What good are 53 years of progress?

[Scott]

Donald Bradman scores 452 not out and that was in 1930. Why aren't they all doing it nearly 90 yrs later? 

[Ben the Ignorant]

15 minutes ago, DaveS said:

There was nothing wrong with the engine, it was shut down by a stupidcomputer that glitched, but restarted when they finally got the computer to behave.

So disappointing. In a Cold War or Space Race documentary, they said every change to an essential piece of rocket or missile caused a complete set of retesting. Must have taken untold amounts of man-hours but nuclear missile technology may be more critical than manned spaceflight. Not sure which documentary it was, but I'm positive it was from Jeff Quitney's YouTube channel. He has all the best, with cleaned-up audio and improved video.

Just checked before posting (I'm also big on testing things ?), and Jeff Quitney's channel seems to be out, but PeriscopeFilm has nearly the same content.

[Alien 13]

I have noticed that a lot of missions to the Moon (and Mars) seem to want to change or add something to whats gone before for no obvious reason.

A mission to the Moon could be relatively easy using "off the shelf" parts that already have a proven reliability, I do recall some news programs that showed the design an build of space components by universities and the quality was abysmal compared to what most aerospace/military industries do every day.

Alan

P.S. Why the windows haters on this thread, Linux wouldn't have fired the first stage 

[saac]

1 hour ago, barkis said:

I think the Israelis did quite well considering.        The technology involved is always going to throw a hissy at some time, whichever nation embarks on  these ventures. We have to be thankful this wasn't a manned attempt, although perhaps a corrective intervention could have been performed.

The recent Boeing airliner tragedies Illustrate the fragility of complicated electronic devices. They don't always perform to rigid requirements.  Sad, but true.

Ron.

 

Exactly, had it not been for Armstrong's intervention and decision to ignore the automated landing systems then the Eagle would have aborted its landing.  Flight safety critical code is an absolute nightmare when it comes to safety case assurance. Maybe the lesson is we shouldn't be overly reliant on  this type of technology.  When I first saw the the glass cockpit of the Space x Dragon capsule my first thoughts were wow how slick and at last spaceships looking like the do in the movies. Then I wondered what are the failure modes, what redundancy is there to allow for a more manual control - where is control yoke bell crank    

in fact wasn't over automation a big concern for the Apollo crews. I'm sure it was an issue and that they had to argue their case for manual flight control. 

 

Jim 

[LukeSkywatcher]

Technically the Israelis got to the Moon.

[Ben the Ignorant]

Technically an omelet is eggs. ?

[saac]

3 hours ago, Ben the Ignorant said:

So disappointing. In a Cold War or Space Race documentary, they said every change to an essential piece of rocket or missile caused a complete set of retesting. Must have taken untold amounts of man-hours but nuclear missile technology may be more critical than manned spaceflight. Not sure which documentary it was, but I'm positive it was from Jeff Quitney's YouTube channel. He has all the best, with cleaned-up audio and improved video.

Just checked before posting (I'm also big on testing things ?), and Jeff Quitney's channel seems to be out, but PeriscopeFilm has nearly the same content.

So what do you  think they did wrong Ben?  If you had their chief engineer's telephone number how would advise him/her to proceed?

Jim 

[Ben the Ignorant]

15 minutes ago, saac said:

If you had their chief engineer's telephone number how would advise him/her to proceed?

Trust but verify.

[fireballxl5]

Not directly related, but something I learnt a couple of days ago and maybe helping to recognise the achievements of the Apollo space programme. The fuel pump on a Saturn V rocket engine developed 56,000 hp. That's approx equivalent to all the cars on three formula 1 starting grids, just to supply the fuel needed to run that engine. Think about that as it's incredible.

[ollypenrice]

12 hours ago, Scott said:

Donald Bradman scores 452 not out and that was in 1930. Why aren't they all doing it nearly 90 yrs later? 

Because the blokes who throw the ball at the one with the stick have got better at it as well!  (I have only a rudimantary grasp of cricket as you can see...)

In the case of the airliner disasters it seems that measures were not properly in place to allow the crew to over-rule the software when it did what software does - which is go wrong. There is one, and only one, group of people who believe that it is currently possible to make reliable software. They are called Software Engineers, they are genuinely very clever and they need to be carefully managed by representatives from the outside world...

Olly

[Gina]

I guess I could be called a Software Engineer (amongst other things) as writing software was part of my career, but IMO writing truly reliable software is virtually impossible.  The more complicated, the more chance of it going wrong.  Actually, it doesn't go wrong but does something the author didn't expect because not all scenarios were tested.  It appears to "go wrong" because it wasn't written correctly.  As they say "garbage in - garbage out".

[Ben the Ignorant]

There is a 2.6 seconds (1.3 second to emit plus 1.3 second to receive) radio lag between here and the Moon, so would it be possible to guide a probe manually until the last seconds, when it flies too low and too slow to crash? And leave the last moment to a computer so its task is very brief and more simple?

[Mr Spock]

NASA isn't perfect. They lost the $125 million Mars Climate Orbiter due to some numpty not converting imperial into metric...

[andrew s]

30 minutes ago, Ben the Ignorant said:

There is a 2.6 seconds (1.3 second to emit plus 1.3 second to receive) radio lag between here and the Moon, so would it be possible to guide a probe manually until the last seconds, when it flies too low and too slow to crash? And leave the last moment to a computer so its task is very brief and more simple?

Mind you, I always crashed the old PET Luna lander simulation.

Regards Andrew 

[teoria_del_big_bang]

14 minutes ago, andrew s said:

Mind you, I always crashed the old PET Luna lander simulation.

Regards Andrew 

I remember that. 

Could you imagine giving that to a teenager today and seeing how long they play with that?

Steve

[saac]

1 hour ago, Gina said:

I guess I could be called a Software Engineer (amongst other things) as writing software was part of my career, but IMO writing truly reliable software is virtually impossible.  The more complicated, the more chance of it going wrong.  Actually, it doesn't go wrong but does something the author didn't expect because not all scenarios were tested.  It appears to "go wrong" because it wasn't written correctly.  As they say "garbage in - garbage out".

I agree wholeheartedly Gina.  I have no direct experience of designing software; however, in my former line of work in aviation, I came to appreciate how difficult a process it is building a safety case for flight critical software. Failure modes that the designer never envisaged, nor could have, push risk tolerance limits to eye watering levels.  And of course, no matter how well tested or assured, we cannot guarantee 100% safety or mission success.  Case in point is the current scenario running with Boeing and their 737 Max aircraft.  Boeing is certainly not lacking in technical expertise, capability nor heritage . Boeing certainly know what they are doing, but even they are not immune from failure.  Engineering is not nor never will be a risk free endeavour , anyone who thinks we can eliminate all failures does not understand the nature of engineering.  

Jim  

[saac]

53 minutes ago, Ben the Ignorant said:

There is a 2.6 seconds (1.3 second to emit plus 1.3 second to receive) radio lag between here and the Moon, so would it be possible to guide a probe manually until the last seconds, when it flies too low and too slow to crash? And leave the last moment to a computer so its task is very brief and more simple?

I wonder why they don't just do that   

Jim

[andrew s]

1 minute ago, saac said:

I wonder why they don't just do that   

Jim

If their broadband is as flaky as mine.....

Regards Andrew 

[teoria_del_big_bang]

1 hour ago, Gina said:

I guess I could be called a Software Engineer (amongst other things) as writing software was part of my career, but IMO writing truly reliable software is virtually impossible.  The more complicated, the more chance of it going wrong.  Actually, it doesn't go wrong but does something the author didn't expect because not all scenarios were tested.  It appears to "go wrong" because it wasn't written correctly.  As they say "garbage in - garbage out".

So very true. And what so often happens is later development on already tested software breaks something in the previously tested software that gets missed. The more complicated the software gets it becomes almost impossible to test every scenario and execution sequence.

I guess even today computer hardware can go wrong but generally they are incredibly reliable and it is more likely the software crashes or just doesn't work as intended, so essential human error rather than a computer failure.

Steve

[Stub Mandrel]

2 hours ago, Ben the Ignorant said:

There is a 2.6 seconds (1.3 second to emit plus 1.3 second to receive) radio lag between here and the Moon, so would it be possible to guide a probe manually until the last seconds, when it flies too low and too slow to crash? And leave the last moment to a computer so its task is very brief and more simple?

A human can only react to data sent by computers and is less able to assimilate multiple data streams simultaneously so would be inferior to a computer in that scenario.

Where Buzz Aldrin was better than a computer was because he could look out the window and see the terrain and make an assessment of how rocky it was, he was also able to hold off the abort because (for whatever reason) he was confident better terrain was ahead. The LEM lacked the ability to assess terrain roughness (although a modern probe might) he also had a combination of optimism and the ability to act outside the envelope - this flexibility is at the edge of what AI can do.

In the 737 crashes, again the pilots appear to have been able to see simple information the computer couldn't (bear in mind the root cause appears to be stuck or damaged mechanical parts sending the computer wrong data) and inability of the computer to work 'outside the envelope' (in this case when overridden by the pilots the computer, after a pause, over-rode the override).

My conclusion is that a human is useful in dealing with exceptional circumstances where the solution is to do something outside the expectations of the designers.

 

Incidentally, by your original logic, does the crash of two 737s indicate no progress has been made in aviation since the early days of trans-continental flight?

 

[Ben the Ignorant]

20 minutes ago, Stub Mandrel said:

Incidentally, by your original logic, does the crash of two 737s indicate no progress has been made in aviation since the early days of trans-continental flight?

The title of this thread is obviously a rhetorical question, not a litteral one.

[Paul M]

I'm not a big fan of automation for the sake of it. It costs jobs and lives.

But I was yet again gobsmacked by Elon Musk's adventures on Friday. For the first time in many years I feel that space flight is getting back on track. 

Falcon Heavy and the new Crew Dragon are what should have been instead of the Shuttle. ISS has also been little more than a place holder For the US space program.

Yup, there could have been a McDonald's on Mars by now. Now that'd be progress

[maw lod qan]

19 hours ago, Paul M said:

I'm not a big fan of automation for the sake of it. It costs jobs and lives.

Yup, there could have been a McDonald's on Mars by now. Now that'd be progress

I guess some automation is inevitable, but I want to be in as much control as possible in my car, and with what little flying I've done.

But no MacDonalds. I prefer a BurgerKing. Flame broiled, and that would eliminate some weight issues on the mission. You could reuse the engines, if you could get it to start.