A Boeing 777 Has Crash-Landed At San Francisco International

[Cat Driver]

There have been a few major shifts in CRM ideologies over the years.

It would stand to reason that aviation would evolve as time passes and the learning curve goes up.

I comment on this subject from the position of having actually experienced the evolution of flying over the last half century actually flying these aircraft....

......how much PIC time do you have in transport category aircraft CID?

[CID]

Cat Driver, how much PIC time do you have on wide bodied aircraft? Or transport jets? Or FBW? What's the minimum experience you recommend before people can comment? On an anonymous forum? On the internet?

[Cat Driver]

CID you are free to comment all you want on this anonymous forum because with anonymity one can say anything they wish within the rules of the forum.

As to my understanding of fly by wire aircraft I worked with Airbus Industries for two years and received training on their technology from five of their factory instructors at their facility in Toulouse...and I am confident enough with who I am that I post using my real name..... ..

You and I have a long history starting from the Air Play days and I am quite aware of your anonymity phobia, but every once in a while I get some pleasure from reminding you we are very different.

[bobcaygeon]

I wonder what the conversation would be if we turned our attention to another accident on a much less automated aircraft. The L1011.
Look up the Eastern 401 accident in the Everglades.

Don't have to look it up, I remember it quite clearly - including the alleged apparitions and the disembodied head in the galley oven. 401 happened at night over a blacked out Everglades while 214 was at high noon on a clear day but the results were the same, a flyable aircraft crashed with fatal consequences. In either case, nobody was minding the store. The Eastern guys were so concerned with troubleshooting a burnt out gear light no one noticed the a/p altitude hold had been disconnected while on the 777, three pilots were unaware of their airspeed trend. FTFA.

Eastern culture is mentioned many times in this thread but do we all forget that this used to happen here too.
Eastern 401 had a well trained ATC controller that did not clearly challenge the crew when the a/c was descending into the everglades. Why was that?? Was the old school military culture at play? The end result was the same when the controller took his own life. Sound familiar. The shame is that other cultures are not learning from our mistakes but just repeating them. Not surprising as we continue to repaet our own mistakes over and over again. Just not in airplanes.

[SheriffPatGarrett]

It's really very simple.

Operating out of Miami to South America, I found weather to be severely CAVU 95% of the times(opposed to mostly the reverse in Canada)

Well, half of the guys pulled their ball caps down and acted as if a check airman was on the jump seat, hand flying the whole procedure
all the times, just to be ready for the other 5% when the crap hit the fan(or the simulator)...while the other half sleepily glided on with the auto-pilot
and on visual.
Guess who usually got in trouble!

As for Cali, Colombia, I've been there several hundred times, I knew that procedure was on the book, but I never heard of anybody ever using it.
Fact I even asked ATC and they did say that nobody did...the way to go was go over the airport then turn around back north in the very wide and flat valley.
To go down in that gut with close by vertical walls...Beside, it was a VOR approach, not an NDB!!! NDB are notorious for steering planes into mountains!

The proper one:


The bad, bad suicidal one:


Well AA had just bought in these routes from Pan Am or Eastern, I forgot which and it was a newbie's mistake, newbies
on IT and no situational awareness...These American Airlines guys used lots of over-the-road trucker CB linguo,
which left the latinos ATC completely in the dark.

It was also a perfect example of automation run amok.

Or maybe it was just a case(for us) of: "Just say no to Mach Eight-Oh"...(we were paid by the hour!)

[pdw]

... I believe what you are referring to is what happens when a swept-wing jet gets on the back side of the power-curve. It rises sharply below L/D max and you really don't want to be there!

So, ... "sharply" is the root difficulty. Not seeing the problem evolve in shorter duration is the root of the problem (with the list of distractions).

[CID]

@Cat Driver

You're a one trick pony aren't you?

[complexintentions]

So, I may suggest that until you have experienced those types of applications in order to give you a direct comparable, you won't understand how easy moving millions of passengers worldwide in a highly controlled environments really is. It's just not that hard, and that three pilots who have been working in that environment for years can still bollock it up that badly, tells you a great deal about what is actually required on a daily basis. That's not to say there aren't good pilots flying automated airliners, it's just that it sure doesn't seem to be required, statistically.

I must admit, I'm beginning to fail to understand how the word "Pilot" applies in modern airliners anymore

STL,

Could you please detail the relevant experience you have in "moving millions of passengers worldwide in a highly controlled environment" that grants you the authority to make these sorts of statements? I mean that sincerely, because unfortunately this just sounds more like rotary-wing machismo than considered opinion. I'm sure you would tend to discount my opinion as well if I were to make similar observations about niches of the industry in which I had little or no experience.

The "three pilots who have been working in that environment for years can still bollock it up that badly" belong to a sub-set of the aviation culture that is very, verrrrry different than that which most on AvCanada are familiar with. It's telling how virtually NO ONE who has been exposed to Far East or similar operations is terribly surprised by the accident or confused as to its reasons. Unfortunately the Asiana crash has inevitably led to a few using it as a reason to bring out their own rants for or against automation, or to lament the general decline in piloting ability, or the whole depressingly predictable "airline pilots as button pushers", "only ________ (fill in the blank to taste) are REAL pilots", etc etc etc.

Never said there wasn't a skill set required. What I said was, that skill set is dramatically different to the traditional pilot's skill set.

This is about the point your whole argument falls down, in my opinion. If anything, Asiana demonstrated that the current skill set required ISN'T that different at all from the traditional pilot's skill set - other than being greatly expanded.

"Pilots" - or whatever your obsession with labelling wants to call us - still need to be able to fly the airplane. And to understand the systems - vastly more complex in their underlying function, if not their operation, than anything in the past, but still retain the manual skills. And unfortunately it isn't just about changing training, it has to do with changing mindsets and attitudes - which in places that have cultures dating thousands, not hundreds, of years old, is very difficult.

I spent thousands of hours hand-flying much less complex machines than the B777 in years past, and I can tell you it was still a lot simpler and easier than acting as PIC on a widebody.

[Flying Nutcracker]

^When I read this I almost grabbed my Bose ANR headset, because i hear yelling!

[CID]

If anything, Asiana demonstrated that the current skill set required ISN'T that different at all from the traditional pilot's skill set - other than being greatly expanded.

??

[Siddley Hawker]

Comment on another Asiana accident.

http://www.pprune.org/jet-blast/519037- ... orget.html

[grimey]

If anything, Asiana demonstrated that the current skill set required ISN'T that different at all from the traditional pilot's skill set - other than being greatly expanded.

??

I'm guessing he mean "required in order to safely operate the aircraft" not "required by Asiana or other airlines". When shit hits the fan, I would hope that a pilot sitting at the front of a plane I'm on has the ability to fly it if all the fancy electronic aids go for a shit. Several recent accidents have apparently demonstrated that this isn't always the case.

[Meatservo]

What he means is that the accident shows us how not being able to handle the aircraft all that well can get you into serious trouble, in a giant jet as well as a Cessna. It shows us that despite all the goodies in the cockpit the aircraft met its demise simply because of inept handling, which should demonstrate that traditional aircraft-handling skills are still important no matter what plane you happen to be flying. The automation adds safety when it's used on a solid base of good airmanship and flying ability, and can cause chaos when it isn't. At least that's what I read into this. Automation is there to make good pilots more capable. It wasn't meant to be used as a crutch for poor pilots to stay safe: but sometimes that's what it's actually used for.

[sky's the limit]

Complex,

I find it astounding how fragile the ego's of some you heavy iron pilots really are.... fascinating.

Oh, wasn't aware I was the only one who sees millions of people moved around the globe in abject safety each day? Guess statistics don't mean much anymore. And I really don't care if you comment on every niche market in aviation, everyone else here does... Certainly didn't stop people spouting off when COugar lost their S92, did it?

I'm on days off now, have fun guys!

Oh, and I guess that "sub-culture" can be applied to the French too... and Canadian pilots as pointed out by several other posters... You lost me at the cultural profiling.


stl

[CID]

Interesting. What I read in to that statement (albeit out of context of the rest of the post) was that it was contradictory. It's the same....but different.

With respect to automation, it's more complex that making good pilots more capable. In many cases, as with the 777, it makes flying it "possible".

Some automation is "nice to have". Other automation is absolutely required for a particular aircraft to be operated within the prescribed level of safety. That's a difficult concept for some to grasp but he airworthiness standards are written with the goal of normalizing the risks associated with flying. (Commercial flying....not so much private)

In theory, buying a ticket on a 777 to fly you half way around the world should have the same risk associated as paying to ride 100 miles in a J-3 Cub. The design and operating standards generally become more stringent as the airplane size, weight, performance and other capabilities increase and sometimes the only way to do it is with automation. Sometimes (as in the past before advanced technology was available) you add crew members. Or engines. Or redundant systems. Or sometimes by imposing limitations or restrictions. Take Commercial SEIFR for example.

A huge piece of the puzzle is pilot training for proficiency in operating THE aircraft not ANY aircraft. Some of the skills and concepts overlap but I really think we need to move away from the notion that there is a logical and mandatory path from J-3 to 777. There doesn't need to be.

So in a strange way, I'm agreeing with complexintentions. These days training MUST be focussed on the aircraft you flying BUT there will be some overlap no matter what model it is. Is it any wonder that airlines limit the number of different types you can fly to 1 or 2?

[Meatservo]

CID, I think that under the banner of "automation" there are several different types of devices, but when we talk about automation in the context here, we aren't talking about all of it. In particular, you are talking about the kind of automation that makes it "possible" to fly the aircraft at all. I'm assuming you mean things like stability augmentation and the devices that manage the engines so they can be started with the flick of a switch and controlled with just one lever. As necessary as these things are, however, I don't think that philosophically they are particularly relevant in the context of this discussion. I think in this discussion the type of "automation" being referred to is that which provides flight-path guidance, replacing the control inputs of a pilot using a traditional stick-and-rudder pedals arrangement. The background stuff is just there, the same way that our cars no longer have spark-advance levers and primers and chokes or clutches, all replaced by "automation", yet a bad driver is still able to steer the thing into a fire hydrant because he is texting instead of looking out the window, or because he takes a corner too fast.

I think we should confine our discussion of manual controls VS. automatic controls to that automation which directly replaces a pilot's hands and feet on the actual flight controls for this discussion. In that sense, inept handling is still a danger regardless of the level of automation in the aircraft.

I guess my original point was that I believe that with a well-trained crew (including the ability to manually provide flight-path guidance using the actual controls) the automation excels as the first line of defense against external threats to the safety of the aircraft. With a less able crew, the automation is reduced to being the last line of defence against internal threats to the safety of the aircraft, posed by the crew itself. Which do we find more laudable?

[CID]

Meatservo (awesome handle by the way) I see your point. But...stick and rudder controls on a 777 are there so that the pilot can take over from the autopilot and fly it manually. Not like in the conventional sense where flying it manually is the default and an autopilot is added so the pilot can let it fly automatically.

In order to allow the pilot to push aside the automation of the autopilot, the engineers had to add a whole new level of automation. Things like artificial feel and three Primary Flight Computers (PFC) that apply the active control law.

In layman's terms, and using the automobile analogy, the driver can still turn the wheel in an attempt to crash into the fire hydrant but the car may not let him (or her). To be more accurate, the Boeing FBW philosophy wouldn't limit the steering wheel to avoid hitting the fire hydrant, it would impose "protection" that would make it difficult for the drive to hit the fire hydrant unless he absolutely wanted to.

That fundamental difference between the Airbus FBW and the Boeing FBW concepts often cause pilots to state that Boeing pilots can always override the flight control computer but that it's not as simple as that. That being said, if the Airbus FBW was in our ill fated fire hydrant hitting car, it wouldn't "let" the driver do it. Even if he really wanted to.

The protections integrated into the control system on the 777 include (from Boeing):

• Bank angle protection
• Turn compensation
• Stall and overspeed protection
• Pitch control and stability augmentation
• Thrust asymmetry compensation

So there is a great deal of automation in the "Normal" manual control mode. Sure, there's a "Direct" mode that allows the pilot to bypass the flight control laws, but that is VERY VERY contrary to the recommendations and standard operating procedures.

One of the design goals of the system (from Boeing) was:

Retention of the desirable flight control characteristics of a conventionally controlled system and
the removal of the undesirable characteristics.

Can that be done if the controls were directly connected to the control surfaces?

[Meatservo]

I don't mean to be argue-y, but this is actually pretty interesting. Your first point, "..stick and rudder controls on a 777 are there so that the pilot can take over from the autopilot and fly it manually." This still assumes that the pilots are able to judge when manual flying is appropriate and also assumes that they will be competent to do so. Fair assumptions, I think, so the philosophy behind installing automatic flight controls remains the same regardless of the intended ratio of manual vs. automatic flight. When the automatics are running, the assumption on the part of the manufacturer, airline, and passengers should be that the pilots are always "flying" the aircraft, at least in their heads, otherwise how can they detect an anomalous energy state /heading/configuration in order to do something about it? The level of automation involved in making the aeroplane "nice" to fly for the pilots is really just an extension of all basic design philosophy, that is, nobody deliberately builds an aeroplane so that it will be HARD to fly! The stability augmentation, "feel", stall warning and protection, turn co-ordinating, etc, is all just electronic versions of what designers have been trying to do all along with sweepback, dihedral, roll spoilers, stall strips, stick shakers, large vertical surfaces, weight and balance limitations, springs, and so on. Even if all of this stuff is now so sophisticated that it will now add power or lower the nose if a ham-fisted pilot pulls back too far on the pitch control, or limit bank angle or whatever, experience shows us that people have still crashed these aeroplanes through a lack of airmanship. I remain convinced that no matter the level of sophistication of the airframe, if the crew is not competent to safely maneuver the machine using the built-in flight controls, however circuitous the route between said controls and the aerodynamic surfaces, then the whole building is built on a crumbling basement. I'm not trashing automation. I'm just saying that automation was always, or should have always been, intended to ADD to flight safety and efficiency, built over top of a solid foundation of airmanship and flying ability, NOT to replace it. The fact that the technology is so useful as to camouflage the risk imposed by a degradation in basic situational awareness, airmanship and physical skill should be identified as a source of risk in and of itself.

[Beefitarian]

In a 172 the flight controls are "yoke, cable, ailerons/elevator."

In a fly by wire plane they are, "Yoke, sensor, electrons- wire, computer, wire, motor (servo even) rods/mechanisms elevator/ailerons.

I have wrote this before.
You don't fly the plane, your control input asks the computer to fly the plane the way you want it flown.

I understand that it happens fast enough to make it seem like flying any other plane, but it is another factor that a pilot must sub-consciously adjust for while "hand" flying them.

[Meatservo]

Depends what you mean by "fly". Asking a computer to do something means that you did it. The computer has no will, and as I commented in another thread, is just as happy or sad being crushed into a ball along with your organs and self-esteem as it is cruising through the tropopause. In the plane I fly at work, when I push on a rudder pedal I am "asking" the hydraulic system to open a shuttle valve momentarily until the rudder moves in response and closes the valve again. The rudder moves because I ask it to. It moves so quickly that I feel as though I moved it. The aircraft yaws in response. Perhaps I would be safer if there was a computer installed that was able to limit the amount of pressure I could send to the rudder in order to prevent me from crashing the aeroplane if I commanded a full deflection at too high an airspeed. But if there was a mountain in that direction it wouldn't matter if the computer didn't allow me to rip the tail off with my bumbling pedal application, the plane would still turn and hit it.

Do you suppose that if an executive committed some kind of business fuckery using his computer, and was later caught, he could argue that he didn't really do it, he just "asked" the computer to do it? Perhaps we could prosecute the guy who made the software, or maybe we should program business computers to recognise ill will and incompetence and thwart it? The computer has no will, and when you "ask" a computer to do something, that means YOU are doing it. Currently aeroplanes act as an extension of the pilot's will. When the pilot lacks situational awareness, airmanship, skill, and moral authority, the computer is still more than happy to ride along with him to his death.

[Beefitarian]

Depends what you mean by "fly". Asking a computer to do something means that you did it. The computer has no will, and as I commented in another thread, is just as happy or sad being crushed into a ball along with your organs and self-esteem as it is cruising through the tropopause.

Absolutely and I hope my position is clear that. If the crew/pilot needs to disable the auto pilot. They need to be capable of "hand flying" the plane they are in.

In the case of a fly by wire plane there will be technique that might even mess with the colonel unless he is current on the particular craft. I think we can all admit he knows how to make a lot of airplanes do exactly what he wants. Wether or not we enjoy his movies.

Just think if the crew is made up of a guy like me that has never flown many types, or a guy that has not flown the one they're in much and another pilot that is out of practice because he rides around too muchreading magazines.

I think most of us can imagine that likely to be bad.

[CID]

The computer has no will....

Au contraire. That is my point essentially. On a 777 under Normal Law, a simple control input doesn't result in an instantaneous control surface deflection. The "will" of the engineers that designed the system is imposed.

The actual deflection of the control surface(s) and the feel of the control input are the result of the overall flight control algorithm. A pilot "flying" a 777 is almost in an artificial reality relationship with the aircraft in that the reaction of the aircraft is somewhat normalized. Kind of like playing with a flight simulator with the crash and overstress detection turned off. (Yah...that's a real stretch but I think it makes my point)

Now let's talk a little about the executive at his desk. You're implying that computers are nothing more than an electronic typewriters. That's not the case. Imagine having a spreadsheet that the company uses to manage investments. Now lets imagine that one of the spreadsheet formulas has a mistake. It could result in the incorrect calculation of a very critical number. Would that necessarily be the fault of the executive? Wouldn't there be other contributing factors like the spreadsheet error?

And how about those guys who sit behind a computer analyzing crack growth rates on the skin of the 777? Any chance the program they use to crunch the numbers could have an error and the engineers fail to detect a potential catastrophic failure? Yes. Of course there are usually a few levels of oversight and error detection in complex design programs but computers take on the personality of the software they are running and any work (or play) you execute on them is modified and/or controlled by the "will" of the programmer.

[pdw]

The computer has no will....

The "will" of the engineers that designed the system is imposed.

Their "will is imposed" on the design of the software so it does what it needs to do in the situation its designed for.

None for the computor ... just the software .... until a mechnical problem

[CID]

pdw, true. But without software, a computer is just a good doorstop. Or boat anchor.

[amraam]

Photos of the aircraft being removed from the crash scene:

http://www.nycaviation.com/photo-galler ... rash-site/

amraam