F***ing unbelievable!

[Black Ram]

WOW. This is freaky.

I think it kinda shows that true fail masters are adept at misusing both sidesticks and yokes.

Kudos to the relief pilot - an astute observer, unlike capt Dubois, and persistently vocal in his remarks, unlike f.o. Robert.

[3WE]

WOW. This is freaky.

I think it kinda shows that true fail masters are adept at misusing both sidesticks and yokes.

Kudos to the relief pilot - an astute observer, unlike capt Dubois, and persistently vocal in his remarks, unlike f.o. Robert.

...nothing to do with the planes control philosophy huh?

I pull back on the stick and it's supposed to pitch up and hold my new attitude, It doesn't matter how hard I pull back, or what the trim configuration it, I just pull back and count a few seconds, and the plane does everything for me automatically...just like it always does.

What's it doing now?

[Deadstick]

I'll admit to being an ignorant slut, but in my experience as a stick and rudder guy, I see the Airbus sidestick as an extension of a video game. I guess it works just fine most of the time, but to pull back on a yoke or joystick gives you a lot of feedback as to what the aircraft is doing. It's a little like the new whiz bang deal with cars that want to hit the brakes if you haven't, or change the traction on the wheels if it doesn't like what you're doing. I want to drive my own car and fly my own aircraft. I have nothing against an aircraft autopilot, but I don't want it deciding for me what level of assistance I get next.

[3WE]

That's interesting Gabriel because I always got a full break and pitch over in the same aircraft.Maybe you were more aft CG?

I don't like falling a whole lot, so when things break (and they do seem somewhat sharp in a Cessna 1X2) I relax the yoke...nose over, and return to flying...

...however, it was my understanding that it you really kept the yoke hauled full back, you would stay in a reasonable tame mushy, moderately fast descent...(somewhat like Air France)

[Deadstick]

I don't like falling a whole lot, so when things break (and they do seem somewhat sharp in a Cessna 1X2) I relax the yoke...nose over, and return to flying...

...however, it was my understanding that it you really kept the yoke hauled full back, you would stay in a reasonable tame mushy, moderately fast descent...(somewhat like Air France)

I've always had sharp breaks, but I'm a skinny little bald guy with a lot of yoke pull.

[3WE]

I want to drive my own car and fly my own aircraft. I have nothing against an aircraft autopilot, but I don't want it deciding for me what level of assistance I get next.

To hell with you...

I don't want the plane to tell the multi-thousand-hour pilots up front what to do when I'm back in 28F with only an air-yoke.

[snydersnapshots]

And they did it yet again!


9s The FO calls GA, the captain disconnect the AT and advances the throttles.[/COLOR]
[COLOR="SeaGreen"]10s The AP is never disconnected
11s so it keeps tracking the glide slope.

I haven't flown the 777, but I HAVE flown the 767-400, which is very similar. In fact, all the Boeings I've flown (except the 727 for obvious reasons: no auto throttle) are pretty much the same.

The quote above is where it all went to hell on 'em. Instead of hitting the Go-Around switches, they hit the Auto throttle disconnect. Hitting the GA switches would have put the autopilot in the Go-Around mode and the auto throttle would have probably commanded about a 2,000 foot per minute climb (I know that's what it does on the 737/757/767. I'm assuming it will do so on the 777 as well).

By hitting the A/T disconnect, they simply shut off the auto throttles, which left the autopilots coupled to the localizer and glideslope. Even with the a/t disconnected, they could still have saved the day by hitting the Go Around switches and manually advancing the power.

I can see how this can happen, for a couple of reasons.

(1) Early in my career (long before I got an airplane with autothrottles and autopilots--back in the days of turboprops, 8 legs a day with nothing but raw data), I made a couple of unexpected missed approaches that were complete goat ropes. It all worked out, but it wasn't pretty. That's when I developed the technique of talking through the callouts and mentally going through the missed approach procedure as we go over the outer marker. Go arounds or missed approaches were relatively common in the Jetstream and Metroliner (I remember one flight where I made 5 approaches, with three misses because I didn't break out (the last approach was a hail-mary while I was giving the company time to coordinate our divert) and one ATC directed break out because the speed they assigned caused us to nearly lose separation on the traffic ahead). In 18 years of flying Boeings, however, I can count the missed approaches on the fingers of both hands...with a few left over.

(2) Because Go Arounds/Missed Approaches are relatively uncommon, the muscle memory when you get to the bottom of the approach has the pilot disconnecting the auto-throttle. In the heat of battle, habit took over. My technique is always to have my fingers or thumbs (depending on the airplane) gently against the go around button (737/767-400) or switch (757/767-200) and my thumb WELL away from the a/t disconnect switch. I'm sure as heck not the ace-of-the-base and I know if there's a way to screw something up, I can find it very easily. I try to figure out tricks or techniques to prevent me from, as Tom Wolfe so eloquently put it in "The Right Stuff," "screwing the pooch."

[Gabriel]

The quote above is where it all went to hell on 'em. Instead of hitting the Go-Around switches, they hit the Auto throttle disconnect. Hitting the GA switches would have put the autopilot in the Go-Around mode and the auto throttle would have probably commanded about a 2,000 foot per minute climb.

By hitting the A/T disconnect, they simply shut off the auto throttles, which left the autopilots coupled to the localizer and glideslope.

Aw, come on! I'm disapointed.

The part that you describe, yes, it was the trigger of the incident, but it was a stupid, innocent and completely forgivable error that could have been very quickly and easily detected and correctd had any pilot been flying the plane.

How is that, immediatelly after comencing a GA, the flying pilot is not motioring the PFD and the FMA? What more important thing he had to do? Or wait. What other thing he had to do, more important or less important?

THEY LOST 300+ FT IN A GO AROUND IN A CATIII APPROACH IN IMC!!!! THAT'S LIKE 20 SECONDS OF KEEPING THE DESCENT AFTER THE GA IS STARTED!!!!
Doesn't one of the pilots need to be 100% of the time monitoring the flight parameters and the FMA during a CATIII autoland?

Hitting the GA buttons (even the correct ones) and go for a cofee while the AP/AT handle it with no human supervision is not an acceptable procedure. Yes, you will get away with it the 99.9% of the times that everything goes well. The first time that something does not go well will also be the last time that this (or any other thing) happens to you. Unless you have a relief pilot in the observer's seat. That guy saved a lot of lives that day.

They miss the master warning chime and light.
They miss the NO LAND3 message (in the PFD).
They miss that AP is still in approach mode (in the PFD).
They miss that the plane is still descending altitude and vertical speed in the PFD)
They miss tat pitch has not gone up and is in fact going down (attitude indicator in the PFD, I know, not a very important instrument in a CATIII approach and a GA in IMC, NOT!).
They fight for the yoke.

Who was flying the plane? Doesn't the PFD have the words "primary" and "flight" in its name for a reason? Maybe the pilot FLYING should glance towards the PRIMARY FLIGHT display every once and then? Especially in a CAT-III approach and a GA in IMC?

Do you really think that this was a case of simply hitting the wrong GA buttons?

[brianw999]

I'll tell you what I think !

I think I won't be flying Air France anytime soon !!

The old days nickname of "Air Trance" doesn't seem to be so much of a joke anymore !

[Black Ram]

...nothing to do with the planes control philosophy huh?

I pull back on the stick and it's supposed to pitch up and hold my new attitude, It doesn't matter how hard I pull back, or what the trim configuration it, I just pull back and count a few seconds, and the plane does everything for me automatically...just like it always does.

Now that you mention it, I think it's unlikely this would have happened in an airbus (trying to use Sully's wording ). I think just applying TOGA would have activated GA mode.

So my point is, no matter what the interface, control input or protections, it all defeats the purpose when the crew is incompetent. I think these comments from the avehrald report sum it up:

""Soon Airbus will come with an aircraft that will go around when you shout "go around!".

Just make sure you speak clearly and don't instruct it to "Go to Ground!"

Perhaps a better use of the research dollars is an aircraft that learns to hover motionless in the presence of contradictory control inputs?"

[Deadstick]

To hell with you...

I don't want the plane to tell the multi-thousand-hour pilots up front what to do when I'm back in 28F with only an air-yoke.

I'm more likely to be found in 4D sipping champagne.

[Black Ram]

Incident: Air France A320 at Marseille on Mar 11th 2013, approach to stall during visual approach
By Simon Hradecky, created Tuesday, Mar 26th 2013 17:24Z, last updated Tuesday, Mar 26th 2013 17:24Z

An Air France Airbus A320-200, registration F-HBNE performing flight AF-7664 from Paris Charles de Gaulle to Marseille (France), was on a visual approach to Marseille's runway 31L. While turning final the alpha protection activated, the aircraft reached a minimum height of 700 feet before the crew initiated a go-around. The aircraft subsequently positioned for another approach and landed safely.

The French BEA reported in their weekly bulletin that the occurrence was rated a serious incident and an investigation by the BEA has been opened.

http://avherald.com/h?article=45fc9192&opt=0

I wish BEA was publicly traded.

[Gabriel]

Now that you mention it, I think it's unlikely this would have happened in an airbus (trying to use Sully's wording ). I think just applying TOGA would have activated GA mode.

If you read another comment in AvHerald from a very nine, smart, and handsome guy named Gabriel or something like that, you'l learn that this kind of incident CAN and actually HAS happened in an Airbus. The only difference was that the mistake, instead of hitting the wrong TOGA button, was advancing the throttles a hair short of full forward.

[snydersnapshots]

So my point is, no matter what the interface, control input or protections, it all defeats the purpose when the crew is incompetent...

Okay guys... After reading Gabriel's and Black Ram's responses earlier this afternoon, I've been stewing about my reply for several hours, and it's all I can do to keep myself from hitting the caps lock in places while writing it.

Everybody is hammering on the crew for screwing up. Did they screw up? Yes--absolutely! Does the traveling public have a right to EXPECT a certain level of proficiency from a crew, to include a smoothly executed missed approach? Again, ABSOLUTELY! I'm not arguing that fact at all. Remember--my family flies in the back of an airliner occasionally too.

Now, instead of calling them incompetent and implying that they're a bunch of dumb$#%$ who don't know how to fly an airplane, why don't we analyze the situation and try to find out WHY they did what they did and maybe learn from it?! Hopefully, that's what they've done in the months after the incident as well.

Everybody here seems to be an expert on the physical operation of an airplane, and I'll be the first to admit that you guys are sharp and pretty much know your stuff with regard to systems and procedures. What I do NOT see on this forum is much discussion about the human factors aspects of a situation like this. Instead of armchair quarterbacking their operational deficiencies, based on our perception of their incompetence, let's assume that they have been flying for many years in many types of airplanes and have built up a certain skill level that has allowed them to advance to the airplane they're on. Then let's try to backtrack and figure out why they did what they did. I realize we're talking about the same airline that had the A340 accident. The operational and knowledge deficiencies in that case are well known and agreed upon, but we're comparing apples and oranges here (as we say here in the former rebellious colonies) as this incident relates to the A340 accident.

Gabriel--you hammered on them for a number of things, among them:

Missing a Master Caution light and beeper: According to the 777 manual, the Master Caution light and aural warning are inhibited from the time LAND 3 or LAND 2 is annunciated until the airplane is either below 75 knots or above 800 feet (on a missed approach) with some exceptions: Autopilot Disconnect (we know that didn't happen), Autothrottle Disconnect (we know the captain did that after the word "alarm" was spoken), No Autoland (Since the FMA went from "LAND 3" to "LAND 2" we know that didn't happen), and Speedbrake Extended (I'm sure that wasn't a factor since the captain probably didn't touch the speedbrake handle that low in the approach. Also, if the speedbrakes had been extended, that fact would have been noted in the report). Thus, I would have to assert that the Master Caution in fact did NOT illuminate, nor did the aural warning go off.

Missing the NO LAND 3/LAND 2 annunciation, which appeared as the airplane was descending through approximately 320 feet. You insinuated they were incompetent for not being glued to the PFD/FMA. Are you instrument rated? When you're approaching minimums, where are you looking? You go to a composite crosscheck where you're shifting your scan between the instruments and outside to try and pick up the approach lights. One thing that is not stated in this report is the weather--it's entirely possible they had a ceiling and/or visibility that was high enough they expected to see the lights. In fact, the report stated: "the aircraft was on final approach descending through about 350 feet AGL with the crew trying to gain visual contact with the approach lights, when the relief pilot called out "Alarm" and the first officer "go-around". " It's also possible that one of them DID notice the change but, realizing from a systems standpoint that the aircraft would still conduct a safe autoland, elected not to say anything and to just continue the approach. I will agree that the IRO made the proper call in directing the go-around (our procedures state a mandatory missed for a change of ASA (Land3 to Land 2) above 200 feet, but only a "NO AUTOLAND" below 200 feet AGL.)

"They miss that AP is still in approach mode (in the PFD).
They miss that the plane is still descending altitude and vertical speed in the PFD)
They miss tat pitch has not gone up and is in fact going down (attitude indicator in the PFD, I know, not a very important instrument in a CATIII approach and a GA in IMC, NOT!).
They fight for the yoke."

This is where the human factors stuff I'm talking about comes into play. They screwed the pooch on this one and I wholeheartedly agree. Now let's figure out why.

I'm doing some speculation here, but this is the scenario as I see it possibly playing out:

The crew has been flying for somewhere between 8 and 9 hours and it's pushing 10 am as they make their approach (not sure of the exact time, but I think that's close from what Flight Aware tells me about the flight as it's flown today). The information we're missing is how long their trip is and how long their layover was in CCS. If this was a trip similar to what I've flown in the past, they flew to CCS, had about a day layover, and flew back to Paris. They left CCS about 8 pm, which means they had to be at the airport around 6:30 pm. If they stay near the airport, they'd have about a 6 pm van. If they're in Caracas proper, about 45 minutes away, you're talking a van at 5:30 or 5:45. Back up an hour from that for a wake up and you're talking about getting up from a nap at 4:30-5 pm. So, they probably slept the night before the flight, got up and had breakfast or brunch and then tried to take a nap from 11 or noon until 4:30 or 5. I can tell you from personal experience, 2-3 hours of decent sleep is about the best they could expect in that case. Once in flight, they can take a break. If the flight time is 8 hours, the first break probably started when they leveled off, about 20 minutes after takeoff, and the last break would end about 40 minutes before landing. Divide 7 hours by three and they each have a break of 2:20. So there's maybe two more hours of sleep. So now they're on the approach to Paris. In the previous 24 hours +/-, they've had two naps which total around 4 to 5 hours of sleep if they're lucky. Any human factors book will tell you that reaction time slows with sleep deprivation.

So now they're almost at the bottom of the autoland when the go-around is called. The captain mistakenly clicks off the autothrottle instead of hitting the TOGA switches. He manually advanced the power while simultaneously calling for flaps 20. Next thought--oh crap, this thing isn't climbing, so he pulled on the yoke, but not hard enough to override the autopilot and put the thing into control wheel steering. The IRO called "Pitch" and the captain pulled back, popping overriding the autopilot and resulting in a nearly 2 g climb initially. I think the captain probably lowered the pitch in response to the abrupt pullup before pulling again and finally getting the airplane climbing.

So, my analysis? Reactions slowed by fatigue coupled with a relatively unusual maneuver (as I mentioned in my previous post). A habit pattern lead to the a/t being disconnected instead of the TOGA buttons pressed, which lead to the airplane not doing what they expected it to.

What am I going to take away from this? A reinforcement of my technique of briefing the go around callouts and procedure while over or inside the marker. Could I do something like this in the airplane? I certainly hope not, but I'll never say never. One thing I will NOT do is label another pilot as "incompetent" until I have flown with him or her and made that determination from personal experience. Believe me, I HAVE flown with some incompetent pilots, though not at this level of the industry. Do I believe the pilots involved were incompetent? No. I truly believe that if they had been in the first hour of the flight when they were fresh instead of the 8th or 9th hour, the outcome would have been completely different.

If you are into flight sims and want to experiment a little bit, try doing a schedule as I described in real-time. Do a flight during the day, when your at the height of your circadian rhythm and have someone count your mistakes. Then do the same flight after a schedule as I've described here and have someone count your mistakes again. I think it will be a real eye-opener.

[3WE]

I'm more likely to be found in 4D sipping champagne.

Hmm...Moving a glass of champaign back and forth is about as effective as moving my air-yoke back and forth, but maybe eith some EtOH in your system, it will hurt less when we both hit.

Let's see, you could set that glass on the arm rest and then sort of rock it around with your wrist...and just before we hit, you can say, "but I've been pushing over the whole time, what's it doing now?"