Post #400,435
3/26/15 3:12:43 PM
|
You're in-lurve with modrin TLAs like "CRM"--as if That 'cooks the rice'
On a TANDEM control pair you need no slowed-down/intellectual-palaver-filtering of words: to grasp: EXACTLY what the other person is attempting to do ... as. it. happens ['it' being: his/her exact inputs to the attitude control System.] You can't get this information Quicker nor more accurately.
Why you persist in calling this irrelevant--have you ever 'flown' anything?--in face of those who regularly count on Mr. Bernoulli to counter the effects of Mr. Gravitational Attraction ... remains {sheesh}-grade puzzling, here.
Conflate abstract 'Stats' with ... the visceral-experience of an actual process within real-time ...one more time? and I'll approach Her Majesty's Master-Debater Accreditation Service about some demerit Awards.
Maybe we each inhabit separate | parallel universes and that's why communication always was doomed, species-wide (?)
|
Post #400,438
3/26/15 3:22:40 PM
|
Thought experiment
Suppose the evidence clearly showed that fly-by-wire was conclusively and significantly safer than manual control of flight surfaces. ("Safer" defined as "less likely to crash.") In the case of a single pilot, an electronic control stick is the correct setup.
When there are two pilots, who in theory should be taking turns, you have the possibility that they are trying to do different things. They shouldn't be doing that at the same time. With manual linkages that can't happen unnoticed. Is the likelihood of that failure mode greater than the increased safety of fly-by-wire?
Sure, it "feels" wrong to imagine pulling up on the stick and it doesn't do anything because someone else is pushing down and you don't know that. But if that happens, you know that half the people in the cockpit are trying to do the wrong thing. In that case I just might prefer that the "expert system" figure out what to do.
|
Post #400,442
3/26/15 3:45:21 PM
|
Excellent point; shall need some pondering..
but for the moment, the massively-nested algorithms are unfathomable by human jelloware: in any realistic time frame for decisions.
We have to rely on homo-sap final Deciders, for an unknowable time-frame.
While we're awaiting the First truly comprehensible Fail-Safe Source Code, I aver that: FEELING what is happening, immediately! beats all [word]-exchanges: for deciding whether
you need to cold-cock a loonie sitting next to you? and fly this thing out of trouble or ... just TELL HIM/HER why you are bloody-sure it is a Mistake.
And then cold-cock if you get back gibbering ƒeare-filled nonsense.
We may get "there"; we sure Haven't Yet.
|
Post #400,446
3/26/15 4:08:02 PM
|
Look at the latest research on self-driving cars
They're finding that for years we've been solving the wrong problem. We've been trying to take people completely out of the equation until we need them: construction zones with poor/inaccurate markings, unusual weather conditions, unexpected mechanical failure, etc. What they have figured out is that people are really bad at maintaining alertness when completely disconnected from the process.
What works better is a drive-by-wire system where the driver makes general inputs - go, stop, lane change - but the car decides what to do with the engine, brakes and steering. Computers are better and faster than all but the very best drivers at this level of control.
Fly-by-wire systems seem to have learned half that lesson. The pilots provide input saying "go higher" or "turn left" and the system decides what to do with the flaps and engines. But auto-pilot has been in use for so long that we've come to expect it. Maybe it would be better to leave the fine control to the systems but make the pilots actually fly again.
|
Post #400,443
3/26/15 3:45:48 PM
|
OH NOES A MODERN TLA! IT MUST BE SHIT AND RUBBISH
CRM was developed in the wake of the Tenerife crash (two Boeings, fact fans!) and other fleshoid-based fuckups.
In 1979.
"Modern".
Heh.
|
Post #400,455
3/26/15 5:07:06 PM
|
Yep, two Boeings wrecked by a European.
|
Post #400,457
3/26/15 6:43:39 PM
|
And this is why you fail.
|
Post #400,503
3/28/15 8:05:05 AM
|
(The CRM idea isn't bollocks, of course) But as a deflection of this issue:
it is. Yah, know about that Nederlander #1 pilot's self-cleared take-off ..and the wimpy PNF who didn't say NO!! ... thus everyone, mostly, died.
We were talking about, not crew sociology generally: but about the Value / Or Not, of instant feedback when two pilots are (hoping they are..) helping each other fly the plane, while under great stress. And, (as re the universe/the whole-fucking Cosmos and Everything ... you were explaining how you already had all-that-flying-stuff permanently Sorted, too.
And I was saying: when you are that presumptuous, and essaying Certainty of such stuff too: you are daft.
So maybe no next plane will stall out exactly as ... when a CRM-certified loon like Bonin was killing F. 447 all. the. way. down. Maybe. But with that side-stick still permitted.. any compulsive Still Could. And, freaked-out? fucking-May.
Ta ... next we take on the detection of secret-psychotics with epaulets and enough daily-patter to seem 'normal'. Easy-peasey problem, eh?
|
Post #400,444
3/26/15 3:45:49 PM
3/26/15 3:47:53 PM
|
Dupe
Edited by pwhysall
March 26, 2015, 03:47:53 PM EDT
|
