4 ms·

I've always felt that when building these incredibly valuable satellites one should consider building 2. I know JWST is probably not a good example with the

I've always felt that when building these incredibly valuable satellites one should consider building 2. I know JWST is probably not a good example with the work required to make the mirror but but in a lot of cases one could probably replicate a rare satellite during production for much less than 2x the cost. Then you have a flight backup, and alternatively you can later double your science with another launch if everything goes smoothly on launch one. As said probably not a reality for JWST but these high risk launches always concern me with no backup option. What a tragedy it would be.

kbutler 9y ago
Satellite insurance is probably cheaper (though rates may rise?) https://en.wikipedia.org/wiki/Satellite_insurance https://en.wikipedia.org/wiki/Satellite_insurance http://spacenews.com/space-insurers-warn-that-current-low-rates-are-not-sustainable/ http://spacenews.com/space-insurers-warn-that-current-low-ra...
- Someone 9y ago
  You can’t really insure a scientific career. Let’s say you work on the design and implementation of a novel scientific instrument for a satellite for over ten years. That’s a job that, relatively, isn’t paid well, but if launched, the instrument will get you papers and recognition. If the rocket launching your toy blows up, all you can get from an insurance is money, and that’s not something you value that much in life. Some satellites also have fairly tight launch windows, for example if you want to intercept a comet.
  - planteen 9y ago
    > Let’s say you work on the design and implementation of a novel scientific instrument for a satellite for over ten years. That’s a job that, relatively, isn’t paid well, but if launched, the instrument will get you papers and recognition. That's true, but building two of everything causes other problems. For a principal investigator (PI), they spend years drumming up support for a concept then hope that it continues to get funded, built, launched, and becomes operational. If we build 2 of everything, let's say it adds 2/3 to the cost. That means less money for other missions overall. That would make many scientists unhappy.
- Robotbeat 9y ago
  Government payloads are not insured. The government is "self-insured."
planteen 9y ago
This happens sometimes. For example, GPS and DMSP (Defense Meteorological Satellite Program) had many blocks of identical satellites. One of the DMSP spacecraft will never launch and has been ordered to be scrapped, since it is now so out-of-date after being in storage for so long. https://en.wikipedia.org/wiki/Defense_Meteorological_Satellite_Program https://en.wikipedia.org/wiki/Defense_Meteorological_Satelli... http://space.skyrocket.de/doc_sdat/dmsp-5d3.htm http://space.skyrocket.de/doc_sdat/dmsp-5d3.htm The out-of-date aspect probably explains why this doesn't happen more. I worked on JPSS-1, a polar orbiting weather satellite. It was a follow on to NPP, but it was not an exact copy. After having experience using NPP in production (in space), naturally improvements and better technology could be integrated. These two satellite launches were separated by 6 years and a new satellite takes 4-5 years to design, build, and integrate. Think how much technology changes in 6 years.
- ceejayoz 9y ago
  Another instance of this: https://en.wikipedia.org/wiki/2012_National_Reconnaissance_Office_space_telescope_donation_to_NASA https://en.wikipedia.org/wiki/2012_National_Reconnaissance_O... "Hey, we've got a pair of better-than-Hubble telescopes just sitting around. Want 'em? You have to promise not to look at Earth with them, though."
zaroth 9y ago
Did you watch the movie Contact?
- phinnaeus 9y ago
  "Why build one when you can build two for twice the cost?"
cdibona 9y ago
So you'll often build two or three test versions of the sat that you subject to serious, and sometimes destructive, testing to simuluate launch stresses, etc. So if a given satellite blows up on the pad, they can use these test articles as a backup. So they don't really start from scratch..
paulsutter 9y ago
Good idea. Or even just make mutiples of the rare/expensive/long-lead-time components, optimize for a second build rather than actually build two. Better still, use a swarm of cheap unreliable satellites for the same function (often not possible)

kbutler 9y ago

Satellite insurance is probably cheaper (though rates may rise?) https://en.wikipedia.org/wiki/Satellite_insurance https://en.wikipedia.org/wiki/Satellite_insurance http://spacenews.com/space-insurers-warn-that-current-low-rates-are-not-sustainable/ http://spacenews.com/space-insurers-warn-that-current-low-ra...

Someone 9y ago
You can’t really insure a scientific career. Let’s say you work on the design and implementation of a novel scientific instrument for a satellite for over ten years. That’s a job that, relatively, isn’t paid well, but if launched, the instrument will get you papers and recognition. If the rocket launching your toy blows up, all you can get from an insurance is money, and that’s not something you value that much in life. Some satellites also have fairly tight launch windows, for example if you want to intercept a comet.
- planteen 9y ago
  > Let’s say you work on the design and implementation of a novel scientific instrument for a satellite for over ten years. That’s a job that, relatively, isn’t paid well, but if launched, the instrument will get you papers and recognition. That's true, but building two of everything causes other problems. For a principal investigator (PI), they spend years drumming up support for a concept then hope that it continues to get funded, built, launched, and becomes operational. If we build 2 of everything, let's say it adds 2/3 to the cost. That means less money for other missions overall. That would make many scientists unhappy.
Robotbeat 9y ago
Government payloads are not insured. The government is "self-insured."

planteen 9y ago

This happens sometimes. For example, GPS and DMSP (Defense Meteorological Satellite Program) had many blocks of identical satellites. One of the DMSP spacecraft will never launch and has been ordered to be scrapped, since it is now so out-of-date after being in storage for so long. https://en.wikipedia.org/wiki/Defense_Meteorological_Satellite_Program https://en.wikipedia.org/wiki/Defense_Meteorological_Satelli... http://space.skyrocket.de/doc_sdat/dmsp-5d3.htm http://space.skyrocket.de/doc_sdat/dmsp-5d3.htm The out-of-date aspect probably explains why this doesn't happen more. I worked on JPSS-1, a polar orbiting weather satellite. It was a follow on to NPP, but it was not an exact copy. After having experience using NPP in production (in space), naturally improvements and better technology could be integrated. These two satellite launches were separated by 6 years and a new satellite takes 4-5 years to design, build, and integrate. Think how much technology changes in 6 years.

ceejayoz 9y ago
Another instance of this: https://en.wikipedia.org/wiki/2012_National_Reconnaissance_Office_space_telescope_donation_to_NASA https://en.wikipedia.org/wiki/2012_National_Reconnaissance_O... "Hey, we've got a pair of better-than-Hubble telescopes just sitting around. Want 'em? You have to promise not to look at Earth with them, though."

zaroth 9y ago

Did you watch the movie Contact?

phinnaeus 9y ago
"Why build one when you can build two for twice the cost?"

cdibona 9y ago

So you'll often build two or three test versions of the sat that you subject to serious, and sometimes destructive, testing to simuluate launch stresses, etc. So if a given satellite blows up on the pad, they can use these test articles as a backup. So they don't really start from scratch..

paulsutter 9y ago

Good idea. Or even just make mutiples of the rare/expensive/long-lead-time components, optimize for a second build rather than actually build two. Better still, use a swarm of cheap unreliable satellites for the same function (often not possible)