. . . according to a post on this website:
Now, let’s not quibble whether it is 10,000x or 150,000x. That’s not the point of my query.
What I would like to know is how different the Apollo missions would have been if they had had the computing power that we have today? What important facets of the mission would be substantially different. Would it have prevented the disaster of Apollo 13?
As the origin of the problem for Apollo 13 was ultimately a defect in a storage tank and not the computer no, more computing power would not have avoided the explosion.
I’m not sure what difference, if any, it would have made. Well, perhaps we would have sent rovers instead of humans and never actually set foot on the Moon.
Honestly, they wouldn’t have been terribly different. Avionics and modern flight software are far more complex than the AP-101 flight computer used on the Apollo spacecraft, but the developers of that system were ingenous in working around its limitations. The biggest advantage would have been increased ground control and telemetry (using modern frequency and code division multiplexing). The Apollo system was fundamentally capable of fulfilling the mission it was desgined for with the technology of the day, and some (such as Norman Augustine of the Augustine Commission and the eponymous Augustine’s Laws) would argue that with the additional capability of modern avionics and software has also come crippling complexity and enormous costs.
The near-catastrophic loss of Apollo XIII was due to a requirements mismatch and some poor cable routing/insulation decisions. We do have more sophisticated requirement tracking systems (e.g. DOORS and CORE) to do requirements tracking and more explicit standards today, but honestly, we still have the same kinds of problems on programs of this scale, especially with the pressure to use commerical off the shelf (COTS) components. In the history of the STS “Shuttle” there were many instances of similar design and quality flaws that were captured only by highly labor intenstive inspection and test processes. No ‘smartphone’ technology is going to make up for the fact that people are almost always the weakest link in any system.
Stranger
The main difference would have been that the astronauts would have wasted all their time playing Angry Birds instead of doing their experiments.
And the roaming charges would have doubled the cost of the program.
But the selfie that Neil Armstrong took as he tripped off the Lunar Module ladder would have been in color and 7MP.
I have seen some comments that the use of the computer system was an enabler of the programme because it allowed some very significant weight reductions. The critical point is that a computer is flexible, and at different ties could be responsible for very different operations. Everything it did could have been implemented in custom hardware - customised for the particular job. But the weight would have gone up significantly. As it was, those computers were a marvel of miniaturisation, using some of the earliest and most primitive integrated circuits (integrated in sense that a package integrated more than one component into a functional block) and the use of rope memory allowed, for the time, very small reliable storage of the program code.
Early spacecraft control systems could be as simple as a mechanical washing machine controller. Probably the biggest enabler was the invention of the transistor. Before that, control system were relay driven. If you go to a building more than a few decades old and stand next to the lift mechanics at the top of the building you will hear the merry click chatter of the relays that control the lift. Simple ladder logic, these controllers still exist in hundreds of thousands of buildings worldwide. The Apollo crafty could probably have had a sizable fraction of their operations controlled by bespoke ladder logic (probably implemented in transistor logic). But a programmable computer has the flexibility to do it all, and be much smaller, and be reconfigurable/reprogrammable if the mission needs change or a mistake is found. The ability to delay freezing design until much later helped allow accelerated pace of the while programme.
During Apollo 13 the computers were turned off for most of the mission, only being brought up for re-entry. All the course control and correction burns were managed by hand, and timed with a wristwatch (the famous Omega Speedmaster.) It all worked out. Under ordinary operations the computer would have handled such tasks. Actual calculation of the dynamics was done on the ground, in the trench in Houston.
One neat thing about the Apollo computers. Even during the moon missions, NASA was working hard on its other job - the aeronautics bit. The very early fly by wire work was proceeding, and they realised that the only computer that would be small and light enough to conduct their trials was an AP-101. A time critical part of doing this was that the rope memory facility was to be closed down when the Apollo mission computers were done. Each mission had different program code (the code was so tight that the entire mission profile was embedded in the code) and the program was literally woven into the rope. The fly by wire programme had to get their experimental code ready in time to weave into a rope memory before the facility was shut down. The programme used a lot of Apollo hardware, and when one of their DSKY input units failed they scavenged one from a returned Apollo capsule.
There is a famous document commonly called “Famous Bugs” that’s been floating around Usenet and the Internet since at least as far back as mid-1980’s, that’s fascinating reading for anybody interested in stuff like that.
Here’s is one link I found. If you have trouble getting to this link, just google Famous Bugs or maybe Famous Computer Bugs and hope to find others.
http://www.textfiles.com/100/famous.bug
Search through this file for mentions of Apollo.
Here is one tidbit:
Here’s another:
Gives you a sort of idea how things were done in those days.
The story that it was a MIT summer student is fanciful and just plain wrong. Probably invented by an MIT student. Does anyone seriously think that NASA had summer students wandering around mission control during the first moon landing? The call was continue with landing or abort.
The person who made the call was no summer student. He was Steve Bales, the engineer in charge of the LEM systems in mission control during the descent. Earlier he had taken the time out to study the entire set of computer alarms, and understand how the system would operate in the face of error conditions. (This was after all part of his job to understand this.) During the landing the alarms did start to go off. The reason was a change to mission profile, and a tiny change to the use of the rendezvous radar (which tracked the CSM) during descent. They decided not to have the radar running, but didn’t realise that just turning it off was not enough - they had to stop the computer from listening to it. The computer ran the rendezvous radar sub-program whilst sucking on noise, rather than real signal. And indeed, took longer than designed, resulting in the real time control executive in the computer running out of time to perform all of its tasks - something that if it had persisted would have required a mission abort.
When the mission was over, four men were awarded the US medal of freedom. The three astronauts, and Steve.
This is not to diminish the critical role that guys in the backroom, Jack Garman of NASA and Russ Larson of MIT (so there is an MIT connection) - who were on a continuous voice loop with Steve - had in providing backup to Steve’s decision.
A very complete paper on the LEM systems and this incident is here.
The “summer student” is probably a reference to Jack German, one of the engineers working for Steve Bales and who supplied key information that allowed Steve to make the call. He was 24 years old at the time, but he was a NASA employee, not a summer student.