Turkish Airlines crash: evidence points to pilots

This morning the Dutch investigators are holding a press conference on the Turkish Airlines Boeing 737-800 crash at Amsterdam last week and I understand that later Boeing will release a safety bulletin. From a reliable source, the contents of the Boeing bulletin will include the points below. Regrettably they don’t reflect well on the pilots, who are dead of course – but I’d caution that there is a long investigation still to come.

  • no evidence of fuel shortage, birdstrike, icing, windshear, wake turbulence, or engine, system or control malfunction
  • the first officer was initially flying the aircraft and was inexperienced in airline operations
  • autopilot and autothrottle were in use
  • the aircraft was initially high and fast on the approach and at about 2,000ft above ground the throttles were pulled to idle
  • the authrottle went to “retard” mode and the throttles then stayed at idle for about 100 seconds during which time the speed fell to 40kt below reference speed
  • the aircraft descended through the glideslope with the captain talking the first officer through the before landing checklist
  • the stick shaker activated at about 400ft above ground and the first officer increased power
  • the captain took control and as the first officer released the throttles they moved to idle due to being in “retard” mode
  • after six seconds the throttles were advanced but as the engines responded the aircraft hit the ground in a slightly nose-high attitude
  • throughout the episode the left-hand radio altimeter read negative seven feet altitude, but the right-hand radalt worked correctly

Boeing will warn crews about fundamentals like flying the aircraft, monitoring airspeed, monitoring altitude, and will give advice about radalt issues.

, ,

81 Responses to Turkish Airlines crash: evidence points to pilots

  1. Paul March 4, 2009 at 10:44 am #

    Why I don’t get is why the throttles remained in retards mode and why TOGA wasn’t activated.

  2. Stevy March 4, 2009 at 11:49 am #

    I think the most important question is not rather why they didnt activate TOGA but why and eventually HOW did they get off the glideslope so badly? What about the CVR?

  3. orko March 4, 2009 at 12:14 pm #

    please wait for official report about the crash. please do not speculate the facts.

  4. Mathijs March 4, 2009 at 12:48 pm #

    Could you please provide more info on your source to make it more reliable? For example a sample copy of the bulletin

  5. Kieran Daly March 4, 2009 at 1:00 pm #

    I understand your question but I’m sorry, I’m not able to help further at the moment. There is an investigators’ press conference starting in a few minutes and later we will publish more material on http://www.flightglobal.com from that. Apologies again.

  6. Magnum March 4, 2009 at 1:05 pm #

    on the last bullit. I guess Boeing is aware of the fact that the crash location is app. 15ft below sealevel

  7. Ian Powell March 4, 2009 at 1:34 pm #

    Ok, the bif question:- what”s the manual say if the 2 radio altimeters are different?!

  8. Joeri March 4, 2009 at 1:35 pm #

    Can somebody explain what “retard” mode is?

  9. Bangalore Aviation March 4, 2009 at 1:37 pm #

    With full respect to Boeing, it has to try to protect its interests, and I am in full agreement with Orko. Speculation serves only vicarious interests not serious aviation folks. We should wait for official communications and results of a proper investigation.

  10. Ian Powell March 4, 2009 at 1:42 pm #

    “We should wait for official communications and results of a proper investigation.”

    But the results are just out. I repeat my question. What’s the flying manual say if the 2 radio altimeters don’t agree?

  11. Hans March 4, 2009 at 1:42 pm #

    1st Why did they let him go 40knts below reference speed ???
    2nd Why did they release the thrust levers after the stall warning and gave them the abillty to retard again ???

  12. Dicks Airbus March 4, 2009 at 1:43 pm #

    The findings in short. Points to radalt defect and this went unnoticed by FO due to last minute checklist operation

    * Capt LH seat, (fully qualified) FO in RH seat, observer FO on jumpseat
    * Normal approach, no changes, to 18R
    * Descent on AP, normal procedure for TK
    * At 1950 ft LH RadAlt indicated -8 ft and passed this info on the A/T
    * From CVR: crew get aural landing-gear warning but not consider this a problem
    * A/T pulled throttles closed (At 1950 ft. because A/P thinks aircraft is landing)
    * stick shaker at 450 ft (Because no reaction by pilots on closing of throttles by A/P)
    * From FDR: full power was then applied
    * FDR stores 25 hours, in this case 8 flights, same problem had occured twice previously before previous landings.

  13. Hugo March 4, 2009 at 1:46 pm #

    The things said in the press conference a few min ago sounds like the same stuff that’s written up here.

  14. Ian Powell March 4, 2009 at 1:53 pm #

    >1st Why did they let him go 40knts below reference >speed ???
    They didn’t know there was a problem with the speed. (They didn’t realise that the A/P had closed the throttles)

    >2nd Why did they release the thrust levers after the >stall warning and gave them the abillty to retard >again ???
    The thrust levers were released when the pilot took control.

  15. jaaph March 4, 2009 at 1:56 pm #

    @Dick:
    last item: which problem had occured before: the RadAlt?

  16. Dicks Airbus March 4, 2009 at 1:59 pm #

    A classic combination of hardware and human failure.

    Few points:
    - Why was crew not watching speed, etc.? What was 2nd FO doing, reading a paper?
    - Definitely an issue with A/T status change feedback (Boeing will need to fix that)
    - Design issue when two radalt provide a certain level of inconsistant data (Boeing will need to fix that)

    Sad, sad, sad.

  17. Mookie March 4, 2009 at 2:00 pm #

    Whoops. Looks like Boeing carries most of the blame.

  18. Ian Powell March 4, 2009 at 2:05 pm #

    >Why was crew not watching speed, etc.? What was 2nd FO doing, reading a paper?

    Very harsh. I would have simply said, why didn’t any 3 pilots notice the air speed.

  19. Dicks Airbus March 4, 2009 at 2:07 pm #

    @jaaph: referring to the radalt failure reading on capt side.

    Yes, perhaps Boeing is now lined up for class action suits. The system clearly does not cross reference the other altimeter.

  20. Dicks Airbus March 4, 2009 at 2:22 pm #

    FROM: THE BOEING COMPANY
    TO: MOM [MESSAGE NUMBER:MOM-MOM-09-0063-01B] 04-Mar-2009 05:29:01 AM US PACIFIC TIME
    Multi Operator Message

    This message is sent to all 737-100,-200,-300,-400,-500,-600,-700,-800,-900,-BBJ customers and to respective Boeing Field Service bases, Regional Directors, the Air Transport Association, International Air Transport Association, and Airline Resident Representatives.

    SERVICE REQUEST ID: 1-1228079803
    ACCOUNT: Boeing Correspondence (MOM)
    DUE DATE: 10-Mar-2009
    PRODUCT TYPE: Airplane
    PRODUCT LINE: 737
    PRODUCT: 737-100,-200,-300,-400,-500,-600,-700,-800,-900,-BBJ
    ATA: 3400-00

    SUBJECT: 737-800 TC-JGE Accident at Schiphol Airport, Amsterdam – 25 February 2009

    REFERENCES:
    /A/ 1-1222489391 Dated 25 February 2009

    Reference /A/ provides Boeing’s previous fleet communication on the subject event. The US NTSB, FAA, Boeing, the Turkish DGCA, the operator, the UK AAIB, and the French BEA continue to actively support the Dutch Safety Board’s (DSB) investigation of this accident.

    The DSB has released a statement on the progress of the investigation and has approved the release of the following information.

    While the complex investigation is just beginning, certain facts have emerged from work completed thus far:

    - To date, no evidence has been found of bird strike, engine or airframe icing, wake turbulence or windshear.
    - There was adequate fuel on board the airplane during the entire flight.
    - Both engines responded normally to throttle inputs during the entire flight.
    - The airplane responded normally to flight control inputs throughout the flight.

    The Digital Flight Data Recorder (DFDR) data indicates that the crew was using autopilot B and the autothrottle for an ILS (Instrument Landing System) approach to runway 18R at Amsterdam Schiphol airport. During the approach, the right Low Range Radio Altimeter (LRRA) was providing accurate data and the left LRRA was providing an erroneous reading of -7 to -8 feet. When descending through approximately 2000 feet the autothrottle, which uses the left radio altimeter data, transitioned to landing flare mode and retarded the throttles to the idle stop. The throttles remained at the idle stop for approximately 100 seconds during which time the airspeed decreased to approximately 40 knots below the selected approach speed.

    The two LRRA systems provide height above ground readings to several aircraft systems including the instrument displays, autothrottle, autopilots and configuration/ground proximity warning. If one LRRA provides erroneous altitude readings, typical flight deck effects, which require flight crew intervention whether or not accompanied by an LRRA fault flag, include:

    - Large differences between displayed radio altitudes, including radio altitude readings of -8 feet in flight.
    - Inability to engage both autopilots in dual channel APP (Approach) mode
    - Unexpected removal of the Flight Director Command Bars during approach
    - Unexpected Configuration Warnings during approach, go-around and initial climb after takeoff
    - Premature FMA (Flight Mode Annunciation) indicating autothrottle RETARD mode during approach phase with the airplane above 27 feet AGL. There will also be corresponding throttle movement towards the idle stop. Additionally, the FMA will continue to indicate RETARD after the throttles have reached the idle stop

    Boeing Recommended Action
    - Boeing recommends operators inform flight crews of the above investigation details and the DSB interim report when it is released. In addition, crews should be reminded to carefully monitor primary flight instruments (airspeed, attitude etc.) and the FMA for autoflight modes. More information can be found in the Boeing 737 Flight Crew Training Manual and Flight Crew Operations Manual.

    Operators who experience any of the flight deck effects described above should consult the troubleshooting instructions contained in the 737 Airplane Maintenance Manual. Further, 737-NG operators may wish to review 737NG-FTD-34-09001 which provides information specific for the 737-NG installation. Initial investigations suggest that a similar sequence of events and flight deck indications are theoretically possible on the 737-100/-200/-300/-400/-500. Consequently the above recommendations also apply to earlier 737 models.

  21. Anonymous March 4, 2009 at 2:27 pm #

    At least I hope certain turks stop calling us and our Towercontrol racists. That stuff really hurt us. And I expect more blaming to come. This is so sad. We are in the Age of Protectionism already. There will be only losers. Peace.

  22. Peter March 4, 2009 at 2:33 pm #

    Still too early for a final judgement,but human factors seem to be very important here.Whatever malfunction,speed is life so watch it,and don’t let the autopilot push you under the glidescope.Seems awfully fundamental to me,and I’m only a private pilot.

  23. Dicks Airbus March 4, 2009 at 2:44 pm #

    Well, if the RA showed -8, it is not failed (as far as the system goes), therefore the AP will not disconnect.

    This is the weak point that started the sad event (a minus reading should be a defect). Guess the system engineers at Boeing will have to take a look at negative numbers & cross reference checking of radalt.

    In addition the crew did not help by the fact that they seemed to rely on the automation just too much. Automation is a supporting system, like fuel – checking still needs to be done as if these systems were not present.

    Sad chain of events based on hardware and human error. We can all learn from this.

  24. Dicks Airbus March 4, 2009 at 2:52 pm #

    A note: radalt shows -2 when on the ground (0 when touching positive pitch altitude).

    Why at RA reading of -8 the AP was not auto-disconnected is a system design failure imho.

    Appeantly this was a single autopilot approach flown with AP B.

  25. Kieran Daly March 4, 2009 at 2:54 pm #

    Thank you to everyone for your intelligent contributions – greatly appreciated.

    I have to say I think those people suggesting that Boeing carries most of the ‘blame’ are not being realistic in view of the conditions on the day.

    Huge questions over CRM here, to put it at its gentlest.

  26. MusicRab March 4, 2009 at 2:56 pm #

    >Sad chain of events based on hardware and human error. We can all learn from this.

    “human error” is harsh at this stage. Hardware AND software errors certainly prime candidates.

  27. Dicks Airbus March 4, 2009 at 2:58 pm #

    @Keiran Daly: Yes, I totally agree with you. Over 100 seconds the crew did not notice the speed decay – what were they doing instead of monitoring the a/c?

    AP kept it on GP, AT was thinking the a/c is on the ground and in idle.

    Definitely very serious CRM issues that let this happen.

  28. ddb March 4, 2009 at 3:47 pm #

    Comments from a layman:
    Boeiing needs to do some adaptation ?
    - when the two LRRA have different readings over a certain limit, shouldn’t there be a visual an auditive warning system. Autopilot should take in account the measurements of both LLRA and, when different, activate a kind of alert-state where lots of parameters are cross-checked and the pilots warned.
    - Why is autopilot not checking the duration of the autothrottle RETARD mode during final approach ? It shouldn’t last for 100 seconds, no ?
    - Apparently autopilot also isn’t checking the minimum reference speed for landing ?
    I thought autopilot was much “smarter”.

    Pilots were probably busy and to much relying on autopilot systems. Although i’m far from being a pilot,
    I’m also running on autopilot rather frequently. Man-machine interaction (or lack of) should be reconsidered ?

  29. pieter March 4, 2009 at 3:52 pm #

    What is the logbook telling about the two times it happended before?

  30. imagun March 4, 2009 at 4:04 pm #

    Dutch press statement – english version (pdf)
    http://www.onderzoeksraad.nl/docs/rapporten/Persverklaring_4_maart_GB.pdf

  31. stEve March 4, 2009 at 5:03 pm #

    Clearly the autopilot should have more sophisticated software to detect erroneous data. However, as a private pilot of a Cessna Skyhawk, I can report that the most basic responsibility of the pilot on landing approach is to maintain adquate airspeed to avoid a stall. I check the airspeed every few seconds on short final approach.

    The crew is at least as much to blame as the autopilot for this tragedy. They NEVER should have allowed the airspeed to get so low. That rule is absolutely fundamental even to a beginning student pilot, let alone a team of 3 professionals.

  32. Sandy March 4, 2009 at 5:14 pm #

    Pilots are there to control the aircraft whether via the autopilot or manually. It is not unusual for trainee pilots too get training on commercial flights (ie with passngers on board). This seems innocuous enough when all is going smoothly, but as we have seen in this case it is a lethal practice. It should be outlawed. Only fully competent pilots should be in the cockpit. I write as a professional pilot and flying instructor on multi-engine jets, with decades of experience.

  33. It's NOT Boeing's Fault March 4, 2009 at 5:15 pm #

    The Autothrottle and Autopilot are separate systems. You can land the 737 manually (without Autopilot as you’d expect in an Autoland situation) yet keep the Autothrottle engaged! Boeing recommends against leaving Autothrottles on during manual landings but many do it. That’s another issue. When the Radar Altimeter, likely the Captains, reads below a specific value (I forget, like 15′) the Autothrottle goes into a Retard Mode “thinking” the plane is about to land. Retard is Idle power. That simple. Being high and fast, a.k.a. the proverbial Slam Dunk, onto a Glide Slope, is an edgy experience and is best done with the Autopilot clicked off. Autothrottles ON will protect you on the slow side but in this case, the Captains Radar Altimeter was broken. Things break. Crew has to be attentive. Not Boeing, forget it. How many hours and years have 737s been flying safely? Broken gage, loss of situational awareness, misundestaing of fairly simple automation. Sad.

    Colgan crash had no autothrottles and go too slow.

    Type B-737,757,767,777

  34. Robert March 4, 2009 at 5:45 pm #

    How could anyone blame Boeing for this? The PIC is responsible for getting the plane on the ground. Instruments fail. It’s not like the tail fell off. Fly the airplane.

  35. Dicks Airbus March 4, 2009 at 5:57 pm #

    @By It’s NOT Boeing’s Fault: Please quote where you read they came in high and fast (this is not the case, they were perfectly lined up on GS).

  36. Steve March 4, 2009 at 6:14 pm #

    Re Not Boeing:

    although I go with most of what you write, the following DOES worry me: the software bluntly makes a critical decision based on a sensor that believes it is already under the tarmac.

  37. hans March 4, 2009 at 7:32 pm #

    It is a general rule that safety related software/electronics shall not rely on a single system. Safety related computers are based on the voting principle: the majority decides. The minimum is a 2 out of 2 system: the outcome of 2 computers are compared by a fail-safe comparator and only if both give identical information the required action will be carried out. If the information is not identical an alarm or – if possible – a safety shutdown shall be initiated. A 2 out of 3 system is preferred. If one malfunctions, the other 2 still have the majority and can take care of a safe automatic operation. 3 out of 5 systems are very common in safety related applications, even in car industry. Common mode errors (e.g. a failure of a common power supply for all systems) shall be prevented. Systematic design errors shall be prevented as well, computers based on diverse technology and software are preferred for safety critical applications. This all is common practice in safety technology. I can’t imagine that that the autopilot of a Boeing airplane is not based on this principle and that it relies on just a single altitude meter.

  38. Mike March 4, 2009 at 9:18 pm #

    @ Hans: I like to hear what you state, because that’s how one would like these systems to operate, but the AT decided to go to idle based on ONE faulty radalt. The other radalt was simply not monitored by the AT. Ans unfortunately neither by the three pilots.

  39. MM March 4, 2009 at 9:34 pm #

    In my opinion, the statement below is the sole reason for the crash.

    - the captain took control and as the first officer released the throttles they moved to idle due to being in “retard” mode

  40. Steve March 4, 2009 at 10:50 pm #

    Hans, thanks for your explanation. Now I understand why the safety board say they will focus on the radalt-A/T coupling: they are not done with this yet. Hopefully they’ll also find a decent explanation as to why the crew let the speed drop unnoticed.

  41. JimmyJazz March 4, 2009 at 11:08 pm #

    The pilots were stuck in RETARD mode.

    Retards.

  42. Tony Y March 4, 2009 at 11:31 pm #

    Posts above suggest a single autopilot was engaged. Also there is insufficent details at this stage to suggest something more certain.

    I had an instance when on a dual channel approach the autopilot pitched 15 deg nose down at 800 feet, rate of descent increased rapidly without warning, next GPWS was activated,so a prompt go-around was initiated without further incident. Also numerous times I have had he autopilot play up once engaged on the LOC/GS because of signal interference ie aircraft moving around the ground instalations distorting the signal. The most recent one was when an A380 was given clearance to line up whilst we were on approach, the automatics went berserc trying to follow the LOC/GS indications which were erratic.
    I have had on a single Autopilot approach in IMC bank angle 40′deg followed by +10′ pitch followed by -10′ deg pitch.
    In such a critical phase monitoring of the automatics is of the essence as sometimes they can do some unexplained things.

    Retard mode, in cruise reaching top of descent in VNAV the autothrottles move tp idle position. Retard mode is activated on a dual channel approach, and is activated passing 27′feet RA. Activated only if FLARE mode engages if it doesn’t then both autopilots disengage.
    Two independent radio altimeters provide radio altitude to its respective flight control computers.With an Radio altimeter inop the autopilot disconnects 2 seconds after Loc/gs capture. Isn’t in amsterdam the platform height 2000′ ft for ILS procedure?

    As for speed reversion the Retard mode is followed by ARM mode, provides minimum speed protection.The AFS commands a speed 5 kts greater than minimum speed. Reaching 5kts greater than minimum speed reactivates MCP speed selection control.The AFS commands nose down to increase speed if thrust levers are not advanced.

    I can only highlight somethings that I have observed in my experience. Those are :mode confusion, type of approach used, complacency, unnecessary chat especially during approach, Radio altimeters MEL’ed, conduct of approach, briefing for it or lack of, what things to look out for mentioned to trainee(this should take place at topd not during approach as I have witnessed) etc are just some of the things that we do not know at this stage. Lets not forget especially on training flights, at low level if you don’t watch it closely things can get rather demanding in a mere second so staying ahead of it all is of paramount importance.

  43. Arjen March 5, 2009 at 4:16 am #

    Human or interface or designers of the interface, who is to blame, the neverending discussion. The interface was wrong, the designers did not design enough security valves and the humans did not act appropiately to the chain of events because they ignored warnings and relied on the autopilot/autothrottle to handle everything correctly. As a flyer and a computer program designer I can understand all the viewpoints expressed, but as a physician I can not. As every physical measure has a 95% confidence interval, one in every 20 measures is wrong. The first thing physicians learn is never to rely on numbers. If the patient can talk, has a firm pulse and is feeling as warm as I am, the chance that the patient will die in seconds is zero. No numbers needed. Any intern that has the guts to rely on a low blood pressure on a monitor without looking at the patient to check whether the monitor is showing the actual blood pressure has a serious problem. No patient would accept that a physician would say, well I treated you wrong because the monitor was faulty. After this somewhat lenghty and probably boring comment as a relative outsider I have to conclude that the only responsibles were the pilots. As I understand two pilots were going through a checklist when the fatal events occured, but what was the third doing then? Landing is a critical procedure. If I were to walk away from an unstable patient at the intensive care to discuss the charts with the intern without asking a third party to warn me for crucial changes, what would be the verdict in court? The only conclusion can be that these kind of errors are allways human. The next step is to kill the designers that were stupid enough to connect throttle control to a single faulty altimeter without cross checking the other flight parameters. I would say that the only indication that an airplane is on the ground is when the suspension of the landing gear is compressed. Let’s put a sensor there! (untill it becomes faulty and we can begin the discussion all over again).

  44. Arie March 5, 2009 at 7:48 am #

    A layman here. Why is nothing being said about the cockpit voice recorder. At what stage did the pilots become aware of what was wrong? Did they say anything?

    And, since it happened before according to the flight recorder. Was it the same pilot? If not, what did he/they do, and most importantly, did they warn anybody?

  45. Oscar Polat March 5, 2009 at 7:57 am #

    I think it is a serious f***-up by both Turkish Airlines for not fixing the faulty altimeter and by Boeing, how on earth the system relies on one altimeter and cuts down the engine power? Unbelievable, I am not a pilot I am a stress engineer, so I am just wondering if there are three altimeters on the aircraft, wouldn’t you say read from all three and make a decision based on that?

    By the way, I have read a comment by a Dutch guy who was complaining that Turks call them racist, but you are, I am sorry to say this, I think it is true, I am a Turkish stress engineer once a Dutch engineer once called me a monkey (meaning not as intelligent as a Dutch person I guess :) . But may be it is true, how on earth Turkish Airlines did not fix the faulty altimeter and why the pilots ignored the warnings?

    It is really sad!

    P.S. If a Dutch guy is reading this, I don’t think women are inferior to men and I don’t oppress my women, just in case you did not know. :)

  46. Francis D March 5, 2009 at 10:34 am #

    @Oscar Polat on March 5, 2009 7:57 AM

    From (another) Dutch guy (total passenger layman but interested):

    1- I don’t think Turkish ppl in general are calling us rascist (there are still many moderates here in Holland despite the success of mr. Wilders)
    2- Having said that, its undeniable that there are big cultural differences worldwide in dealing with blame and guilt (which is the point the “dutch guy” seems to be referring to, however bluntly)
    3- Those differences should not impede us from assessing the causes objectively and learning from the outcome. Something that, in this stage, is still premature pending more info (like VCR transcript)

    Respect and regards to all!

    PS Good to hear u think women are equal (they are ;) )… however, u seem to have more than one? (“I don’t oppress my woMEN) Or is that a typo? ;) Any position on man-man relations?

  47. Francis D March 5, 2009 at 10:55 am #

    @Kieran Daly

    My apologies for the last message being a little off-topic. As a layman I really enjoy reading this blog tho, very interesting.

  48. Another Dutchman March 5, 2009 at 11:37 am #

    Some Turkish papers blame the Dutch investigators for covering up any Dutch mistakes. Partly based on the assumption that the pilots are heroes and should not be blamed at all…
    A bit of an one-sided view, it seems, but could there be some cover-up for the Dutch traffic controllers? Shouldn’t they have seen that the plane was way too low, and have warned sometime during that 100 seconds period?
    Is it peculiar that so far nothing was stated about this aspect?

  49. roeland March 5, 2009 at 1:05 pm #

    By JwS on March 4, 2009 8:23 PM

    you write that most of the netherlands is below sealevel; well that’s true. However, a LRRA measures the height to ground, so it doesn’t care if you’re having your field elevation at -8ft, or +300 ft. That only counts when you are referring to QNH.

    Ok, being an ex aircraft avionics engineer — the trigger was maybe the #1 RA, but not the cause. The crew has been asleep, It’s a CRM issue.

    Also, nobody noticed that the yellow door selectors must have been in manual? This also indicates that at engine-start/take off, the list haven’t been completed either. This also supports the crew being not too good at their work.

  50. Old Lizzy March 5, 2009 at 1:32 pm #

    Once more I am reminded there is rarely if ever a single probable cause to an accident. We all go through CRM training and have the Swiss cheese slide drilled into us about several relatively minor issues/failures lining up nicely to create a “perfect storm” scenario.

    We have had a relatively good record for some time now in the safety section of the industry and it is now shattered by several recent accidents, all of which of course continue to be investigated, but which all lets be honest carry a single message – AVOIDABLE.

    Common themes? Crews distracted/fatigued, not recognizing a failure and responding appropriately. Compacency is always an enemy of an automated cockpit. I have flown 777, 320/319, and it is awfully easy to sit back, relax and work on 5 across or 7 down. You simply cannot stare at instruments all day long. Whereas flying the venerable 727 for instance, kept you on your toes – it needed flying and talking to.

    I would argue we are starting to pay the price for high tech airliners that do not present subtle failures very, well, subtly. We are paying the price for fiscally driven minimum pilot training. We are paying the price for outdated flight time and duty regulations that do not reflect the demands made on a human body that has not advanced like the planes it operates. We are paying the price for the dumbing down of our profession.

    And then in the midst of all these avoidable wrecks comes an event that makes a good case for buying a lottery ticket. The Hudson ditching due to all engine failure/degradation. Incredible odds taking out both engines, followed by benchmark decision making, teamwork and flying skills, followed up by near perfect evacuation and rescue. Made us pilots all feel good for sure. Until Buffalo, and until Amsterdam….back to reality ladies and gentlemen – we are not so good most of the time.

    And then I go back to my own carrier, where we do not have the issue of low time pilots, but we have incredibly low morale. We have been beaten into a pulp by bankcruptcy, 9/11, corporate decisions made by uncaring rapists who should have trouble looking at themselves in the mirror. I fly a 767, with arguably one the simplest fuel systems imaginable – two engines and three tanks. Yet we cannot seem to go a few weeks without having a major fuel event, such as a crew taking off without any boost pumps working, (suction feed works!) flameouts due to crossfeed mismanagement (maybe suction feed does not work!) and the result? Near total thrust loss due to failure to burn fuel from the center tank prior to wing tank selection….. trust me, it really is very simple, or at least it was…..

    All of these fuel mismanagement issues have resulted in more and more SOPS being written, instead of fewer. Example – we are now not allowed to use the crossfeed without referring to a 5 page (yes, five page) flight manual checklist now.

    As I say, no single probable cause at all……

    Let’s get back to some basics everyone.

    Lizzy

  51. Martijn March 5, 2009 at 3:11 pm #

    Does anyone realise that the ground levels at this particular airport are actually -8 Feet?

    (the airport is build on the bottom of an old lake)

    I am not an expert, but assume a radalt is measuring independent of any other equipment so not using any reference to the actual height of the ground, so this must be coincedence?

  52. Cem ARGUN March 5, 2009 at 3:17 pm #

    Obviously Boeing is responsible for a faulty or risky autopilot system. Assuming the RA is active at about 2500 feet, at 1950 feet, a steep change of 1958 feet should have been considered as a major fault and the other RA should have been cross referenced, also the negative value for an absolute to the ground measurement should have been considered as major fault and the autopilot should have been deactivated.

    Turkish Airlines is responsible as they have permitted their airplane to fly and especially landing in autopilot, knowing that it had the same RA defect within the last 25 flight hours. It doesn’t matter whether it’s been fixed, as they did not clearly understand the real cause of the problem.

    Pilots are sure responsible in not noticing a speed half the reference speed for 100 seconds, especially knowing that the RA is defective.

    What about ATC? They have radar & ILS. They supposedly knew the approaching plane’s distance to the airport & its speed with their own proper tower equipment, contradicting the faulty RA altitude communicated by the plane to the tower. Or is it not communicated instantly?

    Boeing has mission critical design flaws, Turkish Airlines has mission critical safety operating procedure flaws, the pilots have mission critical judgment & equipment monitoring flaws. Why on earth the tower has not warned the pilots within 100 seconds? Can all the parties involved might be in such a paralysis?

    Is it a coincidence that this faulty altimeter accident happeneed at Schiphol Airport, one of the few maybe the only major international airport below sea level? As Boeing engineers have not taken into account a negative RA reading as a fault, it might very well be that they haven’t taken into account that airports might as well be below sea level.

    Faulty RA, stupid auto-pilot, airport below sea level, “In God We Trust Airlines”, 3 sleepy pilot, paralyzed ATC and only 9 dead. Now this must be fate.

  53. Martijn March 5, 2009 at 3:42 pm #

    @ Cem ARGUN

    Exactly. Screening some of the discussions, i can’t find anything related to the fact that the airport (and surrounding area) are below sealevel.

    Could this be an accident related to software problems similar to the Y2K issue where software was reaching “zero”?

  54. CAPT. IMAD HASBINI March 5, 2009 at 4:01 pm #

    ITIS THE RESPOSIBILITY OF THE CAPTAIN PERFORMING THE TRAINING AND THE PILOT IN THE JUMP SEAT AND THEN THE TRAINEE FIRST OFFICER AS 100 SEC PASSED AND NO BODY WAS MONITORING THE SPEED AND THE ENGINE INSTUMENTS WHERE THE PILOT DOING THE TRAINING SHOULD HAVE HIS EYES ON THE FRONT PANELS EVEN IF HE WAS READING THE CHECKLIST. FOR ME I CONSIDER IT IS 99% PILOT ERROR AND 1% SYSTEM ERROR AS WHEN THE SYSTEM FAULTED A LOT OF WARNINGS CAME OUT BUT WAS NOT REGARDED CORRECTLY BY THREE PILOTS.

  55. CAPT. IMAD HASBINI March 5, 2009 at 4:10 pm #

    IT IS ALSO BOEING TO BE BLAMED AS THE SYSTEM HAS A DEFECT AS THE PILOT FLYING WAS USUNG AUTOPILOT2 AND AUTO THROTTLE 2 SO THEY SHOULD RECIEVE INFORMATION FROM NO2 RADIO ALTIMETER AND DISREGARDS NO1 RADIO ALTIMETER. AND THERE SHOULD BE A WARNING TO PRONOUNCE THE RADIO ALTIMETER DIFFERENCE AT LEAST BELOW 1000 FEET ALTITUDE.

  56. Edward March 5, 2009 at 4:53 pm #

    @ Martijn
    LRRA measures the height to ground, negative reading indicates instrument failure.

  57. 737 captain March 5, 2009 at 6:20 pm #

    Anyone ever seen the video from Capt. Vandenberg……..

    If you are not sure what the aircraft is doing….. click click, AP off AND A/T off, first fly the aircraft, go to a safe altitude (above MSA ESA)

    100 seconds is a lifetime in aviation terms.

    CRM is certainly one of the major issues here!

    Safe flying to all of you

  58. Joe March 5, 2009 at 11:11 pm #

    A negative reading on a radio altimeter does not indicate a failure. They’re typically set up so that they read 0 at main gear touchdown on approach, and as the nose rotates down, the value goes slightly negative. Most aircraft, sitting still on the ground, will show a radio altitude that is negative by several feet.

  59. stEve March 6, 2009 at 4:43 am #

    I continue to be shocked as more details come out. Why weren’t the pilots monitoring airspeed ??? The most basic and critical item to check continuously is airspeed when on short final approach to landing. Every student pilot is taught that.

    It turns out that when the stick shaker activated indicating an imminent stall, the 1st officer applied full throttle, but just seconds later the captain wanted to take control. For 6 seconds during the change of control, the engines were again at idle as commanded by the autopilot. How could the pilots allow that ? It’s beyond me. When the captain again applied full power, it was too late and a second or two later the plane stalled and dropped like a rock to the ground. Luckily it was already low enough that most survived.

  60. Ahmad Dan-Hamidu March 6, 2009 at 10:18 am #

    A very unfortunate chain of events…triggered off by just one faulty RadAlt. I was wondering…if one RadAlt was faulty and the other was ok, couldn’t the flight augmentation computers have rapidly done a kind of vote against the conflicting RadAlts by maybe…doing a quick check on what GPS had to report in terms of Radio Height?

    Another thing that could be considered is, a review of approach procedures. I also don’t think its possible to know before hand if a RadAlt is working or is going to screw up the way this one did.

    It almost reminds me of an accident I had on my old “MS-Flight Simulator ’98″. I was approaching Lima’s (Peru) airport at a comfy, level, 4,000ft. I was combing the coast at 150IAS with full flaps and gear-down, waiting to capture the localizer for a right-hand turn, to be followed by glideslope capture. All was going smoothly until -like the RadAlt which decided to “step”- my scenery files also decided to “step”. All of sudden, the scenery files error made the ground come rushing up to me without my Altimeter showing a corresponding loss of height (re: approaching a hill/mountain causes Alt to fall even though VertiSpeed remains at “zero”). It happened twice; the first time I escaped by climbing…I actually heard the tires touch the “rising ground” for a split second as my 734 pulled up in time. The second time, I was caught off guard due to monitoring of the localizer indicator’s slide. I could pull up in time and…the ground “hit me”.

    How could I know the scenery would screw me up, how could the Turkish crew know that their RadAlt would screw them. They should not be blamed entirely unless you want to criticize them for not distrusting the flaw as soon as it reared its ugly head. I would have distrusted at first sign of anomalous behaviour, disengaged auto-whatever, handled manually…and always have GPS informing me of Lat/Lon/Alt even though it wouldn’t be engaged.

    Please tell me what you think of my opinion via email or Yahoo-360. Cheers.

    PS: I’m not a pilot…yet.

  61. Edward March 6, 2009 at 3:11 pm #

    @ Joe
    Sorry, Just wanted to point out that the below sealevel story is not relevant, and negative reading at that speed/before tochdown indicates possible error.
    What do you read app say 100 sec before tochdown? ;-)

  62. Sharat Chandrasekhar March 7, 2009 at 7:15 am #

    Quoting CAPT. IMAD HASBINI;

    IT IS ALSO BOEING TO BE BLAMED AS THE SYSTEM HAS A DEFECT AS THE PILOT FLYING WAS USUNG AUTOPILOT2 AND AUTO THROTTLE 2 SO THEY SHOULD RECIEVE INFORMATION FROM NO2 RADIO ALTIMETER AND DISREGARDS NO1 RADIO ALTIMETER.

    No. One hates to speak ill of the departed, but it is sadly, the careless pilots that are to be blamed for this accident – Plain and Simple. There is no plausible excuse for an experienced flight instructor with 15000 hours not to notice a speed bleed down to stall during approach.

  63. Norbert March 7, 2009 at 3:59 pm #

    Is recovery of a stalled 737 still possible at 450 ft? If I stall my glider, I have learned to push the nose down instead of pulling the nose up, because pulling up just further decreases speed and further increases the stall. I realize my glider is not a 737, but shouldn’t the principle be similar?

  64. Zafer March 7, 2009 at 10:31 pm #

    Working in the railway signallization, I know almost all of the rail systems have some kind of fail-safe or redundant systems to prevent such conditions. For example when the signal light turns to red, you make sure it is red, not another colour. It is hard to imagine that the autopilot of this state-of-the-art aircraft obeys to the output of a single altimeter??. 2-out-of-2 is the minumum required. Is it too expensive to put several altimeters and several autopilots in the aircraft and make sure all their results are the same? I would like to talk to the engineers who made the design. Make sure that the electronics systems are working as designed before blaming the humans.

  65. Dirk van der Laan March 10, 2009 at 12:24 pm #

    Hello DD, you have asked many questions that undoubtedly many people with little knowledge about aircraft operations have.
    There are two issues here. First, who is to blame? Second, why has this crash happened? The answer to the first question is in this case simple. The failure of an instrument, that is mainly important for the autopilot, should never have lead the crew to lose sight of the speed and stall. “Always keep flying the aircraft” is one of the basic rules among pilots and the crew hasn’t and the captain is ultimately responsible for that.
    However the second one is more interesting and harder to answer and still under investigation by the Dutch TSB. Key attention goes to the failing radio-altimeter which has had consequences for the automatic operations of the aircraft that many people, pilots included, didn’t know about.
    It becomes too technical to explain what the consequences were, but the main thing is that the captain assumed to auto-pilot to prevent the aircraft from flying too slow and it didn’t.
    Had he known what was happening of course he would have turned off the autopilot. However, since he didn’t know and also was training his co-pilot, he and the other crewmembers failed to notice the failing autopilot and the decreasing speed.
    And when they finally did notice and reacted swiftly it was too late.

    Finally, to answer your question about whether the tower shouldn’t have alerted the pilots, the answer is NO. Speed is the responsibility of the pilots and of them alone. Air traffic control can give advice and notify, but their main job is to keep airplanes from flying into each other.

    And … I don’t see who is hiding anything. I see just a small instrument-failure, bad luck and late reaction of the pilots. That’s all.

  66. joe March 10, 2009 at 5:36 pm #

    Pilot error….. end of story!!! Sad but true!

  67. Dirk March 10, 2009 at 6:52 pm #

    Joe,
    … end of story only when you need a stick to hit the pilot with. Start of story if you want to learn from it. Choose what you would like your critics to choose when you make a mistake.
    “Let he who is without sin cast the first stone”

  68. Karel Idema March 11, 2009 at 11:11 am #

    One of the first actions recovering from a stickshaker or stall condition is AP and AT off and the power levers full forward (against the stop and not TOGA!!). Don’t worry about egt’s in the red. The thrust returned however to idle and remained there for 6 precious seconds as the AT was still in command. I guess the aircraft was at least 10 seconds or more deprived from emergency thrust when badly needed. This last mistake, again not operating according the Flight Manual, might have been the fatal one.
    Certainly this will be checked in the simulator.

  69. Ahmad Dan-Hamidu March 11, 2009 at 2:28 pm #

    Ummm, a thought just popped up:
    Is it possible to transfer FADEC (Full Authority Digital Engine Control) from obtaining altitude data from either RadAlt manually?

    If its possible (at the pilot’s discretion), then it could have been possible to avert the crash by “telling” (settings of course) the auto-throttle (FADEC) to disregard the faulty left RadAlt and rely on the right RadAlt.

    …just a thought.

  70. dirk March 11, 2009 at 3:59 pm #

    Of course a captain with thousands and thousands of hours experience doesn’t know the standard stall recovery.
    Kept the AP and AT connected for fun and didn’t touch the throttles ‘caus he loved crashing.
    Hmmm…

  71. Karel Idema March 12, 2009 at 3:11 pm #

    Let it be clear the autopilot B did a perfect job so the aircraft did not leave the glideslope otherwise the aural glideslope warning would have been triggered alerting the crew. The pitch was increased further and further to maintain the GS causing a speed drop as no power was applied There was no malfunction in the autothrottle system itself. It received information from RA 1 it was time to close the throttles and it did. This action must have been announced on the PFD with IDLE or RETARD i.s.o. SPD in the Flight Mode Annunciators. This annunciation as well as the speed decay, the pitch increase from 2 to 14 degrees, the lack of normal engine RPM and sound, the lack of rapid throttle movements, usual during approach with ATS on, it all went unnoticed by three crew. Why? Is Boeing to blame for this mishap?
    The crew was not intending to make an autoland as is suggested many times. An autoland requires two AP’s in command.
    Furthermore all indications of the primary flightinstruments are compared constantly and a warning will be triggered when appropriate. A RA is not a primary flightinstrument.

  72. Trails March 12, 2009 at 5:16 pm #

    ‘Click – Click’
    ‘Click – Click’

  73. retired guy/investigator March 12, 2009 at 9:43 pm #

    Also answers to Dicks Airbus comments, several additional comments and facts,
    1) Amsterdam Airport Schiphol landing altitude is 15 feet below sea level (-15ft)
    2) Captain (was called very experienced) operated as check pilot for the co-pilot who had prof-check to accomplish
    3) Captain at RH seat as PNF (doing Nav/Com), co-pilot at LH seat (actual handles a/c) as PF
    4) “learner pilot” at observer seat in front of Flight Deck door (obviously not at all observing)
    5)AP/AT/Auto Land is NOT a Boeing design, I asume captains LRRA, who senses the Alt. didn’t go gradually from 2000 to -8ft but must have dropped fast?, System should sens this malfunction and disconnect!
    6) 100 seconds during the final approach (when especial extra attention is needed) is a very, very long time not to observe this and not interfear?
    7) Too bad, lessens learnt for all of us during future flights.
    8) No Flight Log enterings were found about previous Altimeter misreadings recorded at the last 25 hrs, according FDFR it happened twice, it would not hard to find out who/which pilots are guilty for this ignorance! They must feel most guilty for this crash

  74. Ali Oztas March 14, 2009 at 6:51 pm #

    As a frequent flyer and Gas Turbine control engineer (industrial applications), I would like to give some comments about this “hot” subject.
    1) Technical
    Design philosophy of altimeter should comply with minimum avionics safety standards. There are numerous ways to configure redundant initiators (instruments) such as voting, average and median etc. Next step is designing the sequencing software in such a way that if one or more initiators fail, the system should go to its fail safe status and disengage auto control systems. At the same time a single failure of an initiator should not lead to catastrophic failure of the asset (airplane). The captain is obviously is not a control engineer and his primarily task is flying the airplane. I am confident that post implementing these designs in the early test stages, the Boeing Company would have performed fully functionality checks. Sometimes due to commercial reasons an/or technical reasons, final use of this design might be limited. I understood that Boeing Company has their RCM (reliability centered maintenance), IPF (instrumental protective functions) for designing and maintaining their systems. We should trust prestigious company like Boeing for all their compliances.
    Post complete functionality tests of the avionic system, any revealed design deficiencies will be undergo risk assessment for safety and reliability compliances versus the cost. Remaining findings and design philosophy will be translated to operating manuals for the flight operators. The operators should then undergo competency development programs (simulator training, class room trainings etc.)

    2) Human factor
    What ever robust design of the avionics system, there will be always a human factor for success of failure of safe operating your asset (airplane). During transient operations (landing and take off), the captain should monitor the critical instruments such as this altimeter(s).

    It appears that we all are focusing on the consequence of a failure (faulty LRRA as confirmed during the initial investigations). Did we find the root cause of the LRRA failure? Was this instrument maybe wrong calibrated or maybe not calibrated as per maintenance intervals? Failure modes of these instruments are known by the designers, premature failure modes could be excluded. It appears that, LRRA is giving the wrong reading in low altitudes (in this case). How about the readings on higher altitudes and their associated control and operating philosophies? Is same LRRA used for higher altitudes and is there any evidence of faulty readings?

    Let’s quit blaming each other! We should learn from mistakes and share our knowledge for the sake of safe flying.
    My advice for the virtual civilization “warriors” (Dutch Guy and the Turkish): we are living in a multicultural world and we have to respect each other and avoid virtual civilization clashes, at least on our level as professionals.
    From understanding comes respect, do we understand each other?
    TEMPUS FUGIT…

    Ali Oztas

  75. James Gow March 19, 2009 at 10:12 pm #

    For now, give the the flight crew, and their families, a break.

  76. Alan March 25, 2009 at 12:41 pm #

    When I learnt that the difference between flying and falling was but a few knots the law prohibited me from carrying passengers. It seems that the passengers must now pay for our industry to re learn.

  77. chris May 31, 2009 at 8:33 pm #

    unless i’ve missed it : where’s the vital evidence from the cockpit voice recorder(s)& the flight data boxes(s), which should explain lots of things….before prejudging ?

  78. PAMILA December 3, 2009 at 1:48 am #

    That is some inspirational stuff. Never knew that opinions could be this varied. Thanks for all the enthusiasm to offer such helpful information here.

  79. Gwen August 23, 2013 at 12:19 pm #

    Good day! I just would like to give you a huge thumbs up for the great info you have got here on this post.
    I will be returning to your web site for more soon.

  80. Sanford September 9, 2013 at 9:01 am #

    Everything is very open with a very clear clarification of the
    issues. It was truly informative. Your website is very useful.
    Many thanks for sharing!

  81. Fabian November 5, 2013 at 7:15 am #

    This paragraph provides clear idea in favor of the new people of blogging,
    that genuinely how to do blogging.

Leave a Reply