|
|
|
|
|
by 3327
4717 days ago
|
|
This is absolutely amazing... The minimum amount of computation power, memory involved is just staggering. Isn't it actually more advanced today to a degree, in terms of getting the job done? How many lines of code would a moon landing guidance system be today? How many modules and dependencies? Which one wold be more redundant ( I feel the comparison is obvious here if not: simple program whose redundancy is in its simplicity, and a complex program where redundancy is with multiple levels of checks within) |
|
Dr. Brooks at MIT built a lander system with "modern" (at the time, Pentium-3 level) chips for DARPA and it was pretty straightforward, at a talk I attended he suggested that by this time (actually by 2006) he expected the whole package would be less than $100 worth of parts.
The only 'big' thing left between Earth and a regular visit to the Moon is some sort of on-orbit refuelling capability. Once you knock down that particular challenge lifting the parts into LEO which can then be refuelled and sent to L1, and then refuelled again and sent to the moon and back to L1 to top up, and back to LEO. There are existing systems for all the other pieces [1].
[1] So there are liquid fuelled engines, that can be refuelled under gravity, but none have been tested which can be refuelled in zero G. A proponent at ULA proposed a gas pump that would spin creating the necessary force to move propellant from tank to tank without the need for an additional non-dilutive pressurizing agent. But it is a fair criticism that we don't have zero g refillable engines yet.