn't capture everything - a lot of tacit knowledge is captured in e.g. tuning of the machines, undocumented fixes made on-site, idiosyncrasies in manufacturing steps, etc.
  - Rebelgecko 7y ago
    Unfortunately that's already the case with many long running projects. A friend of mine that works on JWST recently fixed a mistake that was made by someone who retired 10 years ago.
sanxiyn 7y ago
NASA sometimes does it. Mars 2020 is an upgrade of Mars Science Laboratory (aka Curiosity) spare.
dylan604 7y ago
“The first rule of government spending is build two at twice the price” — S. R. Hadden
azernik 7y ago
Generally they do build two - one as a ground testbed for interpreting diagnostic data, trying to reproduce issues, and testing software changes/commands. However, at the very low rates we're talking, everything is (damn near) hand-built. Economies of scale don't kick in until higher volumes.

icegreentea2 7y ago

The NASA-Congress decision making complex is kind of a monster. Often makes lot of penny wise, pound foolish decisions. (I should say ${X-Body}-Congress decision making complex is like that).

mhb 7y ago

Is there any economy of scale? Why not wait until the first one fails to build another one?

java-man 7y ago
I would expect the R&D costs to dominate. One does not even need to consider economy of scale. For example, if the R&D is $1 million, and manufacturing one mirror costs $100k, the total price for two mirrors is $1.2 million vs. $1.1 million for one. Plus, consider a non-zero probability of a launch failure.
TeMPOraL 7y ago
Because that requires keeping all the specialized machinery and knowledge alive for the duration of the whole project. Given the length of this project, I wouldn't be surprised if at this point NASA couldn't order a second one built even if they wanted to. The unfortunate thing about advanced technology is that if you don't use it, you lose it. More specifically, you lose the research and machining setups and people with knowledge about them in their head; the main problem here being that documentation doesn't capture everything - a lot of tacit knowledge is captured in e.g. tuning of the machines, undocumented fixes made on-site, idiosyncrasies in manufacturing steps, etc.
- Rebelgecko 7y ago
  Unfortunately that's already the case with many long running projects. A friend of mine that works on JWST recently fixed a mistake that was made by someone who retired 10 years ago.

sanxiyn 7y ago

NASA sometimes does it. Mars 2020 is an upgrade of Mars Science Laboratory (aka Curiosity) spare.

dylan604 7y ago

“The first rule of government spending is build two at twice the price” — S. R. Hadden

azernik 7y ago

Generally they do build two - one as a ground testbed for interpreting diagnostic data, trying to reproduce issues, and testing software changes/commands. However, at the very low rates we're talking, everything is (damn near) hand-built. Economies of scale don't kick in until higher volumes.