ANA Japan roll incident.

[Hotel Tango]

Courtesy of the Aviation Herald:

An All Nippon Airways Boeing 737-700, registration JA16AN performing flight NH-140 from Okinawa to Tokyo Haneda (Japan) with 117 people on board, was enroute at FL410 about 23nm south of Hamamatsu (Japan) when the aircraft violently rolled left causing injuries to two flight attendants and descended by about 6000 feet levelling off at FL350 again. The flight continued to Tokyo's Haneda Airport for a safe landing. The two injured flight attendants were delivered to a hospital with flesh wounds.

The airline reported, that the captain (64, 16,000 hours total experience) had taken a toilet break and was about to return to the cockpit requiring the first officer (38, 2400 hours total experience) to open the cockpit door. The switch to open the cockpit door is located about 10 centimeters/ 4 inches away from the rudder trim switch. By mistake the first officer operated the rudder trim switch instead of the door opening switch causing the aircraft to violently roll left and descend by about 6000 feet before the first officer was able to return the aircraft to stable level flight. The captain subsequently entered the cockpit. Two cabin crew received minor injuries (flesh wounds), no other injuries occurred. There was no turbulence in the area at the time of the occurrence.

Japan's Transportation Safety Board rated the occurrence a serious incident, opened an investigation and dispatched three investigators on site.

[Henri737]

This is not the first time this occurance took place. I know of a similar incident. It's all about ergonomics on the 737: look at the B-hydraulic vs engine anti-ice switches..... Time for a panel redesign.

[tezzer]

Just been reported on the BBC Radio news bulletin.

Wonder if there was a requirement for another "toilet break" soon after !

[westhawk]

It's not just 737s either. In fact you probably couldn't name an airplane where something similar isn't possible or hasn't been done by someone somewhere at some time. Those old autopilot and yaw damper "paddle" switches located adjacent to each other on '70s and '80s era Collins A/P controllers? Many a PF who called for the Y/D on at 400' surprised by the suddenly engaged A/P that doesn't want to go the same direction as the pilot does. Fire bottle switches? I got plenty of work replacing discharged fire bottles when I was a mechanic. Ever throw the wrong switch during an abnormal drill in the sim? Yep, been there and done that too! (a better place to learn such a lesson than in the plane)

Some pilots (and mechanics) learn early on that you really must stop and think before flipping a switch, no matter how mundane the task might seem. Some learn later than others.

Oops indeed!

But the point that better thought out ergonomics can reduce errors in this area is well taken.

[hetfield]

Thx God it wasn't an A300.

[glad rag]

Lucky the fin stayed on >:getmycoat:<:rolleyes:

[HeadingSouth]

in "my" cockpit there are numerous switches in close vicinity of about 4in / 10cm of each other. how about other issues such as concentration, fatigue, other ideas ?

[Dani]

If it would have happened in an Airbus, most of the posting would be about "protections" and FBW induced incidents.
In a conventional aircraft its all about "it happens everywhere". I'm sure this will not happen in an Airbus.

[westhawk]

In a conventional aircraft its all about "it happens everywhere". I'm sure this will not happen in an Airbus.

"It" can happen anytime someone changes a switch position without first making sure that the correct switch is being operated. I doubt that Airbus or any other type of aircraft are exempt. Why would anyone want to bring the A versus B argument to a thread that applies to ALL machines?

[Northbeach]

It's all about ergonomics on the 737: look at the B-hydraulic vs engine anti-ice switches..... Time for a panel redesign.


I have been on the NG for several years now, I like the jet and it does many things well. However the ergonomics and flight deck design disappoint me terribly. For example, on the flight control panel the heading, altitude and airspeed knobs all feel virtually the same. Now I do not have an engineering or human factors undergraduate or advanced degree(s), but why in the world would the experienced aeronautical engineers with the resources and experience that Boeing has overlook this obvious design weakness?

On the MD-80 (McDonald Douglas product) when one reached for the heading knob there was a distinctive feel associated with the knob because each heavily used input knob had enough of a distinguishing structural design to communicate identification information through touch alone. For example, if I were looking out the side window scanning for traffic and simultaneously reached for the heading knob to comply with ATC’s turning instruction, and if in error, I grabbed the altitude setting knob I would get an immediate “wrong action” feedback through my fingers. Granted this touch based knowledge had to be gained through time on the jet, but it developed rather quickly (in my opinion).

To my disappointment this is not so on Boeing’s NG product; the airspeed, heading and altitude knobs on the flight guidance panel all look similar and, more importantly, are virtually indistinguishable from one another using touch alone. Bad design, really bad design especially coming from a company that has been producing aircraft as long as Boeing has.

As far as getting a design makeover, needed - I agree completely, but doubt it will ever happen. The aircraft was certified as it operates today, to totally redesign the flight deck instrumentation and positioning would require the jet to go through the regulatory recertification process. Boeing isn’t going to voluntarily pay for that. They have enough going on trying to get their new baby, 787, out of the production storage facility and operational.

More on topic I suspect positioning of the flight deck door opening switch is left up to the individual company as “ours” are not anywhere near the rudder or aileron trim switches. People are going to make mistakes, this time the jet fell out of the sky, and fortunately it was recovered. I suspect that pilot will not make the same mistake ever again-one would hope so.


Why would anyone want to bring the A versus B argument to a thread that applies to ALL machines?


Some people must exercise their demons; whether they suffer the lasting effects of perceived or very real insults, inferiority complexes, need to establish their own superiority by putting others in their place or driven by a strong sense of national or regional pride. In those cases every accident thread, and many other threads become a place for them to make the same A versus B arguments. Emotionally they seem to have a need to be on the, and identify with the "winning side" and vindicate themselves as if they personally were responsible for the favorable outcome. I find them tedious reading and tend to skip over the comment(s). When the only tool you have is a hammer, then every problem becomes a nail. So one reads the same dribble over and over again.

I believe Airbus has been winning the larger number of future sales and delivering more jets for several years now. And the A380 seems to be popular. The “marketplace” has spoken. Personally I think the competition is good for the industry as it drives innovation and productivity. Look over our collective shoulders and we will see that the Chinese are working hard on developing their own aerospace design and manufacturing industry. A, B and now C .........

No insult intended to the Canadians, Brazilians, Russians, the Spanish or others by not previously mentioning them. They make fine airplanes that compete successfully in the global marketplace.

[westhawk]

I've flown 737 100-800 and they all had a ROTARY KNOB for the rudder trim-- I think the reporter has confused things with the aileron trim (KNOB on the 100/200 and switches from the 300 onwards.) Same setup for the Lear 60. In fact it's two knobs one atop the other on coaxial shafts which must be activated simultaneously to cause trim tab movement. Apparently either Learjet or the FAA felt the extra protection against inadvertent rudder trim operation was needed. They are spring loaded rotary switches that return to the center (neutral) position when released.

[Denti]

@Northbeach, there are at least two different MCP versions on the 737NG, one does have different knobs with different haptic feedback (they feel a lot different), the other brand's knobs are not that much different (there is still some difference). The first one is the old Honeywell panel and the latter one the newer Rockwell Collins one.

@OK456, it might depend on the door system installed, on our 737s the door lock switch is a rotary switch which does not have a push action to normally unlock the door. Its position on the center pedestal changes between aircrafts so muscle memory could well lead you to the wrong position. However the switch is very differently designed compared to the rudder trim.

[iwrbf]

this reminds me of a thread some time ago:

http://www.pprune.org/biz-jets-ag-flying-ga-etc/315587-gulfstream-550-double-engine-shutdown.html

I guess the "double check before you change something" is not as old fashioned as the machine-to-human-interface groupies try to postulate on and on and on and on... similar knobs are as bad as knobs that were thrown around a panel in a randomized modern art way but the machine-to-human interface won't EVER be perfect.

Poor FAs (and poor PAX frightened to hell), but happily they all survived... this time...

[Sky Wave]

I've flown 737 100-800 and they all had a ROTARY KNOB for the rudder trim-- I think the reporter has confused things with the aileron trim (KNOB on the 100/200 and switches from the 300 onwards.)

The 737-700 that I flew had a rotary knob for the cockpit door control and it was very close to the rotary control for rudder trim.

SW

[Dani]

"It" can happen anytime someone changes a switch position without first making sure that the correct switch is being operated. I doubt that Airbus or any other type of aircraft are exempt. Why would anyone want to bring the A versus B argument to a thread that applies to ALL machines?

Because it cannot happen in an Airbus. Airbus don't have trim switches (well, there is one). The ergonomics of a modern airliner is a very important safety feature. Airbusses have so few switches in the cockpit that you would really have to make an effort to mistaken one for the other.

[westhawk]

Because it cannot happen in an Airbus.

Well I guess that's the answer then.

[Burger Thing]

Airbusses have so few switches in the cockpit that you would really have to make an effort to mistaken one for the other.


Apart from the light switches. They are a complete disaster. :}

[cwatters]

6000ft sound a lot.

[CDRW]

Oh Dani Dani - you never heard about the SIA Airbus en-route over Australia, when the captain decided to turn off some fuel pumps, but turned off some hydraulic pumps. It did some pretty interesting things until normal hydraulic power was resumed. "It would not happen on an Airbus" - yea sure.

[Callsign Kilo]

I'm sorry, but what happened to actually using your eyesight to identify the difference between the two knobs? Yes they are close to each other but one is notably larger. It is also in the centre rear of the console with a big black line running through it. Boeing intuitively marked this with large white italics. The word "R U D D E R", or something like that, are labelled on it. This guy also had 2400hrs of experience. Albeit it doesn't tell us how much 737 time he or she has, I kinda presume that they have some level of gumption? Plus he or she must have really deflected that rudder trim knob. It takes some time and I imagine quite a lot of deflection would be required in order to make the aircraft want to perform aerobatics. There is also a notable difference between the way the two switches behave. The door unlock switch has a notable and immediate stop on it. I'm sorry, but I just don't get this one. Neither do I put it down to poor ergonomics. Probably more to do with not bloody looking!

[Green Guard]

so only 10 cm apart !!!

then we all should be very carefull next time when eating,
and using our fingers, forks or chopsticks...especially in dark cockpit

NOT to mix up the mouth for an eye

[cactusbusdrvr]

OK, poor switch position is fine, but why on earth was there not a F/A on the flightdeck during the restroom break? With the threat of the badguys entering the flightdeck just flipping the door unlock switch on does not sound like the brightest thing in the world to do. one of the inflight crew should have been on the flightdeck ready to open the door after verification.

[Dani]

by CDRW:
...the captain decided to turn off some fuel pumps, but turned off some hydraulic pumps.

that's what I call "an effort". ;)

Last time I looked at, the trim switches are not on the overhead panels. So no, you cannot confuse secondary flight control systems with a door lock mechanism in an Airbus (unless you make an effort...).

Of course you could confuse your captains seat with the one of the toilet - the possibilities are endless. What did you want to tell me? :suspect:

[Airbubba]

Years ago a Deltoid captain climbing out of LAX thought he was turning off the EEC switches on a 767 and somehow turned off the nearby fuel control switches instead. They got the motors restarted 600 feet off the water, got salt spray on the plane and continued on to CVG, RAT extended, with the blessing of ATL flight ops:

ASN Aircraft accident Boeing 767-232 N103DA Los Angeles International Airport, CA (LAX) (http://aviation-safety.net/database/record.php?id=19870630-0)

The FAA did an emergency license revocation and actually went to his house to pick up his ticket in person. I'm told he took early retirement.

The EEC switches were later relocated to the overhead well away from the fuel cutoffs.

[5 APUs captain]

The 737 is redesigned from 727 (I am experienced on both).
That's why 737 ergonomics is just poor.

[Don Coyote]

With reference to the first post, it could not happen on the airbus as the rudder trim switch is inoperative when the autopilot is engaged.

[SKYCLEAR13]

All of us can do mistakes.

But some of our new copilots are more chicken than eagle....:ouch:

[cactusbusdrvr]

Dani, read the post before you comment on it. The post was about confusing the fuel pumps for the hydraulic pumps, which is very easy to do on an Airbus as they are above each other on the overhead. HYD, FUEL, ELEC, AIR, in descending order.

[Sky Wave]

, but why on earth was there not a F/A on the flightdeck during the restroom break? With the threat of the badguys entering the flightdeck just flipping the door unlock switch on does not sound like the brightest thing in the world to do. one of the inflight crew should have been on the flightdeck ready to open the door after verification.

So airlines the world over should re-write their SOP's to match your door opening procedures?

We have never required an F/A to come in during a comfort break, nor do most airlines who's procedures I know.

Should we remove our video camera's from our cockpit door?

There's more than 1 way to crack an egg.

[bubbers44]

US carriers require a FA in the cockpit to verify our security procedures are followed. It seems your procedures use a security camera which would accomplish the same thing. Keep doing what you are doing. We will catch up.

[bubbers44]

Ten years ago today is when we had to implement these procedures so we probably do a bit of overkill to make sure it can't happen again. I am sure today the bastards will test us again on the anniversary.

[galdian]

Said with a heavy sigh: "IN Japan.........!":(

It is JCAB reg for all airlines that if one crew leaves then a CA (in some companies at least it MUST be the CManager only) comes in to open the door for the return of the crew; I believe this was done although not clearly stated, IF for some reason it was NOT done then one can expect blood!

Also from JCAB a continuous outside watch is paramount to safety - so if the F/O was maybe low experienced and impressionable he could not spare a glance inside the cockpit, far better to fumble around and get it....wrong. :ugh:

If procedures were followed and there was a CA in the cockpit the question of why the F/O was stuffing around with the door lock rather than allowing the CA to ID the returning crew after a pre-arranged signal and open the door will be a point of interest.

As a final - don't know procedures around the world but do know Australia changed 10-15 years ago so that oxygen only had to be worn above 41,000ft
(or something like that) when single crewmember in the cockpit.
I believe Boeing have stated that unnecessary use of the quick donning masks reduces their life - one wonders if that could cause them to fail when really, really required?? :ooh:

In Japan use above 25,000ft remains - no doubt the rationale "...the B727 oxy mask was very difficult to use therefore..." or similar would be the justification, the fact it is a totally different mask to the 727 (the term "quick donning" a bit of a giveaway to the vast majority ;) ) will probably need pondering and consideration for at least the next 10 years (being generous!)

[Loose rivets]

9/11 caused the demise of my 'retirement job' - when the bank pulled the plug on us. Generally I read of huge advances in technology and safety, but when I read that one crew-member is left on the flightdeck as the door is locked behind them, I'm left bewildered.

I realize there can't be a public answer to part of this issue, but just leaving one crew-member behind something that would even delay assistance has to be just plain wrong. There are several forms of incapacitation that leave the patient not just incapacitated, but literally flailing about. One outfit I worked for trained the cabin crew in containing such an occurrence.

[Lord Spandex Masher]

LR, some of those huge advances in technology allow, erm, "alternative means" of getting back into the flight deck without being let back in, should it be necessary. I'm not telling you how though!

It's never been a requirement at the airlines I've worked for to have an extra crew member in the flight deck if one of us is having a quick widdle.

A number two or longer absence, for whatever reason, does require the extra crew member though.

[Fratemate]

It is JCAB reg for all airlines that if one crew leaves then a CA (in some companies at least it MUST be the CManager only) comes in to open the door for the return of the crew

Not true, mate. With the advent of the CCTV that requirement was removed. Not going to go into minute detail for the obvious security reasons, although I do appreciate how over-the-top that sounds :).

[Al Murdoch]

There's at least one major European airline that requires cabin crew to be on the flight deck when a pilot goes to the toilet. Coincidentally that same airline had an almost identical incident as the one mentioned here. Lost control at FL390 I think, after the autopilot gave up on the full deflection rudder trim whilst trying to open the door.
If it can happen - it will happen.

[galdian]

Fratemate

Thanks for that, wasn't aware the local regs allowed that option - was just speaking from general knowledge/undertnanding.
Questions: is it just an option under the regs or do some Japanese operators actually use CCTV as standard ops ?? Pax or freight or both??

Ta!

[Lookleft]

Our airline requires an F/A in the cockpit at night time if one of the pilots goes out of the cockpit but they put that requirement in the CC manual so no-one bothers. The other problem is that they have reduced the CC numbers and put cadets inthe RHS so I will reduce my caffiene intake and stay put.

[Fratemate]

Galdian,

Standard ops i.e. we use the CCTV all the time on the pax flights. It is, of course, on pax flights only, as there's no nasty people in the aircraft on the freighter (apart from the pilots) so there's no need for TVs (and the toilet is on the flight deck :))

[ManaAdaSystem]

The NG is poorly designed when it comes to switches and where they are placed. You ask for the wipers, and you get engine ignition. You ask for eng a-ice, and you get hyd pumps. Try to operate the MCP in darkness/smoke, and you will have a hard time (as another poster already pointed out) to differenciate between heading, alt, etc.

In this tread we have the super pilots who never make mistakes, and think selecting a wrong switch is impossible and due to poor operating procedures. They are in for a big surprise sooner or later.

The rest of us have already been there and done that, and we are better prepared for these events. Including Pilots trying to retract flaps when gear up is called.:eek:

You are missing the important bit. Had the 64 year old captain had some
bladder control, this would never have happened.:p

[UK019]

We too have the CCTV but it doesn't cover this: Pilot goes for a pee. Other pilot becomes incapacitated. Door entry system fails and we have an operational pilot in the cabin with the pax and unable to get back to the flight deck. End of aeroplane and everyone in it. Sure, it's a long-shot - but not impossible. That's why my Company always has two people on the flight deck in flight. It's usually a cabin crew member but if there's a third pilot, that fulfils the requirement. It's a simple, easy, safe procedure and costs nothing.

[Callsign Kilo]

The NG is poorly designed when it comes to switches and where they are placed. You ask for the wipers, and you get engine ignition. You ask for eng a-ice, and you get hyd pumps. Try to operate the MCP in darkness/smoke, and you will have a hard time (as another poster already pointed out) to differenciate between heading, alt, etc.

In this tread we have the super pilots who never make mistakes, and think selecting a wrong switch is impossible and due to poor operating procedures. They are in for a big surprise sooner or later.

There is no such thing as super pilots who never make mistakes. Yes the NG is based on an antiquated design, however I'm a firm believer in LOOK before you TOUCH. People seem to believe in doing things in a flash of light - where's the rush? This is my observation anyway; and I see it quite frequently. The PMs hand moves like a world wind. A scarce glance is made in it's direction -
bam; system B is depressurised or in this case a passenger airliner approaches a spiral dive. Either people start to look or we adjust the retirement age to directly correlate with the onset of bladder issues! :ok:

[White Knight]

Dani, read the post before you comment on it. The post was about confusing the fuel pumps for the hydraulic pumps, which is very easy to do on an Airbus as they are above each other on the overhead. HYD, FUEL, ELEC, AIR, in descending order.

Except the HYD switches are guarded and the FUEL not........ Bit of a difference methinks.

At least that's how it is on 330/340:ok:

[FlexibleResponse]

White Knight

Except the HYD switches are guarded and the FUEL not........ Bit of a difference methinks.

At least that's how it is on 330/340

White Knight,

The HYD switches only became guarded after SQ managed to switched off all four HYD switches in the cruise. Now, that was exciting...

Perhaps the "bit of a difference" is not as large as one may think when one reflects in the cool light of day..?

[rubik101]

I'm not sure if UK019 is in the same company but the system is the same for us and it works. Peeing pilot is replaced by an F/A. Flirting-chit chat goes on for a few minutes. Peeing pilot returns, knocks or buzzes, non pee pilot checks the camera or the F/A peeps through the peep hole and opens the door. Body contact sometimes occurs and everyone is happy. No risk, no cost, no altitude lost, no switches operated.

[Yaw String]

Not sure the SIA incident involved the Cpt...Anyway, am pretty certain the full crew complement were not on the flightdeck while the "Fuel balancing" was attempted..
Does not matter what age or experience you have...the only thing you can be certain of finding with your eyes closed is...well...you know what:mad::E...Sorry ladies!

[UK019]

Rubik101, having read your public profile (I want your lifestyle please) no I don't think we are in the same outfit, but what you say makes total sense and as you say, it works - with a bit of ''frottage' to boot! ;) On a serious point though there are no negatives in having a rule that says that the flight deck must never be occupied by just one person.

[Dani]

We too have the CCTV but it doesn't cover this: Pilot goes for a pee. Other pilot becomes incapacitated. Door entry system fails and we have an operational pilot in the cabin with the pax and unable to get back to the flight deck.

That's then about a tripple or quadruple failure. Door system has a fail safe mode, and if you know the code you can access from outside. If that's still not working, there is a general code and an emergency code. If you loose your captain (due to bladder leak) and you loose your FO and you loose all your electricity or the whole door mechanism - well that's less probable than an all engine failure I would say.

People always invent so complex failure when dealing with new systems. Come on - you can do better!

Dani

[40&80]

Hope the health and safety people are not reading this.
They would stop 50 ton trucks operating with one crew around the country and also a Mum doing the school run without a responsible adult also in the car..probably not be allowed to operate your ride on lawn mower single driver.

[UK019]

Dani no it isn't. It's a double failure (the door and the incapacitation)

A double system failure is something I have experienced twice in my career (33 years airlines, 10 before that flying elsewhere in aviation) one of which caused a high speed rejected T/O when two utterly unrelated systems failed at precisely the same moment. No-one (including me) could believe it, but the evidence was there in front of our eyes.

As I said in my post, it's a long-shot, but the prevention is so simple and has no cost that I'll stick with it if I may. I don't have a choice anyway - it's SOP in my outfit!

[Dani]

The door doesn't fail per se. You have an electronical, an electrical and a mechanical component. If all three fail it's a tripple fail alone (without the dead FO). Experience show (Helios B737) that one is able to get inside if you really have to and if you are not a terrorist and if you get help from the cabin side.

[poorjohn]

Interesting that a subsequent thread got locked for some reason, while debating the meaning of an initial analysis that purports to show that the aircraft "inclined to the left a maximum of 131.7 degrees". "2.6G" was mentioned, too.

[danger85]

Is there a new feature being rolled out on Boeings enabling reverse flight? According to the Sydney Morning Herald in Australia, "Images from the flight recorder on the Boeing 737-700 showed the plane veered to the right and then sharply to the left, before dropping backwards.":ugh:
ANA plane plunges after pilot presses wrong button (http://www.smh.com.au/travel/travel-incidents/wrong-button-caused-plane-to-almost-turn-upside-down-during-plunge-20110929-1ky4c.html)

[lederhosen]

This is very similar to the recent incident in India where the co-pilot pushed the column unintentionally and was unable to understand what was going on when the aircraft went into a dive.

It also raises the question what the co-pilot was looking at when he twisted the rudder trim. It looks completely different and is in a different place from the door unlock. If he was looking at the instruments how come he did not see the control wheel moving way out to the side?

Above all it raises the point that most pilots have little experience of manual flying these days and when an unusual attitude develops they are initially stumped. Our sim training exercises lack the surprise factor which seems to have been the problem here.

[Denti]

The door unlock can be pretty much anywhere and we have at least 10 different locations in our fleet, at least one of them where it is directly adjacent to the rudder trim. Added to that there are some very bad CCTV setups around as well, for example having the screen on the aft overhead panel behind the left seat. An FO would have to lean out of his seat over the center pedestal to see anything on that, identify the person who wants to enter the flightdeck, keep observing the area why he supposedly opens the door. In that position he would neither see the flight controls and displays nor the door unlock button or rudder trim.

Still, both knops have a different design, but the mode of operation is the same.

Newer CCTV setups display the picture on the lower DU, much more convenient.

Boeing approved retrofits or even boeing designs can be ergonomic nightmares, and the whole door thing is just that in some configurations.

[punkalouver]

It's never been a requirement at the airlines I've worked for to have an extra crew member in the flight deck if one of us is having a quick widdle.

A number two or longer absence, for whatever reason, does require the extra crew member though.

It is on the aircraft I have flown in recent years where there is no keypad to be able to get back in the flight deck. F/A required in flightdeck when one crewmwmber leaves the flight deck.

[Indarra]

Video recreation:

http://www.yomiuri.co.jp/stream/m_news/vn110928_6.htm (http://www.yomiuri.co.jp/stream/m_news/vn110928_6.htm)

[VH-Cheer Up]

Ah, the old "cockpit door/accidental roll" knob confusion problem...

Cabin PA: "And now, ladies and gentleman, a display of aerobatics by the first officer. Please ensure your seat belt is firmly fastened and your tray table is in the upright position".

Origato, gozeimas.

[NutLoose]

See

Wrong Button Sends ANA Jet Nearly Upside Down - Japan Real Time - WSJ (http://blogs.wsj.com/japanrealtime/2011/09/29/wrong-button-sends-ana-jet-upside-down/?mod=google_news_blog)

CGI Film of the incident is interesting

http://www.youtube.com/watch?v=LCm5kW4jzx0&


In a chilling reminder of how a simple human error can override the highest of high technology, an investigation has shown that All Nippon Airways Co. narrowly escaped a catastrophe earlier this month when its plane almost flipped over after a co-pilot hit the wrong button while trying to open the cockpit door for the plane’s captain, returning from the restroom.
The revelation came at an awkward time for the Japanese carrier, just hours after the first of the new-generation Boeing Co. 787 Dreamliner jets on which it has built its future strategy finally landed in Tokyo on Wednesday, more than three years behind schedule.
The celebratory mood was quickly over as Shin Nagase, a senior executive vice president at ANA, apologized and bowed deeply in front of TV cameras at a news conference to apologize for the trouble caused by the incident, which took place Sept. 6.
According to the Japan Transport Safety Board, two flight attendants were slightly hurt when the Boeing 737-700, with 117 people aboard, tipped more than 130 degrees to the left at one point and dived about 1,900 meters in 30 seconds during a flight from Naha on the southern island of Okinawa to Tokyo. Two flight attendants were slightly injured and six passengers became airsick or reported neck pains.

[finfly1]

What dashboard 'button' would cause such a reaction?

[NutLoose]

According to the article


ANA said a 38-year-old co-pilot likely hit the rudder trim controls by mistake instead of pushing the door unlock button to let the captain in after he stepped out of the cockpit to go to the restroom.

[angelorange]

Rudder trim issue according to Flight Global:

VIDEO: ANA 737 rolled near-inverted after rudder trim blunder (http://www.flightglobal.com/news/articles/video-ana-737-rolled-near-inverted-after-rudder-trim-blunder-362687/)

[fireflybob]

You have to be careful that when you attempt to fix one problem that you don't generate a load more.

The locked door policy was introduced to prevent unfriendly people easily gaining access to the flightdeck post 9/11.

Trouble is crewmembers now have to go through all this rigmarole just for a comfort break - it's not the first time that the rudder trim has inadvertently been used in this manner.

It's all about risk management and personally I feel the locked door policy carried extra risks.

[Short Approach?]

Just an ATCO so excuse me if I'm asking stupid questions, BUT.

Why would you have a button which if inadvertently flicked inverts the airplane?

This button adjusts my range... this button the brightness and this one blows up the building... yeah.

[Plectron]

Recovery? What recovery? Instincts? Who is kidding who? What else do you expect from pilots who might have 100 hours total time of hand flying when they retire. Much worse is on the horizon I assure you.

Couple the lack of ability with a sense of entitlement and bravado (on the ground and safely ensconced in the bar with his mates) that needs to be seen to be believed, et voilà, you have airliners rolling over on their backs.

If you haven't been in the cockpit with these guys you don't and can't understand. Once you have, it is all made clear.

Where in hell are the insurance companies and ICAO? Forget the so-called unions, management, and complicit regulatory agencies staffed with former company hacks. Don't believe me if you choose, but it all happens to be true.

I await my mandatory tongue lashing for pointing out the obvious.

[root]

Just an ATCO so excuse me if I'm asking stupid questions, BUT.

Why would you have a button which if inadvertently flicked inverts the airplane?

This button adjusts my range... this button the brightness and this one blows up the building... yeah.

The rudder trim is a round dial on the central pedestal. It is not in your direct line of sight unless you purposefully turn around and look at it.

The dial is of the continuous type. Meaning if you turn the dial to starboard the rudder will incrementally trim farther and farther into starboard.

Coincidentally the button to release the locking mechanism of the door is also a dial. It is positioned quite close to the rudder trim dial. Both dials are not in sight of a pilot who has not turned around to look at them.

In practice flight attendants will often call the cockpit to take orders for food or drinks. Sometimes people add last minute things over the flight inter-phone so the flight attendant takes a bit longer than usual to enter the cockpit. Some pilots, especially FO's are vulnerable to this, grab the wrong dial and hold it to the left. If an FO is busy reading, talking to ATC or something else he might inadvertently grab the rudder dial instead of the door dial. Because they are behind him and relatively close to his seat he won't notice this in his peripheral vision.

Holding the door dial to the left unlocks the door mechanism. You need to keep it turned to the left actively to keep the door unlocked for the flight attendant to enter the cockpit.

As you can see from the video the 737 was tipping over to port side. This coincides with my description of the operating principles above.

edit: Oddly enough: even though the dial itself is not visible to the pilot, the displacement of the yoke should be obvious almost immediately. The AP will attempt to balance the increasing out of trim position by applying opposite aileron. This becomes very clearly visible once it goes over a few degrees as the entire yoke would be turning farther to starboard as time goes by.

[Centaurus]

It takes 29 seconds to hold the rudder trim in the 737 full over to obtain full trim from neutral. It is hard to believe that even the most incompetent pilot would hold the switch that long in mistaken belief it was a door unlocking button. Even if he operated the aileron trim by mistake it would take about half as long to obtain full aileron trim and that includes pressing two switches together. Something not quite right here. Another case of the automatic monkey syndrome where lack of basic manual flying skills is practically non-existent in many airlines.

[ap08]

Another case of the automatic monkey syndrome where lack of basic manual flying skills is practically non-existent in many airlines.
Unfortunately the lack of manual flying skills is very existent ...

[SeenItAll]

Just a question for the professionals -- The information provided suggests that the plane rolled significantly in excess of 90 degrees, yet the FO was able recover it to stable flight. Does that not suggest that this FO had some pretty good situational awareness and manual (aerobatic, indeed) flying skills? Or would this recovery have been automatic?

PS: it also seems to suggest that the 737 retains a lot of the aerobatic capability (without being torn apart) that Tex Johnson demonstrated in the Dash-80 (707 precursor).

[root]

It takes 29 seconds to hold the rudder trim in the 737 full over to obtain full trim from neutral. It is hard to believe that even the most incompetent pilot would hold the switch that long in mistaken belief it was a door unlocking button. Even if he operated the aileron trim by mistake it would take about half as long to obtain full aileron trim and that includes pressing two switches together. Something not quite right here. Another case of the automatic monkey syndrome where lack of basic manual flying skills is practically non-existent in many airlines.

I tend to disagree here.

29 seconds is not a very long time. Very often the attendant or other pilot calls to gain access but takes some time to actually open the door. I've seen this happen quite a few times already. People just hold the switch to the left and keep it there until the other person actually opens the door.

I'm not sure why you are talking about aileron trim.

[EEngr]

29 seconds is not a very long time.Pilot calls to re-enter cockpit. The one on deck reaches over to turn the unlock knob. The one outside jiggles the door. Still locked, calls again (now a bit pissed). The one inside thinks to himself (while still holding the wrong control), "Dumbass. Can't even operate a doorknob." I can see 29 seconds passing with such antics.

[Denti]

Yup, 29 seconds is no all that much. And the control column displacement is not necessarily within primary field of view if the FO is trying to see something on that tiny CTTV screen on the left aft overhead panel (which means he is hanging out of his seat halfway across the center pedestal staring behind and to the left side) where he has to continuously check to make sure the area in front of the flight deck door is clear.

By the way, he didn't use full left rudder trim, it was just two left trim inputs each lasting a pretty short time, that was all it took. Check out the first report (http://www.mlit.go.jp/jtsb/flash/JA16AN_110906-110928.pdf) which is sadly in japanese but clearly shows some of the FDR data as well as the flightpath and center pedestal of this configuration.

For someone with supposedly no manual flying skills he did a pretty fast recovery though, during night without anticipating any upset. Not bad on that at all, while of course the initial action wasn't great but not all that surprising in that 40 years old flight deck, designed without much knowledge about ergonomic man-machine interfaces.

[Ashling]

Wouldn't have happened on an Airbus. Ground the Boeing !!!

[Fox3WheresMyBanana]

I'm at risk of sounding like Mr Angry here, but I am reading a lot of excuses for some pretty rubbish piloting skills.
Most of the controls on that panel are within 10cm of the rudder trim. The two controls in question are a different size and shape.
If you haven't a clue where the switches are by feel, go sit in the cockpit for hours and practice the muscle memory.
If you don't know where the switches are, look. It's not as if he was doing night AAR or anything busy.
If he wasn't looking at the panel, then why wasn't he looking at the instruments?
29 seconds IS a long time.
Coupled with AF447, I would love to know why incompetence of this level isn't being spotted earlier (e.g. in the sim)

[The Dominican]

I for one I'm not too quick to crucify, not going to get into a deep discussion of security sensitive procedures but it is not uncommon to be holding that switch for a while because your partner turn around to answer a question from the CA, it only takes a few seconds of being distracted not to notice that the ailerons are deflecting in the opposite direction to compensate until the AP gives up and then all of a sudden he found himself in an almost hard over rudder situation, single pilot, without enough tolerance in a specific orifice to put a needle trough and this didn't turn into a smoking hole on the ground, we all know that some rudder events weren't that lucky and this was with two pilots working the problem. It is not that I am not critical of the event, I'm just looking at it from the perspective that this paid hobby of ours teaches very heavy handed lessons.

Next time you are doing recurrent, most of the times one finishes early right? tell the checker running the sim to climb you to 350 and then while you are turned looking at the back of the sim, do the same thing, operate the rudder trim until the A/P disconnects, I know it is not a true test because you are expecting it and he wasn't, but it will give you an interesting perspective as to how easy this could have ended in tragedy.

[fox niner]

I tried it in the sim yesterday. Shocking how easy it is to get in an upset like that.
BTW, a few posts back there is a photo of a 737 pedestal. Our pedestal is built up in such a way that the door knob and the rudder trim are directly next to each other.

[pattern_is_full]

Just a reminder (for "fin fly" and "short approach" and others):

The 737 (like most multi-engine planes) has a very powerful rudder for yaw control with an engine failure, which can indeed turn a plane upside down and (via the inverted lift) point the nose at the ground very rapidly if activated unexpectedly.

Search the Web for "United 585" and "US Air 427" to find two fatal 737 crashes caused by a rapid roll brought on by uncommanded rudder deflection (in those cases, a mechanical glitch, rather than reaching for the wrong knob).

[rubik101]

The low-tech solution posted earlier on another thread is still relevant so I apologise only to those who have already seen it.
Pilot wishing to leave the flight deck calls the FAs and waits until one appears at the cockpit door, enters the code and is allowed into the flight-deck. Pilot leaves and door is locked behind him.
Pee and chat take place for a few minutes until the pilot enters the code and the FA looks through the peep hole and the seated pilot confirms via camera, if fitted, that it is indeed the pilot outside the door. Now the safe part, the FA opens the door by using the door handle!
Result, no stupid switch position will ever catch you out again.
Cost; zilch.

[320wonder]

i believe there's definately more to this..... hopefully they come out with a full investigation and report.

normally, if you flick the switch and open the cockpit door, after a while.. if the other crew member doesn't come in, would you keep holding it? or let it go? i would let it go and check the camera to see what is he doing.

i don't know bout the 737s... i fly the A320.. and when i flick the open door switch, i'll hear a loud "Tack" indicating that the door strikers are unlocked.

[DC-ATE]

I flew the -200 & -300 and all we had was a push button to lock/unlock the cabin door. WHEN did they change this switch and WHY ?!?!

[Denti]

Post 9/11. Operation is detailed in your FCTM.

[A37575]

Coupled with AF447, I would love to know why incompetence of this level isn't being spotted earlier (e.g. in the sim)

First of all, the vast majority of simulator type rating training and recurrent training is with full use of automatics. Thus, incompetence at manual flying would rarely show up unless an inadvertent unusual attitude just happened to occur. Apart from specific non-normals (manual reversion for example) hand flying is usually confined to the last part of an ILS and even that would be with FD and AT engaged.

Secondly, most unusual attitude training ( if practiced at all) is kept within the Boeing definition of unusual attitudes which are clearly stated in the FCTM. In fact, these are quite benign attitudes and easily recoverable. Rightly or wrongly, recovery in IMC from inverted nose down flight is rarely practiced in the simulator and it is left to the pilot to learn by reading from the FCTM and QRH sections rather than practicing in the simulator.

To have found himself in the extreme attitude described in the incident report, the copilot must have been way behind the aircraft to have allowed it to develop into such a serious Upset. In fact, it is most probable the copilot himself applied erroneous manual inputs to the flight controls that exacerbated the initial problem caused by inadvertent trim input.

With most simulator training confined to box ticking exercises designed to minimize training costs, it is no wonder that pilots are rarely given the opportunity to assiduously practice manual flight manoeuvres that require good skills such as Jet Upset and stall recoveries at high altitude as well as crosswind landings on slippery runways.

To make things more difficult for new pilots, it is rare to have a simulator instructor set a good example by personally demonstrating how to fly these maneouvres, before handing over control to the student to have a go. The assumption seems to be that if the student has a commercial pilot's licence there should be no need for an instructor to demonstrate. Not all students are aces and an instructor demonstration may be needed rather than a box quickly ticked.

[DC-ATE]

Denti-
Post 9/11. Operation is detailed in your FCTM.

Well, we had the doors locked long before "9/11", so don't know why they had to change the design of the switch. There was NO way to have this happen with the switch/lock the way it used to be.

As to my "FCTM".....it's long gone, as I retired over 20 years ago !!

[silverstrata]

Some observations for confirmation of what people have already been saying here.


a. Some 737s have the door knob and rudder trim right next door to each other.

b. The knob has to be held over, until the door is opened. That can easily be 5 - 10 seconds - easily enough overpower the autopilot for an upset.

c. The presence of a flight attendant on the flight-deck would not assist. Rather it would be a further distraction, and the f/o chats her up.

d. Requiring flight attendants on the flight deck can be counter-productive. You then have to ask 'sir' before you can go to the toilet. And the reply will often be "piss off, I'm busy with the service" (low standards airline, low standards fight attendants).

e. Some f/os would not notice if the aircraft was flying upside down. Autopilot is engaged, so the attitude indicator must be lying. (See also the lack of flying skills in the AF 447 incident, or the Indian 737 upset.).

f. The 737 is an ergonomic disaster.

g. The hydraulic pumps are regularly turned off, instead of the ant-ice switches (why are ant-ice switches 'guarded' like the more important hydraulic switches?)

h. Every aircraft is different, with many switches in different positions. Bit of a pain, when you are expecting to feel one switch, and find something completely different. Then there are also differences between the classic and NG, to add further confusion.

i. The proposal that Boeing will keep the same flight deck for the next 30 years, is an aviation tragedy. In 30 years time, it will be like airlines using the Vickers Vimy today, in the 21st century. May Boeing be economically punished for their stupidity, with airlines turning away from the 737.

j. How can a commercial aircraft without a viable control split facility still be produced and flying in the 21st century?


.

[Willit Run]

Maybe the JCAB needs to increase the ground school time from 6 months to say???? 12-14 months to cover this most perplexing,complicated,intensive task.

[Fratemate]

All this hand-wringing about 737 ergonomics and re-inventing the wheel, with cabin crew coming in while one person is on the flight deck etc is complete bolleaux. The simple matter is Numbnuts didn't look before he operated the switch; end of story. What do you do when you're going to put the eng anti-ice on? Do you randomly reach above your head and turn on a switch that feels about the right place? Of course you don't. Likewise, when opening the cockpit door, do you reach down to the centre pedestal and twist a big round knob that is in a different place, looks and feels nothing like the door opening switch, without looking? Of course you don't. This is purely and simply down to FO Numbnuts not looking before doing and it doesn't need the rest of aviation to be re-invented; rather he needs a good kicking (read educating) and instructions into 'more study' and 'more highlighting' (Japan-based pilots will understand the last bit) :).

[ATC Watcher]

The simple matter is Numbnuts didn't look before he operated the switch;

Easy , this is investigation done like 30 years ago ' the operator has srewed up . no need to change anything ".

I do not know how many millions flight hours the 737 has,and how many tens of thousands crews have flown it since 9/11, and this appears to be a one-off incident. The question is :can it happen again with a different "Numbnuts" ?
If the answer is no, then send the numbnuts to re-education .
If the answer is yes, then change something.(i.e relocate the door unlock knob elsewhere , preferably in a standard location )
End of the story.

[B412]

There isn't an aircraft type of any description that does not have the possibility of a problem from incorrect switching.

The adage "Identify", "Confirm" and then "Select", was taught years ago, and still applies. Those simple steps can be done multi crew, or single crew as was the case here.

A lack of basic airmanship seems to be the culprit here.

[silverstrata]

So, somehow, customers were permitted to reposition the newer small Rudder Trim Control rotary knob much further forward.


Lack of future planning by Boeing, leaving no room on the pedestal. Then along comes an acars PFM box, and nobody knows where to put it. So it gets jammed at the back, and the rudder trim knob gets positioned further forwards.

And that can be half the problem. You look for a rotary knob half way along the pedestal, which is nowhere near the trim knob, and then find the trim knob instead, because someone has moved it.

737 pedestals are pick'n'mix, because no two aircraft are the same. You can spend several minutes trying to get an ADF frequency out of an HF box, if you are not careful.

BTW - why do we still have ADFs in the 21st century? Can an ILS DME not beep at 4 dme, to mark the OM?


.

[rogerg]

BTW - why do we still have ADFs in the 21st century? Can an ILS DME not beep at 4 dme, to mark the OM?



Many places require you to perform an ADF approach, maybe that's the reason.

[Denti]

Don't need an ADF to fly a NDB approach, as tuning, receiving, identifying and displaying the underlying NAVAID is not required anymore for non precision approaches on the 737.

However, many ILS installations do not have a DME co-located with the ILS.

[Plectron]

Gee, with all those changes to the B737, it's truly amazing anyone can fly one safely. That darn old MD-80 and all the derivatives must be a devil for all the DC-9 guys to deal with too.

The fuel control knobs are right next to the props on many piston aircraft.

My head is spinning with the possibilities there.

[stepwilk]

Back in the old days--yeah, my days--many controls had distinctive shapes. The landing gear handle ended in a nicely molded miniature wheel and tire, the flap handle was a segment of a flap, the engine levers each had differently shaped (and colored) handles. Obviously we can't do that with six zillion knobs and switches in a 21st century air transport aircraft, but at least there was a time when people were thinking about it.

On the other hand, how many people retracted the gear when they thought they had the flaps? Or leaned the mixture past lean cut when they thought they were dialing the props back?

[renard]

I fly planes with an electric rudder trim. If you turn the trim knob full scale at cruise speeds for more than a second the seat of your pants will tell you that the plane is yawing big time. How can you loose the plot so much that you can hold that switch for 29seconds?????

[fdr]

How can you loose the plot so much that you can hold that switch for 29seconds?????

because we/they and you are human.


That is the failing, and also the saving grace... as computers also suffer form the problem of being computers... ie divide by zero etc... BSOD... "wait one..." "re-syncing".


Bad design abounds in the B737, it is not a glorious bit of design, but hard to argue with the market, it has been effective.

[silverstrata]

>>Many places require you to perform an ADF approach,
>>maybe that's the reason.

That's not the point. The point is that airfields that cannot afford a minimum of a VOR (preferably an ILS) should be off-limits to commercial traffic. This would free up two slots in the center pedestal of every commercial aircraft in the world, and make things much safer. ADF approach for an A380? You must be joking.

We started off with ADFs, in the '30s (?). Then we had VORs, ILSs, INS and GPS. You would have thought that the first on the list could be deleted by now. And an ever more crowded and complicated cockpit inevitably means ever more scope for errors. GIGO works on flightdecks, every bit as much as it does for computers. Witness the French A320 three-degree crash.

Its like comms boxes. Some numbskull decided to add yet another digit to the box, making frequencies more and more unintelligible (who was this g!t?). When any competent ergonomicist would notice that the number of frequencies (if we used all channel numbers) only required 4 digits, not 6. So frequency hand-overs are being garbled daily on the airwaves, with the attendent dangers of miss-communication, all because aviation has lost an understanding of KISS.


Not sure what someone meant by not needing an ADF for a non-precision. Dual GPS approaches are not yet authorised, so an ADF will need to be tuned and idented in such cases.

[J4CKO99]

Just a thought, have read some of the thread and can see how this happened fromt he very detailed explanations, but, am I right in saying that the incident would not have occurred in the pre 9/11 world as the cockpit door was not locked back then so the door would just be opened manually by the returning crew member ?

Potentially, to my mind, if that is the case this would be an indirect consequence of 9/11 and the changes that were made, sort of a bit like how Burglar alarms and Anti Virus Software seem to cause as many problems as threats they solve but in a much more serious way.

Would it have gone that way anyway regardless of the terrorist threat ?

[silverstrata]

am I right in saying that the incident would not have occurred in the pre 9/11 world as the cockpit door was not locked back then so the door would just be opened manually by the returning crew member ?


Quite correct. Normally, you would simply open a door-knob, like entering any office. And passengers, trying to find the loo, frequently did.

The Indian 737 upset was another security door issue, that nearly ended in catastrophe. The captain had to wait the full 30 seconds, before the door would open on the security code, while the aircraft was basically out of control. Had ther been no security door, control would have been regained 25 seconds earlier.

Aviation safety is always two steps forward, and one step back, as none of the Goons in the CAA ever think things through. TCAS was another, where the Swiss mid-air was a direct consequence of not doing a full analysis of the possible consequences of this new technology - and so a safety measure ends up killing people. Comms frequencies are the same, as I have just illustrated. The number of times I have heard an aircraft go off frequency because of these complicated numbers, at the most inopportune moment, leaving ATC to struggle with a non-responsive aircraft.


Aviation has long been 'Tombstone Engineering', but it looks as though it has now become 'Tombstone Health and Safety' too.

[stepwilk]

We started off with ADFs, in the '30s (?). Then we had VORs, ILSs, INS and GPS. You would have thought that the first on the list could be deleted by now.

But then how would you listen to the ball game?

[Lord Spandex Masher]

Are you serious?

A five digit number becomes complicated and confusing by the addition of one more digit...which is always either a 5 or a 0?!

Do you get confused between flight level 35 and flight level 350 too?

[silverstrata]

>>>five digit number becomes complicated and confusing by
>>>the addition of one more digit...which is always either a 5 or a 0?!


All I can say is that you have never flown into LHR or LGW at peak hours. What is easier to recall, 132.475 or 3247?

And the proof of the pudding is the number of missed frequency changes, especially while in 'foreign' regions. I count about six or ten missed frequencies every day. Thats six or ten opportunities for extra ATC stress and a potential incident. And for what reason? Because some desk-jockey has only half a brain and did not ask users what they want.

Besides, when you are on a very busy frequency, with no room to get a word in edgeways, what is simpler and shorter to say - 132decimal475 or 3247 ?? So ATC are struggling to control the numbers of aircraft in thier sector, because of unwieldy phraseology. KISS wins every time, while complexity leads to incidents and accidents.

[Lord Spandex Masher]

I've just timed myself saying both of your examples. 1.0 second for the long one. 0.8 seconds for the short one.

Not exactly a life saving difference.

Both are just as easy to recall, for me anyway, having just heard them.

Funny you should mention those airports. I've been based at both, at separate times, for a total of 15 years.

[Denti]

Not sure what someone meant by not needing an ADF for a non-precision. Dual GPS approaches are not yet authorised, so an ADF will need to be tuned and idented in such cases.


Nothing to do with GPS, however i suggest you check FCTM page 5.39 (Raw Data Monitoring Requirements). Raw data monitoring is recommended but not required both for classics (EFIS update 7.2 and FMC version > 7.1) and NGs. The MEL allows dispatch with the ADF inop as long as one FMC is operative even for NDB approaches.

[silverstrata]

Nothing to do with GPS, however i suggest you check FCTM page 5.39 (Raw Data Monitoring Requirements). Raw data monitoring is recommended but not required both for classics (EFIS update 7.2 and FMC version > 7.1) and NGs. The MEL allows dispatch with the ADF inop as long as one FMC is operative even for NDB approaches.


Sorry, but you cannot do an ADF approach without an operating NDB and ADF - period !

If you do not have a ground station and the appropriate cockpit instrument, you are doing an RNAV FMC approach, not an ADF approach. And we have no RNAV approach plates in our Euro-Jepps, and so I'm presuming that RNAV (FMC/GPS) approaches are not yet certified to any of the airfields we visit.

Your reference in the 737 FCTM is referring to whether "raw data" or "map display" is selected - not whether the ADF is working or not. In other words, you can use map display mode, if you have overlaid VOR or ADF pointers for cross-checking (which many systems can do) or you flip between raw data and map display to check that the map is in the right position.

Note the note at the bottom of this section: - "Compare VOR and ADF systems to detect possible map shifts". You cannot do a map cross-check, if the VOR or ADF is u/s or not fitted !!


This is what happens, when people use a (map-shifted) RNAV FMC approach, instead of the raw NDB/ADF.
1996 Croatia USAF CT-43 crash - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/1996_Croatia_USAF_CT-43_crash)


Ditto the B-Med Airbus going into Addis Abeba. The FMC plot and terrain display were off by some 3 miles, leading to a go-around just 60 ft over the high terrain. Now while the VOR would give occasional signals of its unreliability (the VOR was the primary fault), the FMC gave no indication that it had a map-shift. This was the biggest complaint in the report, that the FMC knew it was getting bum VOR information, and therefore may have a map-shift, but did not bother to tell the pilots. And the pilots could not check for a map-shift, because they were comparing bum data with bum data.
Report: British Mediterranean A320 at Addis Abeba on Mar 31st 2003, wet VOR nearly causes two CFITs (http://avherald.com/h?article=42cba6a2&opt=0)

Incidentally, the New Scientist report on this indicated the crew were highly influenced by the terrain display, which showed them passing nicely through the valley. However, the terrain display was likewise mapshifted. Not sure where this info came from, because the AAIB report underplays this aspect (that the crew were merely following the FMC and terrain plot, and disregarding the raw VOR signal).

One would have thought that following an erroneous and displaced VOR radial would bring you eventually to the VOR, whereas this flight paralleled the VOR inbound, which is what would happen if you followed an erroneous FMC position.

Search Results - page 1 - New Scientist (http://www.newscientist.com/search?query=Aircraft+risk+navigation+error&fromdate=&todate=&rbauthors=&rbissueno=2409&resultview=keyword)



And I am not even sure that GPS-RNAV is entirely a solution at present (as was recommended in the B-Med report). It works well with high accuracy and reliability, but Europe is still concerned that the US may degrade GPS signals or switch the system off during periods of international tension. This is the primary reason for Europe developing and launching the Galileo GPS system. When that is up and running, and there is full confidence that erroneous signals can be detected and warnings given, then Europe may proceed to RNAV approaches.

[A37575]

The 737 (like most multi-engine planes) has a very powerful rudder for yaw control with an engine failure, which can indeed turn a plane upside down

Slight exaggeration maybe?

Operation of the rudder trim does not give full rudder deflection. The two accidents you mentioned were thought to have been caused by full un-commanded deflection of the rudder but in the reverse sense to the movement of the rudder pedals. Very few pilots could have recovered successfully in the limited time at their disposal in both those events.

But even with full rudder trim deflection applied inadvertently, the 737 in the ANA incident would never have got to the extreme attitude that happened without some crazy manual flying inputs by the pilot.

[stator vane]

i've read as many of the replies as i could and though i have seen it implied, would i be out of line to say in simple terms, this was the work of a complete idiot, jobsworth, waste of space pilot?

yes, we have all touched the wrong switch and have pushed the wrong button but if we are simply paying attention to the present situation, we see the wrong result and make a correction right away. but to hold any knob, switch, handle regardless of the shape for such a length of time and not notice the aircraft beginning to roll over is the work of a total idiot.

can't blame the airplane for this one!

[silverstrata]

But even with full rudder trim deflection applied inadvertently, the 737 in the ANA incident would never have got to the extreme attitude that happened without some crazy manual flying inputs by the pilot.


I think you negate the role of the autopilot.

During the two left rudder-trim inputs, the autopilot will have been counteracting with aileron - hence the slight opposite right roll to start with. Then the autopilot will go out of limits and disconnect, and so the resulting uncorrected roll rate to the left may well be quite dramatic.

[BOAC]

Exactly as SS says - I know of one previous incident caused by mis-management of fuel balancing where the bank angle exceeded 90 degrees on A/P disconnect.

[lederhosen]

New captain at big airways and he was over the alps at the time iirc!

[grimmrad]

Because he thought that this is the door knob: suspected train of thought: "Ah captain wants back in. Lets open the door - uh oh, something is wrong - why is the captain not coming, help! Lets kick the button a bit more so he finally gets into the cockpit - uh oh, its getting worse - c'mone door, open.. (Keeps on turbning the button frantically..."

[stepwilk]

why is the captain not coming, help! Lets kick the button a bit more so he finally gets into the cockpit - uh oh, its getting worse - c'mon door, open..

You can't be serious.

Well, maybe you are...I recall some of your previous posts.

[Artificial Horizon]

I can understand what has happened, the FO would have activated the rudder trim and would have then been distracted by looking at the aircraft deviating away from normal. That initial confusion was no doubt replaced by panic when the aircraft had a serious 'upset'. It would have been when action was taken to recover the aircraft that he actually 'released' the trim. No doubt had was frozen on there whilst brain tried to process what the hell was going on. I think some recognition could be given to him for actually recovering the aircraft back to 'normal' with out further 'upset'. We have already seen several times in recent history that in these situations a bit of mishandling in 'upset' situations can lead to the loss of the aircraft. Yes he co*ked up getting into the situation but thank god had enough basic handling skills to recover from it. These basic skills are what seems to be lacking in certain areas.

[Tee Emm]

recognition could be given to him for actually recovering the aircraft back to 'normal' with out further 'upset'. We have already seen several times in recent history that in these situations a bit of mishandling in 'upset' situations can lead to the loss of the aircraft. Yes he co*ked up getting into the situation but thank god had enough basic handling skills to recover from it.

Think again. He already has caused a major upset (unusual attitude) by putting the aircraft into an almost inverted nose down situation of his own making. . And does this therefore prove he had enough basic handling skills in the first place? I don't believe that for one minute. He then eventually gets the aircraft right way up. I would say that was the result of pure bloody good luck - not good basic handling skills. The facts will eventually be published although maybe kept in-house. The guy has a pilots licence but that's about all.

[sheppey]

Found the following on a website blog. Does the 737-700 AP have a CWS function?
Re the ANA 737-700 incident. There is not much to go on because only media reports are available at present. So who did what to which is unclear. Maybe when the F/O hit the wrong switch (he got the rudder trim rotary knob) the door didn't click open or whatever the door is supposed to do to unlock. Someone mentioned earlier it takes 29 seconds for the rudder trim to move from centre to full scale left or right.

The ANA incident happened at night. This is significant as we shall see. Assume the F/O held on full rudder trim maybe not for 29 seconds but a significant amount of time. Eventually the autopilot either went into CWS roll as the control wheel turned to maintain the present heading - or the autopilot disconnected altogether causing the aircraft to roll. It maybe the first officer then lost control when he tried to hand fly at high altitude at night in IMC on instruments. Most of his 737 flying would have been on full automatics from shortly after lift- off to shortly before touch-down.

Another possibility is if the CWS roll did remain in place (in other words it had not yet disengaged itself) and the unusual position of the control wheel alerted the F/O to something going wrong. If he then tried to correct to wings level too harshly with CWS roll engaged it may have caused the autopilot to disconnect in the roll mode and pilot confusion with hand flying (lack of instrument flying skills) could eventually lead to an upset.

Or if the F/O inadvertently hit the stab trim switch on his control column at any stage while the AP was engaged in any mode, the AP will disengage in all axis leaving the F/O to hand fly and he probably did not have the ability to do so. It may have been a different story in day VMC but to have a typical F/O steeped to use automatics throughout his career on the 737, then a IMC recovery is going to be a real problem.

[AlphaZuluRomeo]

Am I missing something ?

Some FDR traces are available on the PDF (http://www.mlit.go.jp/jtsb/flash/JA16AN_110906-110928.pdf) by the japan agency.
The duration of the rudder trim inputs are shown on those traces.

Why do people still elaborate theories about "29 seconds, what a fool to keep that for so long", when the traces show 2 inputs, each one with a duration of ~5-7"

29 seconds is the time it takes to go from trim neutral to full trim. Nowhere was it said that in ANA's story the knob was held for 29 seconds!!

[SMOC]

Are you serious?

A five digit number becomes complicated and confusing by the addition of one more digit...which is always either a 5 or a 0?!

Do you get confused between flight level 35 and flight level 350 too?


Thread drift,

Lord SM, if you've done some human factors in terms of short term memory you should know that the magic number is 7.

One, two, three, decimal two, seven, five uses it all and doesn't leave space for heading, speed and/or altitude changes to be included.

Also the fact you've done 15 yrs around London doesn't mean everyone has, some of us fly in Europe less than once a month.

The brain also uses chunking in short term memory such as twos (1,2) threes (3,4,5) and fours (5,6,7,8).

132.475, two chunks 3247 one, leaving more short term memory for HDG / SPD & ALT in the same instruction.

Now lets talk about fatigue ...............

[Pontius]

One, two, three, decimal two, seven, five

On a point of pedantry, one doesn't pronounce the 3rd decimal place. Thus, your frequency should be pronounced "One, two, three, decimal two seven" (leaving apart, that it, the pronunciation of "tree, fife" etc :)

[The Dominican]

Am I missing something ?

Some FDR traces are available on the PDF (http://www.mlit.go.jp/jtsb/flash/JA16AN_110906-110928.pdf) by the japan agency.
The duration of the rudder trim inputs are shown on those traces.

Why do people still elaborate theories about "29 seconds, what a fool to keep that for so long", when the traces show 2 inputs, each one with a duration of ~5-7"

29 seconds is the time it takes to go from trim neutral to full trim. Nowhere was it said that in ANA's story the knob was held for 29 seconds!!

Of course not, but pilots just love to talk out of body cavities that weren't meant to talk out of

[Denti]

@Pontius: check your knowledge of the relevant R/T procedures:


Amendment 80 to ICAO Annex 10, Volume II

Amendment 80 to ICAO Annex 10, Volume II - Aeronautical Telecommunications - became applicable on 24 November 2005. It introduced a procedure requiring all VHF voice communication channels to be indicated by the use of 6 digits (4 digits for the case of channels ending in two zeros), irrespective of whether 25 or 8.33 kHz channel spacing is used, e.g.:

118,005 specified as "ONE ONE EIGHT DECIMAL ZERO ZERO FIVE"
118,025 specified as "ONE ONE EIGHT DECIMAL ZERO TWO FIVE"
118,100 specified as "ONE ONE EIGHT DECIMAL ONE"

The use of the term "CHANNEL" for 8.33 kHz channels was discontinued.

[silverstrata]

Denti.
Nothing to do with GPS, however i suggest you check FCTM page 5.39 (Raw Data Monitoring Requirements).



No answer, Denti, about this novel idea of doing ADF approaches without an ADF?? !!

I presume you are a training captain with a well-known low standards airline. They were mostly crap instructors.

.

[Trojan1981]

Is there a new feature being rolled out on Boeings enabling reverse flight? According to the Sydney Morning Herald in Australia, "Images from the flight recorder on the Boeing 737-700 showed the plane veered to the right and then sharply to the left, before dropping backwards."

ANA plane plunges after pilot presses wrong button


Like a tailslide? Just to put the fun back into airline flying... Now I'd pay to see that!

[FairWeatherFlyer]

For visual comparison, a picture of two knobs (thanks to Randy):

A Passenger Airplane Basically Flew Upside Down Because of a Dumb Pilot (Updated) (http://gizmodo.com/5844628/a-passenger-airplane-nearly-flew-upside-down-because-of-a-dumb-pilot)

It turns out there's a few variations on a theme:

Japan's ANA: Crew error sent jet into nose-dive - Article Comments - View topic - General Discussion - Article Talk : Denver Colorado Neighbors (http://neighbors.denverpost.com/viewtopic.php?p=2129345#p2129345)

Does anyone know which one matches the ANA panel?

[Two's in]

I was sure this was just a Far Side Cartoon...

ANA pilots unaware for 17 seconds that plane was almost turning upside down ? Japan Today: Japan News and Discussion (http://www.japantoday.com/category/national/view/ana-pilots-unaware-for-20-seconds-that-plane-was-almost-turning-upside-down)

TOKYO —

The pilot and copilot of an All Nippon Airways plane were unaware for 17 seconds that their aircraft almost turned upside down during a flight last year, the Japan Transport Safety Board (JTSB) said Friday in a report on the incident.

The ANA flight, with 117 passengers and crew on board, dived 1,900 meters in 30 seconds in the incident off the southern Shizuoka district on Sept 6, 2011.

The maneuver happened when the co-pilot, in trying to unlock the cockpit door for the captain who was returning from the toilet, mistook a command button for the cockpit door lock switch nearby.


http://i302.photobucket.com/albums/nn98/hiiyah777/wingsye5.jpg

[kenhughes]

Been there, done that. :hmm:

http://www.pprune.org/tech-log/463127-ana-japan-roll-incident.html

[BobnSpike]

I think the post was more about the cartoon than it was the actual upset event, and a recent report on the incident was inserted to provide context.:}

I have also been told my life would be much less complicated if I would stop thinking.

[ACCP]

I've always felt the Door Lock and Pitch Trim Cutout buttons were a bit too close together on my machine. And they look the same!!

[Pugilistic Animus]

That cartoon is the funny I hope they just let it go...I think in history simialr events have happened numerous --flap/gear mix ups on older pistons and such...The cartoon is accurate, as to how some disasters are caused...by the OEM not thinking...:\

[nelsonmadiba]

Hi,

Just read on the italian news Corriere della Sera that an All Nippon 737 went into a dive by mistake.
Apparently the pilot instead of opening the door to the captain,who had gone to the loo,pushed a wrong button and put the aircraft into a steep dive,losing about 6'000 feet in 40".

Has anybody heard anything?could it be just fake or maybe exagerated?

Cheers

[AlphaZuluRomeo]

Not fake, but hardly a "news" today...
The event occured sept. 6, 2011, and I think a search will pull some topic about it in the forum ;)

[cldrvr]

We all know you Italians like to take things a bit slower, but this is taking things to a whole new level. How you guys doing paying of your trillion euro debt? Still on track for 400 years?

[pattern_is_full]

This is a wonderful exemplar of a typical Boeing-style upset,
Yaw x Roll = Dive!
No pitching moment needed for the resulting steep nose-low attitude.

Hmmm - did you have in mind an example of a plane (except maybe symmetrical-airfoil aerobatic planes) that won't end up nose-down after a yaw/roll to inverted?

[fleigle]

Those darn Boeings, rolling and upsetting people, they should just cause the vert stab to fall off and fix the problem permanently.!!