Pilots Really "Six Figure Professionals"?

Doug Taylor said:
So whenever the check airmen says, "Using your systems knowledge, work out this problem" he's actually doing you a favor helping you think outside of the box.
Click to expand...

Of course using system knowledge and thinking outside of the box got a lot of people in a lot of trouble in the past. Which is why nowadays the emphasis is on flying the airplane, following procedure, isolating the problem and flying the airplane.

On another thread: Putting the go/no-go decision into the hands of automation is not someting I'm interested in. Don't misinterpret automation as this magic box which has the big picture which knows how to fly an airplane and the percentages of risk involved with aborting or continuing the takeoff roll.
Click to expand...

Of course that's exactly what magic boxes do best. The whole history of aborts and outcomes can by analyzed and the best course of action deduced for a given scenario. All with no time or emotional pressure. Then when the box recognizes the scenario the correct response comes out while the pilot is mid "what the.......".

Automation is a box programmed by engineers which may or may not understand the dynamics of flight and will execute responses programed by engineers which may or may not understand.
Click to expand...

Again, it's a false argument if you talk about the failings of automated systems and not the dismal record of line pilots in this area. And I'm not faulting the pilots. It's impossible to make a rational decision in the split-second you have on an abort. Not enough time, not enough information, not a fast enough processor and inevitably emotion comes in. The good news is this one can be pretty easily solved by future technology.

Remember the Airbus crash? Why would a plane fly so low and no want to land? Adding power or increasing back pressure isn't consistent with an airplane in the landing phase so lets go ahead and override the pilot's response consistent with a safe landing.

WHAP!
Click to expand...
Not only is this old technology, very old, but it's partially French. So not a good example.

Or you're accelerating through about 70 knots and there's a flock of Canadian geese not paying attention and an antelope traversing the runway during takeoff roll in a DC-9. Do you continue? Do you abort? What would the automation do with what it can't see? What if the automation *could* see it, what would it do? If you hit the animals, had an engine fire and aborted, would it command evacuation? Or does the captain determine if it requires evacuation?
Click to expand...

Don't see any problems with designing systems to handle all those scenarios. Certainly no problem with automation seeing obstructions ahead (see self-driving car link above). Are you saying captains make consistently correct evacuation decisions? I don't think the data would support that conclusion.

Relax Doug. This is about after your career and after I'm dead. So you can take some solace in that. If I'm wrong be kind, remember I'll be dead.
 
Doug Taylor said:
If you're doing on off-airport landing, do we really think we're going to have the automation to help us survive?

How did all that glass help when pilots deadsticked the CRJ in Missouri?
Click to expand...

Sure. DARPA is doing it in vehicles. It cannot be too long before computers will be able to accurately see and think for themselves. When that happens, they will be more efficient and consistent than their human counterparts.

http://www.grandchallenge.org/
 
Speaking of automation.. anyone seen the cockpit of the new Eclipse? It has about 10 switches and thats it. I think a C172 has more.
 
I'm with Doug on this one, guys.

Computers can't see.

Computers can't hear

Computers can't smell

Computers can't think

They never will. If we design something that can do this, then we're not talking about a computer. That would be something completely different than what ya'll are dicussing. As of today there is no algorythem for language, computers can't see things for crap (they have problems seeing things unless they have real good defined lines and contrasting colors).

And in the auto land example, we're just talking about flying an ILS down to the ground. In the off airport landing example we're talking about a computer trying to see cows, kids, farmers with shotguns, fences, potholes and a bunch of other crap. Then it has to take that information from the analog world and make it digital, THEN it can process it. We don't have to do that which gives us a huge advantage.

This kind of crap is never going to happen. We're not talking about the space shuttle which has little to run into. We're talking about a VERY confined environment to be operating within.
 
Guys, for me this has nothing to do with ego or being defensive about being replaced by computers.

The automation can fly the airplane much better than I can (except for landings where it stinks...and the ILS during signal interference...which is 90% of the time).

My point is that an airline pilot makes all kinds of important decisions on every flight...and is the "CEO of the Airplane" (my apologies for stealing Keven Garrison's article title). Checklists, automation, policies, regulations, procedures, etc...are all aids to making proper decisions. Just because a decision is easy...doesn't mean it's not critical. Some decisions have been planned out ahead of time, some haven't. Some are easy...some are difficult. They are still decisions to be made.

The challenge of being an airline pilot isn't really the technical aspects of flying the airplane. It is but one aspect of airline flying. The challenge is to ensure that the operation runs safely, comfortably, with high regards to peak customer service within the confines of safety. It is coordinating different corporate departments, government agencies, customers, the airplane, the crew, the environment...all which claim a part of each airline flight. All within a very dynamic environment. Just because the decision to abort has become automated...doesn't diminish the managerial role that still exists...I don't find that role extinguishing even with future automation. Maybe it will...who knows.

But automation...current generation...and the generation over the horizon...shouldn't diminish the executive role of the crew. The oversight of the operation is a challenging game...and one that airline pilots should welcome. Just as we should welcome automation as a crutch to improve our performance.

I agree with the premise that automation improves safety...and helps the decision making process. If the V1 decision can be improved in an automated fashion...I welcome that. But the oversight of the entire operation...is still very important and so multifaceted and dynamic and variable...that the human is the best and most efficient way to complete. The pilots are involved in untold decisions surrounding the oversight the entire operation.
 
John Herreshoff said:
I'm with Doug on this one, guys.

Computers can't see.

Computers can't hear

Computers can't smell

Computers can't think

They never will. If we design something that can do this, then we're not talking about a computer. That would be something completely different than what ya'll are dicussing. As of today there is no algorythem for language, computers can't see things for crap (they have problems seeing things unless they have real good defined lines and contrasting colors).

And in the auto land example, we're just talking about flying an ILS down to the ground. In the off airport landing example we're talking about a computer trying to see cows, kids, farmers with shotguns, fences, potholes and a bunch of other crap. Then it has to take that information from the analog world and make it digital, THEN it can process it. We don't have to do that which gives us a huge advantage.

This kind of crap is never going to happen. We're not talking about the space shuttle which has little to run into. We're talking about a VERY confined environment to be operating within.
Click to expand...

Check out the link I posted above. The vehicles in the grand challenge had only GPS waypoints loaded. They navigated the course by "SIGHT".

It really will not be that long now. It will probably take longer for us to fell "comfortable" with no pilot in the cockpit, then the technology to develop.
Certainly, much like the nav and FE have, the co-pilot will be leaving the scene sooner or later.
 
Seen the data before and I interpet it WAY differently than you do because of my background of being employed to tame these silicon pieces of crap.
 
anyone seen the cockpit of the new Eclipse
Click to expand...

yes its amazing!! So much glass. im not fond of the joystick thing. i like the good old yoke.

you have to acknowledge that pilots often blow these types of scenarios.
Click to expand...

indeed. But, the computer cannot make ethical decisions. this could be good or bad in different situations.
 
boeing 777-300 said:
Then why havent we developed smell-o-vision on the Food Network??
Click to expand...

Because people would discover how much of a sham Emeril Legasse is.

If you've got to narrate your show like an episode of the 1960's Batman, give it up yo.
 
While I don't think that we are that close to seeing this as reality, there are a couple points to make:

1. Not only did that DC9 crew choose a bad landing choice on that road, but they also shouldn't have gotten into that situation, as it was a direct result of misinterpreting the radar!

2. The Airbus "into the trees" was more due to the intentional override of the features of the system so they could do fly by the way that crew wanted to, then a failure in the design of the system itself. That is not to say that I think Airbus does everything perfectly, but the data analysis does show that their "hard envelope" protection works and is the better way to go, and this has been admitted by Boeing flight safety folks in my presence.

3. Wish I could say I could land better than the MD-11 autopilot, but the most I can say is that I can land "as good" most of the time, and every once in a while get one that is just slightly smoother, but that is rare. It does an impressive job, crosswinds, etc., and all. Really irritating, it is!

I am not sure that we have reached the point that computers could handle every scenario as well as a good pilot can, but, I have to say that we probably would reduce the accident rates overall due to the poor decisions that still happen too often, and that is the bottom line.

Still, I think we are a very long ways off to seeing the public buy into it at any level. Very long ways. So, as that is really the driving factor, even for freighters (for a variety of reasons), this is probably not on the list that you have to worry about!
 
flyover said:
Can so.



Can so.



Can so.



Can so.
Click to expand...

You drive a 747 with steam gauges, give me a ring next time you're back in a server room trying to get a printer server to behave
 
Well, all the flight control stuff needs 10^-9 reliability. Some parts can get by with 10^-6. Systems that you can get by with secondary means can go even lower (which is what happens with the FMCs, for example). How often does your PFD fail, or, more to the point, fail such that the redundant systems are not adequate? I can think of only one time that I know of, and that was on a well publicized transport LOC event. However, to be fair, the failure occurred directly due to the engineers not understanding the big picture on aircraft dynamics (they were not flight test engineers, obviously!). Flight test pilots/engineers, had they been in the loop on the particular decision that led to the failure mode, would fixed it, as it is fixed now!

Ground based computers, even for corporations, have no such requirements. As a result, they are often pushed beyond what they can quite do, as marketing drives more functionality (or, more often, just "bells and whistles" that look good on the advertisement). Obviously, any sort of aircraft control architecture would have to be quite a bit tougher!
 
So how many planes you want to crash as engineers fix all the bugs? You need a stop gap in there to fix their mistakes; aka a human.
 
John Herreshoff said:
You drive a 747 with steam gauges, give me a ring next time you're back in a server room trying to get a printer server to behave
Click to expand...

LOL. One of my sons would tell me the same thing based on his experiences.

Obviously almost none of the technology we all have experience with has anything to do with what we're talking about here. Even the next generation of jets, like the 787, will just be a bridge to that technology. It's the military that will lead the way and fund the research in this area and then prove the technology.

And we won't flip a switch and go from total human oversight to total computer control. But by the time we get to that point there won't be much question about it being safer, more efficient and reliable.
 
Well, I can assure you that I have seen the actual numbers that show how many MORE people would have died absent the hard envelope protection that the FBW airbus have. Engineers have made many errors over the years that have resulted in accidents, and the example I related is really not so different than errors that engineers have made that have resulted in structural failures, power failures, instrument errors (which this was, technically) and others. As long as humans design the aircraft, you will have things happen. However, the number of fatalities due to pilot error still exceeds those.

I am not saying that I am pleased with the facts, it's just that I am not going to deny them.
 
You must log in or register to reply here.
Back
Top