American Morning

Tune in at 6am Eastern for all the news you need to start your day.
June 8th, 2009
09:45 AM ET

Was Air France jet flying the wrong speed?

[cnn-photo-caption image= caption="CNN talks with an Airbus expert about what might have caused last week's Air France crash."]

Sixteen bodies have now been recovered from the wreckage of Air France Flight 447, but the cause of last week's crash is still a mystery. New reports say at least a dozen other flights traveling in the same area at around the same time had no problems with weather conditions.

John Cox is president and CEO of Safety Operating Systems. He’s flown Airbus jetliners and is familiar with their controls and systems. He spoke to John Roberts on CNN’s “American Morning” Monday.

John Roberts: What do you make of this idea that some 12 other planes flew through that area? Air France had four flights, a couple from Sao Paulo, and others went through that area of thunderstorms without any problems.

John Cox: Well I think it says the weather itself is probably not the single cause for the accident. This accident, just like all I've been associated with, will end up being a series of events. We'll end up learning that series one piece at a time to understand what happened.

Roberts: Air France sent out some literature over the weekend regarding one of the air speed indicator sensors, the Pitot tube, and there might be some problem with that. With your knowledge of Airbus…What do you know about problems with anomalous speed readings?

Cox: The fleet has shown a little bit of this issue to come up before. I'm aware of three or four previous cases in fleet history but it's typically been short duration. What we learned by the information that the airplane up-linked to the Air France maintenance facility, is they had a lot of very confusing signals that the pilots would have been confronted with. And exactly what caused that, it could possibly be Pitot tubes or air speed indications that would be causal in some way but why the crew was not able to satisfactorily determine which of the air speed indicators was bad, there's a procedure for it. Where that procedure either wasn't acted on properly or failed, that's going to be something the investigators are really going to have to look into carefully.

Roberts: According to Air France, there have been some problems with this sensor icing over in certain conditions. Could that potentially have happened at an altitude of 35,000 feet, flying through the thunderstorm area? Could that sensor possibly have iced over and therefore given incorrect speed readings?

Cox: It's possible. Usually at cruise altitude it is so cold you really don't get a lot of icing. But this particular sensor has shown on rare occasion this icing issue to come up at cruise before. I think the fact there's something unique here, that as an accident investigator, when you find something unique you really want to concentrate and look to see if that could have an effect on the accident.

Roberts: There has been a problem, as you mentioned, at cruise. This showed up in October of last year with a Qantas airliner that was flying in the South Pacific area. Three hours into the flight, everything seemed to be going fine. The computer which controls the aircraft with the autopilot on suddenly got these different speed readings and went haywire thinking it was going too slow. It put the plane into a dive to pick up speed. Apparently 100 people were injured as they were knocked into the roof of the cabin. Could something like that potentially have happened here, and while the pilots recovered from that maneuver, if you’re flying around thunderstorms with updrafts, cross currents and lightning in the area, could that have maybe caused the catastrophic cascade of events?

Cox: I don't think so. In the case of the Qantas airplane, the failure was from an air data inertial reference unit, known as an ADIRU. It broadcast some spurious information and the airplane reacted, the autopilot system attempted to follow it. Although the ADIRU on the Air France accident airplane also displayed some problems, it was a failure-type message. So I don't believe we're going to see anything that would connect these two. The Air France airplane has a number of system failures that the Qantas airplane did not. Also, these ADIRUs are pretty reliable and you don't see that. Additionally, the autopilot on the Air France airplane had disconnected, so you put all that together and I don't believe we're going to end up seeing a correlation between the Qantas and Air France airplanes.

Roberts: What do you think about the potential for foul play here?

Cox: As an investigator, until you know exactly what caused it, you need to leave everything on the table. I wouldn't rule it out but there's been nothing right now that shows me it would be a prime suspect.

Filed under: Transportation
soundoff (79 Responses)
  1. C. Viewer

    There is another possible theory that needs to be looked at that makes total FBW aircraft susceptible to total electrical failures regardless of the number of redundancies .

    September 5, 2009 at 2:55 pm |
  2. JHoran

    A possibility is that some recently added electronic components had a reduced floating point error capacity which when mixed in with the overall system resulted in singularities shutting down key components

    August 22, 2009 at 3:57 pm |
  3. Chi

    I blame the Bermuda Triangle

    June 10, 2009 at 5:05 pm |
  4. Rob

    Can someone talk to the 'never to exceed' speed, and how it may or may not relate to this tragedy? One of the 'experts' on one of the news shows (maybe CNN / HLN, etc.) talked about about very narrow flight envelope of safety – something like 10 knots too fast and the plane breaks apart, and 10 knots too slow and it will stall. This seems like too tight a margin of error for disaster for a commercial aircraft. Thanks – Rob

    June 10, 2009 at 11:49 am |
  5. Kenny Strydom

    Having no time to radio in anything indicates almost instant catastrophic failure.
    They will calculate the position of impact of the aircraft very accurately by using the known wind and water currents, calculating location from the location of bodies and debris found.
    Every body already found will tell investigators if they were inside the plane when it crashed, or if they were free falling (terrible either way).
    They already know much more than we do, the results will definately be very interesting.

    June 9, 2009 at 2:25 pm |
  6. Chih Lai, Ph.D.

    To reply Nick's comment on June 8th, 2009 7:37 pm ET.

    I don't know which phone number to call. So, if anyone here thinks this information might help and knows where to call, please picks up your phone.

    June 9, 2009 at 12:29 pm |
  7. michaelfury

    Maybe while they are searching the bottom of the Atlantic they will find these black boxes:

    June 9, 2009 at 10:50 am |
  8. YanG

    I realize that current black box has a pinger, but you have to be within reasonable distance, i.e. near where the box is, in order to pick up the signal. Seems self-defeating if you do not know where the box is. The practical solution would be to either have the data recorded remotely (dont buy storage argument, computer memory is cheap & you can recycle unnecessary data), or have a distinct beacon which will transmit far enough to help rescuers/investigators to locate reckage. Accidents in air travel will happen, its a certainty, not a possibility. Besides concentrating on safety prior to the disaster, manufacturers/airlines have to have a better plan, then what we are faced with right now.

    On another note, did you know that US navy has a sonar system which allows isolation of any sound in the Atlantic. Left overs from Cold War, curious as to why rescuers are not utilizing it. These were used in location US Sub Scorpion as well as Titanic wreckage.

    June 9, 2009 at 7:09 am |
  9. Ken Walker

    Until the data recorders or "black boxes" are found, analysed and reported on, all comments will be speculations, informed or otherwise. These speculations can add value .... so let's be pragmatic.

    I think most people have been interpreting the information transmitted by the aircraft (ACARS) to Air France in one linear fashion. I do not believe that the speed and altitide recording "pitot" tube malfunction initiated the problems.

    My take on the reporting of errors (including the wrong readings from the pitots) were the result of a spiralling aircraft after losing its vertical stabiliser (tail fin) and then literally 'falling out of the sky'.

    What caused the start? Well, that is still in the realm of speculation and it could be anything from the severe storm cell, metal fatigue, incendiary device etc. but most likely a shear caused by the atmospheric conditions it flew into.

    Then there is the 'coffin corner' that it was most likely in. At an altitude at which a drop in speed would cause a stall or an increase in speed would damage the airframe. They may have reached that space by climbing to overcome the weather and the speed (velocity) changes could have precipitated by the weather.

    June 9, 2009 at 2:08 am |
  10. Nick

    From all comments here, I think the most interesting is from Chih Lai, about the possibility that other aircrafts might contain flight data about 447 in their collision avoidance systems memory. I hope the gentlemen made his phone calls.

    June 8, 2009 at 7:37 pm |
  11. Stephen

    Nicole, if you are wondering why an experienced pilot would ignore his GPS speed, you may want to read up on the difference between air speed vs ground speed. I don't think the GPS speed is really relevant to the discussion.

    June 8, 2009 at 4:53 pm |
  12. Chris G.

    Bad airspeed readings can fool pilots and computers alike into doing something wrong. I was flight engineer in a flight where the pitot tube froze, and what happens is that the airspeed indicated in the cockpit slowly decreases, leading the pilot to think he's approaching a stall. In my case, he put the nose down to increase speed, in the middle of some mountains, even though the co-pilot was telling him the speed was ok (GPS) and I was telling him the pitot tube heater had died.

    June 8, 2009 at 3:55 pm |
  13. Colin form la

    "Why don’t black boxes have a built in self-inflation device that would effect itself when the black box is submerged…. Wouldn’t it be easier to find the black box if it was floating???"
    I was thinking the same thing.

    Interesting discussion here, aside form the 'flames' regarding Airbus vs Boeing.

    June 8, 2009 at 3:31 pm |
  14. Dave H

    Why don't black boxes have a built in self-inflation device that would effect itself when the black box is submerged.... Wouldn't it be easier to find the black box if it was floating???

    June 8, 2009 at 3:14 pm |
  15. Rick

    Re: transmitting all data in real time-not really practical. Some of the black boxes can collect up to 2,000 data points, and do so ten times per second. That’s a lot of data, especially when more than 99.99999% of it is useless, from routine flights.

    Good point but isn't there a way to compress transmissions? And the digital box at the receiving end would only have to retain the last 30 minutes of transmission, just like the black boxes do today. Seems to me the benefits outweigh the technical obstacles.

    As to the comment that the pilot's union opposes Big Brother listening to cockpit conversations I think there's an easy solution. Recordings would only be reviewed in the event of an accident. That would happen anyway so I don't see any additional intrusion.

    June 8, 2009 at 2:49 pm |
  16. Colin in Florida

    Several points to make:

    Re transmitting all data in real time-not really practical. Some of the black boxes can collect up to 2,000 data points, and do so ten times per second. That's a lot of data, especially when more than 99.99999% of it is useless, from routine flights. The newer digital black boxes generally work just fine. The problem here is that the ones we are interested in are probably 10-15,000 feet below the surface. By the way, they do have 'pingers' on them, that transmit a sonar pulse. The battery is good for about 30 days.

    Re: Fly by wire. Boeing and Airbus do not use identical systems. They may do the same thing, but they each developed their own software. While fly-by-wire is obviously the present and future, but it IS software and WILL have occasional problems. Example: The new F-22 Raptor fighter's software 'crashed' when they first flew the plane to Japan, as it did not handle crossing the international date line correctly. In another example, an airbus A320 crashed in 1993 because "The aircraft skidded off the end of the runway after landing. The aircraft touched down with sink rate low enough (i.e. a really soft landing) that the onboard flight computers did not consider it to be "landed," which inhibited (i.e. the software prohibited the pilot engaging) thrust reverse and brake application for nine seconds." Here is an instance of the pilot wanting to do the right thing (thrust reversers and brakes), but the software prevented him from doing it! Oops.

    As to the Turkish Airlines crash, it 'appears' (cause not yet fully determined) that the radar altimeter malfunctioned and said the plane had landed, while it was still at 400 feet. The auto throttle system pulled the engines back to idle, to provide a better landing, but the plane stalled. This 'appears' not so much as a system failure, but poor maintenance and management, as the radar altimeter had been logged by MULTIPLE previous pilots as having given the wrong data intermittently. It was 'checked' but the mechanics could never get it to fail, so they kept using it. Bad management compounding inadequate maintenance.

    As to flying in lightning storms, plane are hit y lightning all the time. The principal of the Faraday cage protects the interior.

    As to GPS speeds, the idea is fine, but there are several problems-GPS speeds are averages spread over a second or two, so which would you believe if the GPS speed said X and the pitot speed said 1.5X speed? Also the problem of too much information to the pilot in a short period of time.

    As to the air data internal reference unit (ADRIU), while I know nothing about them but what I have read in the press, I do know something about systems redundancy design. It appears that the ADRIU has three airspeed inputs. The theory is that if one fails, you still have two working. But the problem is that if one fails, but now the two inputs give different readings, which is right? You have now way of knowing. It actually takes five redundant systems in something like this to provide complete redundancy. Trust me on this-I would explain it but it takes way too long.

    By the way, Max is very, very incorrect in his statement that "1996-2005 Airbus was involved around the world in only 3 deadly accidents but Boeing in 32 !!!" From 1995 to 2005, Airbus suffered 7 fatal accidents. Boeing had, by my count, 57. But Airbus is a very young company (first delivery wasn't until 1974). Boeing had thousands and thousands more planes flying in the time period Max mentioned. In other words, Max's data is very misleading. It's like saying Toyota has more crashes than Rolls-Royce. Well, of course they do, as they make more cars in a couple of hours than Rolls does in a year.

    And from 1998 to 2004, Airbus suffered 10 fatal crashes, a lot for the few Airbus planes then flying.

    Source for my data:

    Lastly, be aware that all these accident numbers are inflated, because authorities count everything-planes involved in hijackings (lots of them in the old days), planes shot down by missiles, accident caused by air traffic controllers, etc. and different authorities count different ways. My 'favorite' is a Continental Boeing 757 accident. A elderly passenger was loaded into the aircraft on a wheelchair. The passenger stood up so the flight attendant could stow the wheelchair. While the attendant was stowing the chair, the passenger (likely suffering from dementia?) proceeded to walk across the plane and out the open catering door, falling to his death. Hardly the fault of the plane, but it is counted by some authorities as a fatal accident!


    See : for a table of data for various aircraft.

    But again, you have to read these with a LARGE grain of salt-The Boeing 757 is arguably the safest airliner flying. Yes, it has had several fatal accidents, but either none appear to be the fault of the plane (the Brgenair and Aeroperu flights were probably caused by blocked pitot tubes, again giving bad airspeed data). See Wikipedia and other sources for more details.

    June 8, 2009 at 1:57 pm |
  17. Scott Currier

    Hydraulics have caused their share of problems too. Whether it be fly by wire or hydraulics there has to be redundancy with diversity. Don't run the cables next to each other and call it redundancy. That's not redundancy.

    I think it was Aeroperu that lost a Boeing 757 due to the pitot covered with tape by a maintenance person.

    My thought was "This is your captain speaking, we are experiencing some interference, does anyone have a GPS on board either operating or not?"

    Get the Gps and check your groundspeed at least.

    The poor guys call the airport and ask them what their speed is but the idiots are using secondary radar and are parotting back the same faulty readings that the plane is sending them. Primary radar, hello, ever hear of it? If a plane needs to verify it's location, altitude, and speed, don't use secondary radar. Use primary and give as much information as you can.

    At any rate, will be interesting to see what they find.

    June 8, 2009 at 1:07 pm |
  18. Chih Lai, Ph.D.

    I worked on airplane collision avoidance system (ADS-B) before. As I know some transponders on airplanes broadcast its information (location, speed, etc.) and receive the similar information from nearby (>100 miles) airplanes.

    Some transponders record this information on their data cards. Since few pilots from nearby airplanes reported seeing flash in the sky around the time flight 447 disappeared, we suspect that we may find some information about 447's status if we check data cards of transponders from those airplanes.

    Just some thought.

    June 8, 2009 at 1:06 pm |
  19. Ricardo

    I agree with Phil. Be patient. Most of the information I read, shows technical inaccuracy, typically wrote by people who don't know nothing about airplanes or flying them, sucking every word spoken by those who should be quiet and are just trying to keep French image clean (pilots, company and builder).

    June 8, 2009 at 12:58 pm |
  20. Aok

    You are correct. Could not agree more. Bill and Mike T should know that the 777 and 787 are fly by wire and were needlessly equipped with yokes . What matters is the problems are corrected.

    June 8, 2009 at 12:52 pm |
  21. Chi

    Geez can't people have an opinion on what type of aircraft they prefer? It doesn't make someone an idiot! Personally, I prefer Boeing to support American companies. MADE IN THE USA!!!

    June 8, 2009 at 12:49 pm |
  22. Ken

    Just out curiosity, does anyone know if the plane sends it's location along with the 24 automatic error messages sent to the maintenance facility?

    June 8, 2009 at 12:43 pm |
  23. Tim

    That A380 picture still has a nice joystick. It looks like they used the Microsoft Sidewinder. Complete with fire and missle launch buttons.
    Looks more like a complex computer game than a plane cockpit.

    June 8, 2009 at 12:41 pm |
  24. Gothic Skier

    I'd like to offer some input into this dialogue. In the region where the aircraft went down the prevaling winds are the trade winds. The northeast trades are especially strong this year supressing the Inter Tropical Convergence Zone southward by several degrees of latitiude. Over a period of several days, this could have caused debris to move to the southwest from where it came down.
    Just a thought regarding the weather and other nearby aircraft not being seriously affected. On 8 Aug. 1966 a Braniff Airlines BAC 111 broke apart in the air and crashed near Falls City Nebraska on a flight from Kansas City to Omaha as it flew into a line of thunderstorms. Four minutes earlier, another Braniff flight departing Omaha reported only moderate turburlence at worst as it flew into the same line. As a footnote, British Air Corp. hired Dr. Ted Fujita (of the tornado F scale) to assist in the crash investigation and his work demonstrated the hazards posed by severe updrafts and downdrafts common in thunderstorms.

    June 8, 2009 at 12:32 pm |
  25. Bernard J

    Remember the Airbus that snapped its tail in freak turbulences and crashed on N.Y shortly after the Sep. 11 horror, can it be the same scenario for AF447?? I hope the final result of the investigation will give a clear answer.

    June 8, 2009 at 12:24 pm |
  26. Horacio

    Four minutes of automatic transmission with the maintenance facility, and pilots missed the due report with Senegal? Were they alive, or had some device taken their lives, and producing a hole in the cockpit thus loosing pressurization?

    As I understand, the first two bodies found were complete, it seems they were not wearing the seat belts, which would most likely rule out bad weather.

    Autopilot controls the aircraft, but pilot controls it, in case of a malfunction, a flip of a switch disconnects it. Different radings of air speeed due too pitot tube problems?, Capt Cox rightly states there are ways to determine the correct
    one. Besides, all GPS units, even the cheapest ones, give you exact ground speed, too bad we didn't have them in my days.

    Let's hope the black boxes can give us enough information to come to a satisfactory conclusion.Meanwhile, may they all rest in peace.

    June 8, 2009 at 12:19 pm |
  27. Christos

    There have been far more Boeing accidents than Airbus, so these ridiculous comments by Bill and Mike T have no basis. These are just personal opinions and nothing more.

    The point is flying is much more safe than people's everyday activities, especially driving.

    June 8, 2009 at 12:19 pm |
  28. stuart

    let me tell you a story- im a surgeon. in 1980 when i was training, a patient came in with head trauma. we placed a device called a "bolt", a sensor which measured intracranial pressure- if the pressure went up, it indicated bleeding or swelling. the monitor showed normal readings, but the patient deteriorated right before our eyes, and it was entirely consistent with an increase in the very pressure the "bolt" was supposed to measure. the patient died with normal readings. at autopsy, the "bolt" was in normal position, but a little clot had formed, which interfered with the sensor- hence the normal readings. almost thirty years later, i remember this as if it happened yesterday. i have never relied on sensors or technology as a result. if it does turn out that the pitot tubes malfunctioned because of ice or debris, its an eerie parallel. you just cant beat human input and judgement. the major advantages of fly by wire are decreased crew requirements, less mechanical linkages, and lighter, hence more fuel efficient planes. as for boeing, god knows what the 787 dreamliner will be like- last i heard it was at least six month late for launch- wonder if its got anything to do with the fly by wire-

    June 8, 2009 at 12:17 pm |
  29. VBuren

    To Frank,

    "Fly by wire" means that the flight controls are managed by electronic, i.e., "wired" connections, not by hydraulic systems as in older generations of aircraft. Although computers are now incorporated in more and more critical and key flight management systems and functions, the human element is (or should) always be aware of and actively monitoring these systems to intercede and override the computer if anomalies present themselves.

    June 8, 2009 at 12:12 pm |
  30. Kaz

    On another message board many retired and current airline pilots are discussing this accident and I find their comments quite interesting. One in particular opined that fly-by-wire systems are great for military aircraft like the F16, F18 and F22, all aircraft that a pilot can punch out of if the computer ultimately does something funky. Not so great on a passenger plane without 300 ejection seats.

    June 8, 2009 at 12:07 pm |
  31. Dave S.

    I agree that there is certainly enough redundancy in the on-board airspeed (pitot-static) system to resolve a discrepancy if one is suspected. The key is recognizing that there is a discrepancy, as John Cox explained, and then appropriately acting on it (following the cross-reference procedure).

    Several have suggested that GPS or radar-calculated speed could be used as a primary reference. This doesn't work for one very important reason: ground speed and airspeed can be radically different, especially at cruise altitude when windspeeds can easily blow at well over 50 MPH. Airspeed is what the pilot (or automated pilot) is interested in with respect to flying; groundspeed is only important to getting to the destination on time.

    I agree with John Cox; it's really tough to speculate what happened when so little is known. I hope we can find the black boxes and figure this one out, but it's a long shot.

    June 8, 2009 at 12:07 pm |
  32. jim

    Forgive me if I'm wrong, but I vaguely recall a story within the last several months where pilots (whether from a Union or from a specific airline) either voted against (or condemned) plans to send cockpit voice recorder data and other data in real-time via telemetry due to concerns that airlines would not use this info. for safety purposes, but rather for monitoring conversations/actions of pilots. Kind of a "big-brother-is-watching" fear...

    June 8, 2009 at 12:06 pm |
  33. Michael Seufer

    Hi Frank,

    If you look closely at the photo, the flight 'yokes' are mounted on either side of the pilot and co-pilot's seats. Some military fighter jets have this configuration as well.

    The Airbus does indeed have sophisticated computer systems however.

    To others regarding the flight data recorders – the part of the ocean where the plane purportedly crashed, is part of the mid-Atlantic ridge. This is a very, very deep and nasty area of the ocean floor.

    There is a long, long underwater mountain range filled with crevasses and rough, en-even terrain. It's so huge, it stretches from the South Atlantic all the way up to Greenland. The ocean floor is pulling apart along this region, hot magma is coming out in various areas. There are a lot of earth processes and seismic activity going on down there. It's very beautiful, but very nasty at the same time. This is a tough, tough area to find something as small as a flight data recorder. Even with our current technology.

    To illustrate, the US Navy only has one deep sea manned submersible, called "Alvin", that's active right now and it can only go to a maximum depth of 14,764 feet. We're looking at 20-25,000 feet in some areas down there. This requires remote-operated underwater vehicles to get to that depth.

    If flight data recorders are found in an area like that, it will be a very, very impressive thing.

    June 8, 2009 at 12:05 pm |
  34. Greg

    As others have pointed out, until the flight data recorder ("black box
    ") is hopefully located, most of the accident analysis will be speculation at best.

    Somewhat related, it should be noted flight data recorders (FDR's) or "black boxes" are actually painted with heat-resistant bright orange paint so they may be easier spotted in an accident. It would be a bit counterproductive to paint them black!

    "Black box" is a bit of a misnomer that seems to persist, along with others terms used like "air pocket."

    June 8, 2009 at 12:04 pm |
  35. pete adams

    if it aint boing i aint going?????????i have heard this slogan for a long time and i beleive in it,there are still a veriety of airlines that fly boeing,there is no reason to fly these planes made by airbus,anyone can build a fancy plane and market it to be state of the art,but it takes a real company to keep one in the air

    June 8, 2009 at 12:03 pm |
  36. jgoodfrey

    Plastic aircraft (major parts are composites) like the Airsus A300 series have weakpoints such as the rudder assembly; it even has composite lug nuts (based on A300 crash over NY 8 years ago when the pilot stressed it). And there's no way to detect fractures as can be done with aluminum. Severe turbulence may have caused at least the rudder to break off. These problems plus fly-by-wire computerized decisions, which may be impossible for the pilot to quickly react to, may be the cause.

    June 8, 2009 at 12:03 pm |
  37. Aviator

    Please, read here :

    Or try "Norbert Jacquet" with Google.

    June 8, 2009 at 12:01 pm |
  38. Pat C

    GPS measures speed over the ground which often isn't all that helpful for determining airspeed. Inertial platforms also measure groundspeed. When indicated airspeeds are suspect, probably the most useful device is an angle of attack indicator; several Boeings I flew had them). In place of that attitude reference will keep the speed in a safe range.

    June 8, 2009 at 12:00 pm |
  39. Bruce K.

    Whatever the cause-This was horrible and only time will hopefully heal the wounds of all the losses-BUT the need for a fix to prevent in the future is necessary regardless if a Boeing-or-Airbus aircraft... Flying is still the best mode of travel and I will continue and let the Lord guide me through my travels. My condolences to all the families/friends of the deceased...

    June 8, 2009 at 11:58 am |
  40. Jason

    For those against fly by wire, read up on past accidents and you will see that most of them are through pilot error or pilots misinterpreting the computer and taking the wrong action against the computer.

    As with all things in life, nothing is 100%, but accidents due to computer error are significantly less than human error.

    June 8, 2009 at 11:57 am |
  41. Dean in Phoenix

    TWA 800...lets engineer routed electrical wires THROUGH the fuel tank rather than around it. Implausible. Then no one noticed the wires going through the fuel tank upon presumably multiple inspections. I don't know what happened...but this is a unbelievable scenario. There would have to be more to it.

    June 8, 2009 at 11:57 am |
  42. Moe

    Maybe the plane git hit with a piece of falling space junk..its bound to happen one of these days.... so much crap in space

    June 8, 2009 at 11:56 am |
  43. DP in DC

    MikeT must not remember the tail problems with 737's that caused a few crashes back in the 80's(remember Piedmont?).

    June 8, 2009 at 11:55 am |
  44. jim

    Would you put your life in the hands of this computer, in a lightning storm?

    June 8, 2009 at 11:52 am |
  45. brad

    June 8th, 2009 11:43 am ET

    Rick- `Isn’t it about time we do away with the black boxes and have each airline transmit in-flight data in real time? The airplane sent out 24 messages to Airbus maintenance. If it had sent out the flight data we would know exactly what happened.`


    June 8, 2009 at 11:48 am |
  46. Ken from Dallas

    If my Magellan portable GPS can tell me when I'm exceeding the speed limit on a freeway, why can't similar technology be used as a redundant airspeed indicator for automated, fly-by-wire systems?

    June 8, 2009 at 11:47 am |
  47. Rick

    Isn't it about time we do away with the black boxes and have each airline transmit in-flight data in real time? The airplane sent out 24 messages to Airbus maintenance. If it had sent out the flight data we would know exactly what happened.

    June 8, 2009 at 11:43 am |
  48. Kyle


    I would suggest doing some research on the safety of these aircraft before drawing such conclusions. I am a pilot and have flowing various Boeing (yoke style) and Airbus (joystick/sidestick style) aircraft and have to say that the joysticks handle better and there is no risk of stressing the control surfaces of the aircraft by means of pilot input since they are fly-by-wire. This is a likely cause of the AA A310 crash in New York in November 2001 (the A300 and A310 use analog yokes instead of sidesticks). These fly-by-wire aircraft are very safe and while things will go wrong, air travel is still the safest means of transportation. By the way, the A380 still uses sidestick controls, the keyboard is not used to control the actual movements of the aircraft.

    June 8, 2009 at 11:42 am |
  49. Bob

    GPS and other data would give ground speed, but is there any other way, besides Pitot tubes, to measure air speed? Overspeed or stalling only happens when you are outside the envelope on air speed, it is irrelevant of ground speed.

    June 8, 2009 at 11:40 am |
  50. dutchbound

    Everyone is the expert now-a-days. More than likely all the other posts given have no or less experience than the flight crew flying this fateful flight.

    If it ain't boeing it ain't going, what a stupid remark, period.

    Wait for fly by wireless.

    June 8, 2009 at 11:36 am |
  51. Serge Crespy, Collingwood, Ontario

    Earth's emission of methane hydrates could (theoretically) collect into a large pocket, in the upper atmosphere. Such a volatile gas would easily ignite during a lightning storm; or, by a jet engine's flame. Could the Air-France jetliner have flown into such a pocket? ... A far-fetched theory?

    June 8, 2009 at 11:36 am |
  52. Art in Chicago

    I agree with Mike in KC. Both Boeing and Airbus manufacture great products and I have flown AF many times before and find their service and personnel to be outstanding.

    We simply have to wait this one out and hope we can piece the information together to get to a cause. Frankly with the miniscule cost of computer storage, it seems the "black box" function could also be done remotely through telemetry. Not sure what the cost would be to fit such an option on a plane, but it would eliminate the event dictating the recovery of crucial information.

    June 8, 2009 at 11:35 am |
  53. Tony

    Tragic event. I would be very amazed if the black box would be recovered. With ocean current and no exact location of the crash site the search is beyond a needle in a hay stack.

    I truly believe this aircraft bursted in the air...

    June 8, 2009 at 11:34 am |
  54. Bill M.

    Mike T's quote is not from an idiot.

    In aviation it is a widely know quote.
    If it ain't Boeing, I ain't goin'!!'"
    Another popular one is "Scairbus".

    Pilot's have issues with these fly by wire planes.

    June 8, 2009 at 11:32 am |
  55. Jason

    The computers themselves are not to blame. It's the poor implementation of those automated systems that should be blamed. Airbus has had a history of poor implementation. For example, the computer should probably not react on its own when the speed changes more rapidly than is normal or possible. When multiple sources of information are available, they should be used. For example, airspeed and Mach can be measured directly or computed by combinations of GPS, INS, outside air temp/pressure, etc. A robust system would smartly combine all of these measurements, be able to recognize faults in any one system contributing to the estimation, ignore that one system, and then report it to the pilot. Not saying erroneous speed was the cause, just giving an example of a bad and good implemetation of a system.

    June 8, 2009 at 11:32 am |
  56. 3000 Hour Pilot

    David in KC...

    TWA 800? Really? That was closed a long time ago – fuel vapors in the ullage of the main fuel tank were ignited by chafed wiring that some idiot engineer had routed through the fuel tank rather than around it. Don't say you're a conspiracy theorist. Something as simple as a nitrogen inerting system would have prevented the TWA 800 accident.

    Nicole – most aircraft I've flown the GPS speed was not immediately apparent; you have to have the correct page pulled up on your FMS. Beyond that, we at least only used it for SA, the primary means were always the airspeed indicators and the relatively failsafe pitot-static system. Once you start getting into ADCs rather than direct reading gauages, however, you introduce more margin for error. I've never flown an Airbus, so I'm not familiar with their systems or design philosophy directly. However, in the Learjet I used to fly it was pretty common to have a discrepency between the pilot and copilot airspeed indicators due to the 2 seperate systems and some installation error. We would use the higher or lower one for the "correct" speed whichever was more conservative (i.e., the lower speed for approaches, the higher for airspeed limits.)

    There's a logical chain of events that seems to be developing as consensus, but I hate to see the media jumping to conclusions. Can we have some more reasoned reporting like this until we get actual information about what happened?

    June 8, 2009 at 11:32 am |
  57. Christine Bishop

    Wy ?? can ,t the find (invent ) a flooding device for black boxes on he planes ,if we can go to the moon we should be able to make our safety recording devices easear to find in water

    June 8, 2009 at 11:32 am |
  58. Bayram

    Did not a Turkish Airlines Boeing plane crash as a result of similar problem (faulty speed sensor) because they were , flying/trying to land at a wrong speed?

    June 8, 2009 at 11:32 am |
  59. Max

    I agree with David in KC that MikeT is a BRAINLESS IDIOT.
    Betw 1996-2005 Airbus was involved around the world in only 3 deadly accidents but Boeing in 32 !!!
    So -silly MikyT, get your facts straight ...You just jealous that the Europeans are manufacturing more modern, more high – tech aircrafts

    June 8, 2009 at 11:31 am |
  60. John

    My father (a pilot) was on an aibus 320 back in the 90's. He was up in the cockpit with the pilot (this is before 9/11) in the jump seat when it began a porpoising type of oscillation which was quite uncomfortable and the crew were unable discover the reason why but apparently it was caused by some computer type of malfunction. It went on for quite a while and then eventually stopped. The pilots were mystified.

    While pilot error is the cause of a certain percentage of accidents I think that over time we will discover that computer error will also be cause for many accidents. When it comes to flying no matter how many safety features are added to aircraft, over time, Murphy's law prevails. The air accident record shows this.

    June 8, 2009 at 11:31 am |
  61. YanG

    With current technological developments, and the fact that we have a vehicle approaching the edge of our galaxy (Gallileo, Cassini) we do not have an automated beacon which would pin point us to the location of the plane/black box. Have the airplane industry become so coplacenet that we do not plan for these type of situations.

    June 8, 2009 at 11:30 am |
  62. aaron

    fly-by-wire is actually found in advanced systems. The old way is archaic and complex since it introduces a hydraulic system into a network of flight computers , flight directors, autopilots, and inertial sensors which all communicate via wires not hydraulics. Fly-by-wire actually assists the pilots a great deal as well and is found in all the newer Boeing models including the military's own F-22 Raptors. So please give it a rest on the fly-by-wire, Captain Sully did just fine with it.

    June 8, 2009 at 11:28 am |
  63. gjokkel

    The latest two Boeing models (777 and 787) use the exact same ""fly-by-wire" design as Airbus does and it has a very good reason. The fly-by-wire system protects the aircraft from inappropriate pilot input (which in fact crashed SEVERAL aircraft in the near past: FedEx MD11 anyone?).
    It the current case multitudes of parallel disturbing factors had lead to the tragic accident: night, ocean (no visual reference point), extreme weather, speed sensor failure.
    These conditions are enough to disorient even the best pilot as it happened to two B757s in the past few years.

    June 8, 2009 at 11:28 am |
  64. Matt B

    The Boeing 777 is a fly-by-wire aircraft as well, along with the new 787.

    June 8, 2009 at 11:24 am |
  65. scott

    Hey Frank i see the yokes on the outside arm of both seats !!!

    June 8, 2009 at 11:24 am |
  66. Thomas

    In the 380 cockpit to the side of the 'keyboard' you will notice a 'stick'. That is what takes the place of the traditional yoke, not the keyboard.

    June 8, 2009 at 11:23 am |
  67. Vlad

    What really frustrates me as a former telecom engineer is that in the 21st century we are still relying on antique black boxes. Why the flight computers can not transmit GPS coordinates or even full telemetry together with the error codes via satellite on any error event is beyond my understanding. The technology to do so (and at reasonable cost) existed decades ago. I think this is a borderline criminal complacency on behalf of aircraft manufacturers not to implement such system.

    June 8, 2009 at 11:22 am |
  68. Myles

    Mike – Those are some really interesting points that are definitely worth considering ... keep in mind though that the Boeing 717 as well as the 777 and soon-to-be 787 are all fly-by-wire aircraft with similar parameters built into their automatic flight control systems. The 717 in particular is set up in such a way that the autopilot will not allow you to stall the aircraft if you fly it outside of a set zone (as interpreted by flight control computer). Speaking of Boeing, a Turkish Airlines 737 in fact crashed a few months ago because the aircraft falsely interpreted its altitude above the ground and prematurely pulled the throttles back to idle–this is not to say that the 737 is a poorly designed aircraft (quite the contrary) but it does illustrate the point that the automation issue is not unique to Airbus.

    If we assume for a second that the speculation (which is all that it is at this point) is true regarding the Airbus's technology being the cause of the crash, it is still important to realize that in the grand scheme of things, automation and the advent of computers on jetliners have contributed to a pretty impressive safety record that is ever improving, despite what happened here. This would be why Boeing has now adopted fly-by-wire technology as well.

    June 8, 2009 at 11:22 am |
  69. Tom Wittmann

    There are several inconsistencies regarding the Air France and Airbus info which needed to be addressed:

    1) First it was said that the location of the aircraft when issuing the automatic update was not known. Now, they say that the debris and corpses were found 60 miles from this location. So what is true?

    2) If the location was known and after 6 days the debris was only 60 miles away, why this debris field not found immediately?

    3) If the Pitot problem has any relevance, it could have a provoked an out of control spin. But this would not, at least during the first phase, impede the pilots to communicate and much less, affect initially the electric system

    4) The fact that other airliners have not reported any concerning weather conditions is of course very strange. Does a independent confirmation (other than by Air France) of the broadcast where the Pilot issued this report?

    It seems that the French are suspecting that the accident could be originated by a extremely grave problem at Air France and/or Airbus, and therefore are not informing what they do not like. If so, they will go to the steps trying to find the Black Box, hoping this will not happen!

    June 8, 2009 at 11:19 am |
  70. Joel

    I think Airbus may be hurting if they don't find out the real cause of the crash and the clock is ticking. They are so many theories that all could be right, and many of them pointing at technical problems with the plane. If the don't find the flight data recorders and determine the cause and if there is a problem, fix it. People may be more reluctant to fly in an Airbus. Airlines may be more reluctant to purchase Airbus. Not knowing is the worst.

    June 8, 2009 at 11:19 am |
  71. Mike Smoth

    I can’t believe that in this day and age they are still having instruments problems to figure out the air speed. I also can’t believe it is based on variable air pressure gages when they could use radar beacons as well as satellite positioning to calculate the speed.

    June 8, 2009 at 11:18 am |
  72. nicole price

    What I don't understand is all the talk of mechanical airpeed indicators being the potential issue.
    Aren't there multiple pitot static tubes located on the fuselage?
    Aren't there multiple GPS antennas for navigation and secondary reference used on these aircraft as well? (yes).
    I find it hard to believe an accomplished pilot would ignore his GPS speed.

    June 8, 2009 at 11:17 am |
  73. David in KC

    MikeT, you are in fact an idiot. Do you remember TWA 800? I thought the 747 was built by Boeing??? We STILL dont know what happened to TW 800 and it will probably be another few weeks before we even know what happened to this AF 330.

    I suggest next time you post something so infamatory you check your facts first. Airbus and Boeing build excellent planes and I will gladly step aboard either one until something proves different.

    June 8, 2009 at 11:17 am |
  74. Jack

    The crash of flight 447 while cruising at altitude is to say, at the least, unusual. Due to the wide area of the crash site it would appear the plane was traveling at a high rate of speed and not a a slower rate which would have caused the plane to stall. If the plane broke up at 30,000 ft I suspect the passengers would not have suffered due the altitude where the air is very thin and cold. Nonetheless, the loss of life is terrible and for the sake of all the passengers and their families, we really need to spare no expense to discover what caused this aircraft to fail.

    June 8, 2009 at 11:10 am |
  75. Roman Robles

    I really appreciate it when someone who is knowledgeable on a topic is permitted to offer rational, reasoned, non-alarmist answers. John Cox was impressive in this interview.

    June 8, 2009 at 11:09 am |
  76. MikeT

    All of the Airbus aircraft are fly-by-wire aircraft whereby the computer interprates the pilot's inputs through the sidestick, rudder pedals and throttles into movements of the flight controls. Pilots don't acutally "fly" Airbuses, the computers do. Also, unique to Airbus, all Airbus aircraft have a ALAW function that prevents any Airbus from entering a stall. When the flight computers detect a stall condition (whether the autopilot is ENGAGED OR NOT), the Airbus will attempt to recover by pointing the nose down to increase airspeed.

    If for some reason the pitot system malfunctioned and showed an erroneous low airspeed indication, and subsequent erroneous stall, the computers would would increase throttle and point the nose down and the pilots would have to try to override this function.

    The problem lies if the aircraft were already cruising at a normal speed of .80 mach and the flight computers thought that the aircraft were stalling, its recovery actions would overspeed the aircraft which in turn would eventually cause the catastrophic failure of the wings and other parts of the aircraft as the aircaft approached and exceeded Vne (never exceed speed).

    Since other pilots did report seeing a bright flash in the skies, it might be a possibility that an overspeed indiation caused failure of the wings which contains fuel, causing fuel to leak and ignite and eventually causing an explosion.

    What's the old saying? If it ain't Boeing, I ain't GOING!

    June 8, 2009 at 10:59 am |
  77. Phil Leonelli

    This is all speculation. Wait until we find the black boxes.

    June 8, 2009 at 10:59 am |
  78. Frank

    It seems the "fly-by-wire" age of "flying computers" might be at least partially to blame – what will always be needed are aircraft that "fly by pilots (a/k/a humans)", first...

    In fact, the cockpit view of the latest Airbus aircraft, the double-decker A380, shows the primary piece of equipment not to be the traditional yoke – but a computer keyboard:

    June 8, 2009 at 10:58 am |