3 ms·

At the very least the "young" engineers have been educated with materials from and by NASA engineers. I don't know what you think Falcon is running on, but the

At the very least the "young" engineers have been educated with materials from and by NASA engineers. I don't know what you think Falcon is running on, but they are definitely using more advanced and faster computers than those used for landing probes on planets. Even those raptor engines are quite complex in terms of regulation. Also, SpaceX certainly used lots of computing power for simulation of all sorts. Neither the computing power nor the algorithms have been around for decades, necessarily. Point is reuseability isn't easy. Nor necessarily cheaper. SpaceX can't do it on the scale of SLS, by the way. They had more trouble than anticipated with Falcon heavy at "only" 26t to LEO. SLS can do almost four times as much. Starship may be able to do it. But it is far from ready for prime time.

bottlepalm 4y ago
I didn’t say it was easy, I said it has been possible. If the investment was made 50 years ago, the ROI for science and technology over the following decades would have been massive. The Shuttle did prioritize reuse to some extent, but the design was frozen and never iterated on. Buran automated the landing and had plans for landable side boosters. DCX also demonstrated propulsive landings. The thing is these were all side project. NASA really never prioritized economical systems as demonstrated by living with the Shuttles exorbitant costs and building SLS. Even though by driving down costs they could of performed orders of magnitude more science. Obviously politics is to blame, not feasibility.
- bayesian_horse 4y ago
  I highly doubt that 50 years ago there were microcomputers capable enough to land a rocket like the Falcon 9 propulsively on Earth. This requires extremely low latency, even prediction, much like balancing on a stick. The basic idea an algorithm is simple, but you have to deal with a lot of uncertainty. Buran automated landing is a different thing and more in line with aerospace autopilots. DCX was thirty years ago, used cutting edge technology at the time and didn't come back from anywhere near orbit. Reuseability has largely been rejected as too difficult, costly and risky. SpaceX has had a tough time showing it can work, most of the other companies have not yet succeeded.

bottlepalm 4y ago

I didn’t say it was easy, I said it has been possible. If the investment was made 50 years ago, the ROI for science and technology over the following decades would have been massive. The Shuttle did prioritize reuse to some extent, but the design was frozen and never iterated on. Buran automated the landing and had plans for landable side boosters. DCX also demonstrated propulsive landings. The thing is these were all side project. NASA really never prioritized economical systems as demonstrated by living with the Shuttles exorbitant costs and building SLS. Even though by driving down costs they could of performed orders of magnitude more science. Obviously politics is to blame, not feasibility.

bayesian_horse 4y ago
I highly doubt that 50 years ago there were microcomputers capable enough to land a rocket like the Falcon 9 propulsively on Earth. This requires extremely low latency, even prediction, much like balancing on a stick. The basic idea an algorithm is simple, but you have to deal with a lot of uncertainty. Buran automated landing is a different thing and more in line with aerospace autopilots. DCX was thirty years ago, used cutting edge technology at the time and didn't come back from anywhere near orbit. Reuseability has largely been rejected as too difficult, costly and risky. SpaceX has had a tough time showing it can work, most of the other companies have not yet succeeded.