4 ms·

I did the same screenshot. Blown away by how simple the controls look. Everything is on the UI. Endeavor spacecraft controls: http://static3.businessinsider.co

I did the same screenshot. Blown away by how simple the controls look. Everything is on the UI. Endeavor spacecraft controls: http://static3.businessinsider.com/image/4f803e53ecad04f562000035/space-shuttle-endeavor.jpg http://static3.businessinsider.com/image/4f803e53ecad04f5620... I understand these are two different machines, but now it feels like we are in the future.

nickff 12y ago
To be fair to the engineers who designed the space shuttle, they could have automated almost all of the launch, docking and landing procedures, but they had to please the astronauts as well, and the latter wanted to 'fly' the craft. Some (such as Chris Kraft) have even argued that it would have been better to do this, so that the shuttles could be tested and demonstrated without humans in danger, to verify safety after Challenger and Columbia. It should also be made clear that even when an astronaut is 'manually' controlling the space shuttle, they are only providing inputs to the outer loop of a very complex feedback control system.
- Someone 12y ago
  "so that the shuttles could be tested and demonstrated without humans in danger, to verify safety after Challenger and Columbia." You could demonstrate, but I don't think it was possible to verify safety with such small a fleet in a decent timeframe. Let's say you accept a 1:100 chance of disaster, and you want to verify that experimentally. You certainly can't do that with fewer than 50 normal launches. That would have taken NASA over 5 years (assuming a vastly sped up launch schedule made possible by the lack of danger to humans) Alternatively, suppose you decide to test beyond the normal flight parameters, and launch a few shuttles with ice on them, a few with O-rings that aren't 100% up to standard, a few in high winds, etc. Even if you try to stay just on the edge of destructive testing, you would run out of shuttles to launch before you know it.
  - mikeash 12y ago
    I believe the approach would be to look for sub-critical failures, not testing to destruction (or lack thereof). For example, there were a lot of O-ring leak incidents in the years leading up to Challenger. Almost every flight had a problem with it to some degree. It just wasn't bad enough to destroy the craft until various factors came together. Similarly, leading-edge damage from debris was a problem throughout the Shuttle program, but it wasn't severe enough to lead to disaster until Columbia got unlucky. Once you were confident in the cause of the disasters, it would be reasonable to mitigate them, then fly a couple and look for any sign of these smaller failures, which should also be stopped by the fixes.
- Florin_Andrei 12y ago
  > even when an astronaut is 'manually' controlling the space shuttle, they are only providing inputs to the outer loop of a very complex feedback control system. That's pretty much how the human brain works, with the astronauts in this case being the cortex.

nickff 12y ago

To be fair to the engineers who designed the space shuttle, they could have automated almost all of the launch, docking and landing procedures, but they had to please the astronauts as well, and the latter wanted to 'fly' the craft. Some (such as Chris Kraft) have even argued that it would have been better to do this, so that the shuttles could be tested and demonstrated without humans in danger, to verify safety after Challenger and Columbia. It should also be made clear that even when an astronaut is 'manually' controlling the space shuttle, they are only providing inputs to the outer loop of a very complex feedback control system.

Someone 12y ago
"so that the shuttles could be tested and demonstrated without humans in danger, to verify safety after Challenger and Columbia." You could demonstrate, but I don't think it was possible to verify safety with such small a fleet in a decent timeframe. Let's say you accept a 1:100 chance of disaster, and you want to verify that experimentally. You certainly can't do that with fewer than 50 normal launches. That would have taken NASA over 5 years (assuming a vastly sped up launch schedule made possible by the lack of danger to humans) Alternatively, suppose you decide to test beyond the normal flight parameters, and launch a few shuttles with ice on them, a few with O-rings that aren't 100% up to standard, a few in high winds, etc. Even if you try to stay just on the edge of destructive testing, you would run out of shuttles to launch before you know it.
- mikeash 12y ago
  I believe the approach would be to look for sub-critical failures, not testing to destruction (or lack thereof). For example, there were a lot of O-ring leak incidents in the years leading up to Challenger. Almost every flight had a problem with it to some degree. It just wasn't bad enough to destroy the craft until various factors came together. Similarly, leading-edge damage from debris was a problem throughout the Shuttle program, but it wasn't severe enough to lead to disaster until Columbia got unlucky. Once you were confident in the cause of the disasters, it would be reasonable to mitigate them, then fly a couple and look for any sign of these smaller failures, which should also be stopped by the fixes.
Florin_Andrei 12y ago
> even when an astronaut is 'manually' controlling the space shuttle, they are only providing inputs to the outer loop of a very complex feedback control system. That's pretty much how the human brain works, with the astronauts in this case being the cortex.