Was Emirates’ 777 Crash Landing “An Accident Waiting To Happen?”

Filed Under: Emirates

Last Wednesday an Emirates 777 crash landed in Dubai. I haven’t really been following the play-by-play on this too much (well, aside from the painful-to-watch evacuation), though up until now I haven’t really heard any reputable theories as to what happened. In this instance the pilots are alive, there are lots of eyewitnesses, and presumably it’s easy for them to recover all the recorders, so I’m surprised by the lack of information so far.

Well, my friend Jim passed along a fascinating article from The Australian which is written by a pilot with over 45 years of flying experience, and who was a senior captain at Emirates for over 15 years. He has a pretty straightforward explanation/theory of what happened to EK521, which he says could happen to anyone. Not surprisingly, automation is to blame.

Based on what we know, the plane’s wheels touched the runway, and then the pilots attempted to perform a go around, which should be a pretty standard procedure. However, during the go around the plane fell to the runway.

In case you haven’t heard it, here’s the ATC audio from the incident:

So, what’s the pilot’s explanation of what happened? Basically when you initiate a go around, you simply push a “TOGA” switch (which stands for “Take Off, Go Around”), and then the plane automatically powers up its engines to sufficient power for the go around, while you make sure you maintain a positive rate of climb and raise the landing gear. It’s one less thing for pilots to worry about in a go around, so that they don’t manually have to push the throttles forward.


But according to this pilot, that also ended up being a problem, since the plane’s wheels had already touched the ground, which meant the “TOGA” command wasn’t activated as usual:

But in the Dubai case, because the wheels had touched the runway, the landing gear sensors told the autoflight system computers that the aircraft was landed. So when the pilot clicked TOGA, the computers — without him initially realising it — inhibited TOGA as part of their design protocols and refused to spool up the engines as the pilot commanded.

Imagine the situation. One pilot, exactly as he has been trained, clicks TOGA and concentrates momentarily on his pilot’s flying display (PFD) to raise the nose of the aircraft to the required go-around attitude — not realising his command for TOGA thrust has been ignored. The other pilot is concentrating on his PFD altimeter to confirm that the aircraft is climbing due to the aircraft momentum. Both suddenly realise the engines are still at idle, as they had been since the autothrottles retarded them at approximately 30 feet during the landing flare. There is a shock of realisation and frantic manual pushing of levers to override the autothrottle pressure.

But too late. The big engines take seconds to deliver the required thrust before and before that is achieved the aircraft sinks to the runway.

What remains to be seen with this theory is what caused the go around, especially after the plane’s wheels had already touched the ground. Not that a go around as such is unusual, but it’s certainly rarer to have one after the wheels already touched the runway. Apparently there was a sudden wind gust, which may be to blame.

Bottom line

I recently had the opportunity to fly a full motion Boeing simulator, and even did a go around. The process was indeed very easy — I pushed the “TOGA” button, pulled the nose up, and put the gear back up. It’s perfectly plausible to think how both pilots could have been busy and not realized that the auto-throttle hadn’t actually been activated. After all, pilots are taught to rely on their systems.

It’ll be some time before we get an official report of what happened, though this explanation from a former Emirates captain and pilot with 26,000+ flying hours sounds plausible. As pilots are increasingly taught to rely on automation, it presents a whole new set of dangers.

What do you make of this theory on EK521?

  1. Automation is increasingly an issue. Same for the Asiana crash at SFO a couple of years ago. In particular, from what information emerged, Asian pilots in particular apparently lack skills in hand-flying a plane, and combine this with the Asian cultural senior/junior reluctance to question issues, yes, an accident waiting to happen

  2. Interesting. What’s even more interesting is how and where you were able to fly a full-motion Boeing simulator! Please tell…

  3. Did you really ask “What do you make of this theory on EK521?” Not quite sure why travel blog readers would know. I certainly don’t.

  4. Boeing has Sims available to the public for a fee. I flew a sim in Singapore and it was an amazing experience. I think I paid around $200 for a 45 minute session and you can choose any airport in the world to land to.

  5. PPRUNE.org has had a lot of conjecture on this topic– the one big mystery is the “cycle time” of the gear, since it’s pretty clear gear weren’t down-and-locked when the TOGA failed miserably. Did they have enough time to fully pull the gear?
    There are interesting questions about the timeline– it’s far above our paygrade to figure out.

    As you correctly point out, this is one where they should figure out exactly what happened. All the pieces are there for a quick solution to the mystery.

  6. The pilots decided to go around well before the plane touch down according to the recording. Then how could the Air/Ground Sensing system prohibited the EEC to set the thrust to TO/GA?

  7. Boeing and Airbus bring full aspect sims to the major aviation shows. 50% of selling an aircraft is convincing the chief pilots of airlines to buy into them.

    @Tim – the TO/GA system requires time to react – in particular engile spool timing. If in that time, sensors in the landing gear sense significant forces associated with hitting the tarmac, they will indeed disable TO/GA and initiate auto-brakes and reverse thrust if set to auto. I am surprise the EK pilots did not recollect this since there was severe emphasis on this point provided by Boeing in a recent recertification course which I attended.

  8. I don’t see a failure of automation — the equipment did exactly as designed. Instead, we see a lack of training and understanding by the flight deck. They were never taught that once the wheels hit the ground, all autopilot systems on that aircraft are disabled. This is where curious pilots, like QF’s Richard de Crespigny excel — they’re not content with step-wise instructions. They want to know what’s behind the controls, what makes an aircraft work. In their pursuit of knowledge, they learn about these little nuances and quirks.

    Having an automated car, I can certainly appreciate the situation the pilots were in. My car’s equipped with just about everything Bosch Mobility Solutions can put in a car in terms of automated driving. Today I ended up in a near-zero visibility rain storm, the type where you can’t even see the front of your hood. The car handles zero-visibility fine, such as in fog. BUT when the radar’s signal gets attenuated so much by rain that it loses sight of the objects around it, the computer gives up and expects you to drive in full-manual…but that’s not what it tells you. Instead there’s a single beep and the screen displays “Radar Sensors Dirty.” Hopefully you notice the max speed dot & little steering wheel icon on the screen have disappeared. I was warned about it, so it’s not a surprise to me, but I can see it biting someone who isn’t paying attention.

  9. A Pilot without class 1 medical, so I am not going to speculate here. Going to wait for the facts, unlike the author of this blog.

  10. Ben, Lucky…whatever you prefer to be called. Please, Im begging you here. You are an awesome travel blogger, extremely well informed and the very best on boardingarea by far. Refreshing honest, personal, open opinions and above all informative. I could go on with the compliments, well meant and deserved. However….

    You are not a pilot, and certainly not a jet pilot, and beyond that not a heavy jet pilot. Please…. refrain from opinion. You know virtually nothing about flying an aircraft. Trust me. And if you think some sim time qualifies you to even pass comment it doesn’t. Im a Boeing and Airbus rated Captain, and I cannot even comment without a lot more information. I doubt you will even read this but hey, if you do, hopefully you will listen.

  11. If you are going to write an aviation article, you might want to get the basics down, prior to trying to sound like an expert. The PFD is the primary flight display, not Pilot Flying Display…amateur.

  12. One this is for sure…. With a Captain from the UAE, and First Officer from Australia – the blame for this incident will not fall on the Captain. Anyone who has ever worked in the Middle East already knows that the expat will be the fall guy!

  13. I dont know about TOGA inhibition on a 777 but what Pilots are taught is to rely on instruments and not on systems ! That is a tremendous difference. But all saying a monkey could fly a today’s automated aircraft is again proven to be wrong. What pilots urgently need are manual flying skills and that is what even legislation if not the operators (and Emirates unfortunately is an operator that tells its pilots not to touch a thing unless in emergency -but as we see then it’s too late- this is not their first automation issue, if it is one, see runway excursion in Munich a couple of months ago) have found to be important. On that finding new type rating courses had to be instated especially for new on the market aircraft. To all the flightsimulator pilots it is not that easy to fly, judge and land a plane as your PC or the fellow airline manager is telling you ! Look at the leftover 777 junk closely, they thought they could handle it !

  14. I don’t think that this was a failure of the aircraft’s sytems – the sytems reacted as they should have done. However, the increasing danger is whether pilots know how to use those automations correctly and to their advantage. The equipment can be relied on, as it did exactly what it was supposed to do. In this case, however, it was the pilot lack of understanding of the equipment which caused the crash. He/She should have known that when a wheel touches the ground the TOGA system can’t be used.

  15. Why are you surprised by the lack of information? The authorities have made it very clear that any public speculation or unauthorised release of information by individuals will be cracked down on immediately. Nobody is going to leak info when they know it will wind them in jail or worse.

    I spent this morning in a meeting of our local airport (just up the road from DXB) SAG/LRST where the main agenda item was the contingency planning feedback from this accident. I’ve been involved with multiple accident investigations in different parts of the world and I can honestly say that GCAA is actually being more transparent about this investigation TO THE RELEVANT STAKEHOLDERS than most other lead agencies are. The public is not considered a major stakeholder at this stage of the investigation, but the way it is progressing so far I have confidence that we will get a solid and accurate report eventually and that this will not be swept under the table.

  16. @sean: Exactly as it should be to avoid misinformation and speculation.

    To all who are complaining about Ben/Lucky with his speculation: this is not a journalism website. He is a blogger whose entire business is based on opinions and personal taste. Therefore, don’t read to much into articles such as this! He isn’t a journalist, but a blogger. There is a major difference.

  17. Pilots were having toga party. What a shock.


    General public may not be stake holders, but every EK ticket holder is a stake holder. They can reciprocate silence with non-spending.

  18. Thanks, Lucky – I appreciate your relaying The Australian’s article. Given the lack of information, what I wanted was at least some intelligent speculation. Given the writer’s experience, it’s certainly sounds an interesting explanation.

  19. It’s lucky’s blog and he can speculate however he likes, and I think it’s clear he isn’t offering an expert testimony. Even if he is dead wrong on this, at least he has created a forum where a lot of intelligent people can offer their own opinion and facilitate a constructive discourse about what happened here.

  20. Bmlynn

    In response to your question asking Ben how he gets the funds to buy miles. Just read some his past posts. He owns a company that provides services to people booking award trips for a fee. he has employees as well. Also he receives referral fees from credit card companies based on sign ups from blog readers. That is common for all blogs.

  21. I think the correct statement should have been “just like the Asiana crash pilot error was to blame.”

    With Asiana piolots did not monitor the air speed which would have alerted them to the fact that the automation was in idle as it was designed to do based on the flight scenario.

    With Emirates the TOGO was disabled after wheels down by design and if the pilots “forgot” that doesn’t make automation the cause of the crash. Pilot error is still the cause of the accident.

    Better pilot training about how the systems work in less frequently encountered flight scenarios is the only way to fix accidents caused by pilot error surrounding the use of automated systems.

    The part everyone is getting right is that there is a gap in understanding of how automated systems are designed to function in circumstances that pilots have very real reason to believe they will face. Airlines should either demand systems be redesigned or demand better comprehension of the existing automation.

    The day automated systems do not operate as designed and documented in flight training manuals is the day you can say that the automated system caused a crash.

  22. Of course Lucky can speculate as much as he wants. It’s a free country!

    However believe it or not, operating an aircraft that weighs ~700,000 lbs with complex systems in a three dimensional world with infinite variables of weather and more is not a piece of cake. It takes decades of required experience to even make the copilot seat, then possibly decades more to safely operate as commander. The human aspects of operating as a crew add more layers.

    Aviation is a very complex and involved science. Lay opinions and speculation are unlikely to hold water. As me commenting on heart surgery gone wrong. It would be ridiculous and arrogant.

    Here’s the thing also. Thousands of people read this blog daily and listen. A wrongly opinionated speculation by a passenger (yes he is a passenger that’s all) and one that want ant even onboard the flight in question can take hold with people.

    Thousands of people also work for Emirates worldwide and count on them for their livelihood.

    That’s just my opinion.

  23. To start with I’m a B777 pilot too.

    This is so not true. You don’t push TOGA and wait for auto throttle to do the Magic. When you are So closed to ground, what you do is a procedure called Balked landing. Followed by a go around. In which you manually advance thrust levers to vertical Position, arrest the rate of descent, begin to climb and then get the gear up and press toga so that auto throttle sets the maximum GA thrust.

    I wonder how this capt was flying for EK so many years or is it the EK training which never taught him this?

  24. @BoeingFlyboy – similar experience here as a 747 and 787 certified. Not that we really know what happened here but task fixation?

  25. @BoeingFlyboy

    Author must be a paid consultant building a theory against FO. In any case walls are closing in on this poor expat FO and every one will lend a hand to make it quick.

  26. What on earth is a “pilot’s flying display (PFD)?” A PFD is a primary flight display!

    @Lucky: I don’t take any “news” related posts on your blog seriously but the credibility of your source in this situation is extremely suspect at the least….

    I don’t fly any (or even certified to) aircraft, but I certainly own a dozen jets and even I know that much….

  27. Perhaps manual vs automated touch and goes can help to understand and interpret all manual and automated modes.

    Single ils and go arounds examinations are a must, but we all know that.

  28. Yes, as pilots we are taught to make best use of automation, and rely on it.
    But, you don’t select gear up before you are sure you have a positive rate of climb and thrust.
    A go around is being trained so often in the simulator that you should be able to feel/hear if you have go around thrust or not!

    No matter what the final conclusion will be, it’s not only the automation to blame.
    Will it as well be the pilots or the way of training??

  29. Doesn’t this case sound similar to the Habsheim Airshow crash with Air France Flight 296 in the late 80s?

    I think the belief was the flight system had engaged Alpha Mode Protection due to flying so low and to the ground with landing gear down at low engine speed that the engines didn’t respond to the crew’s throttle inputs as the flight system believed the plane was landing.

  30. My dad is a 30+ year pilot at a major U.S. airline. I had the opportunity to sit in on part of his recurrent training recently (classroom + sim), and the instructor spoke critically of the hands-OFF flying approach most foreign airlines promote (these airlines tell their pilots to engage autopilot at the earliest opportunity, as SOP).

    U.S.-based airlines, for all their soft product faults, teach HANDS-ON flying. This is important, as it promotes less dependency on automation (and faster response times in CRITICAL situations).

    Sure, there’s always going to be human error (this is the argument against hands-on flying). I’ll take a proficient cockpit crew over zombie pilots any day, though.

    Not exactly related to this TOGA scenario, but it’s important to understand the two schools of thought (automation vs hands-on flying).

  31. First of all: Let’s be very careful to jump to conclusions before we have the final report.
    That said, if Byron Bailey’s evaluation would be confirmed, even if partially, the conclusion should be:
    Sometimes, no matter how fascinating the new automated features and processes are, we should reinstate some good habits from the past. Example: On board of good old Swissair 121 (ATL – ZRH), I witnessed that they consistently had one hand on the throttles during take off and landing. And not just one of the crew, both PF (Pilot flying) and PNF. And even the flight engineer checked the throttle settings very frequently, especially during that phase. My friend witnessed the same procedure during another SR flight. It was a very good best practice for a simple direct feedback from the true throttle setting. And it would be a very quick and efficient fix to make this a standard work instruction…
    Not to forget: Good job cabin crew and crew trainers!
    Also, let’s not forget to pray for the family of the firefighter.

  32. Well done to the cabin crew for there work getting people off the plane and RIP the poor fireman that lost his life very sorry to here that

  33. In such case it would be preferable to do a manual Go Around during which you may not loose altitude if worry about the thrust and the pitch first.
    It was a rejected landing and Boeing shall include in QRH a detailed Rejected landing procedure as well as to correct the Go Around procedure.

  34. Sometimes the non-experts get it right too!
    I’ve been reading the comments and am amazed how many “experts” commented here that the TO/GA switch is not disabled on touchdown.
    The report from the GCAA clearly says that it is.
    Jared says that the recertification instructors keep stressing on that point. Maybe the pilot flying wasn’t paying attention, or he forgot the operating manual.
    BoeingFlyboy calls it right. Once they heard the “Long Landing” call and decided on a GA at such low altitude, the PF should have taken over manual controls.
    This seems to be a combination of three compounded errors by the PF and PNF.
    1. Even with a long landing, there was enough runway left to land the plane safely. There was no need to go around.
    2. Once the PF decided to go around after touchdown, he just pressed the TO/GA switch expecting the Flight Director to take over the thrust management. Neither the PF or the PNF seems to have had a hand on the throttle to manually over ride if necessary.
    3. It took them a full 12 seconds after the TO/GA switch press to realize the Flight Director hadn’t increased the thrust, by which time it was too late.

  35. We’re not taught to “rely on automation”, we are trained to monitor it and confirm our selections using feedback from the instruments (which we DO rely on) to confirm our instruction is being followed.

    the pilot commanded TOGA activation, but didn’t keep his hand on the throttle where it belongs to confirm they are advancing, he didn’t read out the “Flight mode Annunciations” as he’s required to do, otherwise he would have seen that it wasn’t doing what he asked. The pilot monitoring failed in his duty to do just that, and elected to make a radio call as a top priority, dividing the cockpit, and ignoring the basic flying principle of “Aviate, Navigate, THEN Communicate”

    both are to blame, the question just remains how their training failed, and whether there are records of poor performance that have been ignored. That data is yet to be released.

Leave a Reply

If you'd like to participate in the discussion, please adhere to our commenting guidelines. Your email address will not be published. Required fields are marked *