it makes sense to me that the pilot who said "I did not do it" actually did do it without realizing it, was supposed to be putting the landing gear up when he committed a muscle memory mistake. it happened around the time the landing gear should be up, and this explanation matches what was said in the cockpit, and the fact that the landing gear wasn't retracted. I think this idea was even floated initially by the youtube pilot/analysts I watch but dismissed as unlikely.
The landing gear lever is rather prominently featured in the 787 in a panel central to the cockpit layout so that either pilot can easily reach it. For decades and across many manufacturers, the landing gear lever has traditionally featured a knob that deliberately resembles an airplane wheel. It's very hard to mistake it for anything else. It's actuated by simply moving it up or down.
The fuel control switches are behind the throttle stalks above the handles to release the engine fire suppression agents. These switches are markedly smaller and have guards on each side protecting them from accidental manipulation. You need to reach behind and twirl your fingers around a bit to reach them. Actuating these switches requires pulling the knob up sufficiently to clear a stop lock before then rotating down. There are two switches that were activated in sequence and in short order.
The pilot monitoring is responsible for raising the gear in response to the pilot flyings' instruction.
I would find it very difficult to believe this was a muscle memory mistake. At the very least, I would want to more evidence supporting such a proposition.
This idea strikes me as even more unlikely than someone shifting their moving vehicle into reverse while intending to activate the window wipers.
Or even crazier, a manual shift on the steering column. Nothing weirder than pushing down the clutch and then changing the gear with your hand on a knob off to the side of the steering wheel.
I think the aircraft being familiar makes it worse: if you're used to going through a certain motion to do a thing, it may be one of things your brain can do without really thinking about it much, which is where the danger comes in.
I've engaged my wipers when meaning to shift gears before, in my truck which has a steering column shifter. After driving the truck for years. I have ADHD and I very often let my brain go on autopilot for things I do every day, and sometimes it just does the wrong thing. It doesn't matter how complicated or "intentional" the task has to be: my brain will memorize it to the point that it can execute it on its own without me consciously thinking about it.
I think it's totally plausible it was a muscle memory thing, if the at-fault pilot's brain works anything like mine.
(Side note: I actually took some flying lessons, including going through all of ground school, and realized that my brain is just not cut out for flying. I am the type of person to "cowboy" things if I feel like they're not worth doing, and flying is an activity where the tiniest missed checklist item can result in death, so I realized I have a statistically high likelihood of crashing due to some boneheaded mistake, and stopped taking lessons.)
There is no temporal nor spatial adjacency to the switches. The switches are equivalent to the ignition on your car, you operate it in the beginning and end of your trip, and there is nothing during the trip that will involve manipulating this switch.
There’s been at least 2 times I’ve turned my ignition switch while driving. (Luckily it was into the “on” position instead of off.)
Everyone in this thread thinking “these actions are temporally and physically distinct and therefore impossible for anyone to confuse” isn’t really thinking about the problem the right way. It’s not that I’m actually confusing two actions. It’s that I’m accidentally allowing my brain to perform one action when I meant to let it perform another action. “Allowing” is an important word here, because it illustrates that my brain is capable of doing this on its own without me thinking about it, and often will do it on its own, if I let it.
I firmly believe this is not the case. Putting more obstacles between me and the thing I’m trying to do, just trains my brain at rote performance of the task. It’s still “autopilot” as far as my brain is concerned.
This could be anything: starting a car, taking off a medicine cap, typing my password, clicking around cookie warnings. If I have to do it repeatedly, my brain will be able to perform it subconsciously, and I will do it without realizing it.
For fuel cutoff switches, it doesn’t matter that there two of them in a row. If you cut off both of them every single time, and you fly every day, your brain is gonna automate that task.
But the 787 doesn’t have an easily confused layout like that. The landing gear lever and fuel cut off switches are not two stalks on the yoke. Aircraft cockpits are deliberately designed in such a way that important things have differently shaped actuators that feel different from each other. Precisely so that you are not accidentally flipping the wrong switch by accident.
Actually the birth of Human Factors was related to this...
Alphonse Chapanis, a psychologist working at the Army Air Force Aero Medical Lab in 1942, investigated the issue and discovered a design flaw. He observed that the controls for the flaps and the landing gear in the B-17 cockpit were nearly identical and located close to each other.
This is exactly how the investigations are NOT conducted. You don't find the evidence that confirms your theory and call it a day when the pieces sorta fit together. You look solely at the evidence and listen to what they tell you leaving aside what you think could have happened.
One of the nice things about finally having the preliminary report is I get to stop hearing all of the same assumptions/theories/YouTuber said/"a guy I know got a leaked report"/etc in water cooler talk at work because the preliminary report solidly disproved all of them so far. If anyone even had and stuck with an idea matching this report it wouldn't have stood out in the conversations anyways.
The collection of comments on this post remind me it'll just be a brand new set of random guesses until the final report is released. Or worse - the final report reaches no further conclusions and it just has to fade out of interest naturally over time.
It's human nature to want to guess at possible explanations for things that are unusual and unexpected.
If hearing those guesses annoys you, nobody is forcing you to read through comments on a thread of people making them! (I hope - sorry if you are being forced after all.)
> If hearing those guesses annoys you, nobody is forcing you to read through comments on a thread of people making them!
It’d be nice if we could only read insightful comments, and unread the wacko comments, but we can’t. This discussion has actually provided a lot of useful comments from people who seem to know what they’re talking about, but also a lot of really wild speculation.
Idle speculation is far from the only thing you won't find me supporting just because it's human nature. Thankfully, HN comment threads tend to include a lot more than just that kind of discussion, which is why I read them. Indeed there are lots of great details I didn't glean or fully understand in the report covered in the comments.
That doesn't mean I will always agree with the comments (or that everyone will always agree with mine) and that's okay. It'd be a very limited value discussion if we could only ever comment when we agree. It seems exceedingly unlikely any of this has something to do with users being forced to be here though.
I... don't believe that's the case? Though happy to be proven wrong. Unless perhaps you mean a moderator of a different forum, though that wouldn't really be relevant to their reading a thread of HN comments on a subject that annoys them.
Not me! I'm not even sure I'd be the right type or the right energy level to do it if the opportunity even came up. Hats off to the team here for the dedication.
Double engine failure was confirmed, not disproven. RAT deployment was confirmed, not disproved. Pilot error, confirmed, not disproven. Preliminary and final aviation reports are mostly guesses.
I don't think it's fair to say that pilot error is confirmed yet. It seems like a reasonable hypothesis, but what if the electronics glitched out and acted as though the cutoff switches had been flipped (the first time), even if they hadn't? All of the currently-disclosed facts still line up with that scenario IMO.
Even as a matter of safety/investigation hygiene, "pilot error" should be the conclusion of last resort, arrived at after months of poring over data, and because nothing else seems viable.
If we decided to pin all aviation incidents on pilot errors, we wouldn't even have invented checklists (what do you mean you forgot, try harder the next time).
"Natural" pilot errors lead to lessons that can be incorporated into design/best practices. That does not seem to be the case given current understanding: no flaw in any switch design seems apparent, and it does not sound like something you could do by accident.
So "pilot error" is not the "cracking the case"-grade conclusion it is being made out to be, it is an act of investigative resignation. In the days following the crash, allegations of mixing up flaps and landing gear were floated, and they all turned out to be wrong. This is not even accounting for the fact that the pilots are not around to plead their case, and basic human dignity requires us to defend their case until evidence clearly points a certain way
This is the difference between the official investigation and casual online comments for which it's OK to consider it pilot error once there is sufficient indication for that.
> I get to stop hearing all of the same assumptions/theories/YouTuber said/"a guy I know got a leaked report"/etc in water cooler talk
This was a really disappointing incident for aviation YouTube - I unsubscribed from at least three different channels because of their clickbait videos and speculation.
> There is no possible way to confuse these two actions.
This is obviously an overstatement. Any two regularly performed actions can be confused. Sometimes (when tired or distracted) I've walked into my bathroom intending to shave, but mistakenly brushed my teeth and left. My toothbrush and razor are not similar in function or placement.
That's just your brain associating the bathroom with the act of brushing your teeth, and therefore doing it automatically upon the trigger of entering the bathroom. It bears no resemblance to the accidental activation of a completely different button.
The other poster's correction: "it’s like brushing your teeth with razor" is apt. Touching the fuel cutoff switches is not part of any procedure remotely relevant to the takeoff, so there's no trigger present that would prompt the automatic behavior.
Good analogy. Things I do every day in front of the mirror, but I occasionally attempt to squeeze some soap on my toothbrush. Or I have to brush my teeth and I find my beard foamed up. Or I walk out of the shower after only rinsing myself with water.
Not a bathroom one, but the number of times I've tried to pay for public transport with my work/office fob is mental. Generally happens on days where I'm feeling sharper than average but also consumed with problem solving
If someone confused their steering wheel for the brake you'd probably be surprised - there are indeed errors that are essentially impossible for a competent person to make by mistake. No idea about the plane controls, though.
Even in modern "fly by wire" cars the steering wheel and brake pedal have an immediate effect. They are essentially directly connect to their respective control mechanisms. As far as I understand both of the plane controls on question just trigger sequences that are carried out automatically. So it's more like firing off the wrong backup script than scratching the wrong armpit.
The only two production cars on sale where the steering wheel is mechanically decoupled from the wheels are the cybertruck and a variant of the Lexus RX.
Essentially impossible is not the same as impossible. We already know that an improbably sequence of events took place because a plane crashed which is highly unusual.
Technically an overstatement but not by much. Correctly restated, its highly unlikely these actions were confusing pilots. It's as if you mistook flushing your toilet twice when instead you wanted to turn on the lights in your bathroom.
I don't agree with the "twice". A frequently performed manipulation like the fuel cutoff (usually performed after landing) collapses down to a single intention that is carried out by muscle memory, not two consciously selected actions.
Not really, though. They're both (retracting the gear, and cutting off fuel) just toggle switches, as far as your brain's conscious mechanisms go. Doing them both on every flight dulls the part of your brain that cares about how they feel different to perform.
(I'm not strongly arguing against the murder scenario, just against the idea that it's impossible for it to be the confusion scenario.)
I meant philosophical toggle switches, not physical ones. The gear can go between down and up. The fuel can go between run and cutoff. Given enough practice, the brain takes care of the physical actions that manipulate those philosophical toggles without conscious thought about performing them.
Genuinely curious - could heavy marijuana use cause confusion between landing gear and fuel cutoff? Or some other drugs? (Wondering if they screen pilots for alcohol before they board an aircraft.)
The other day I was eating dinner while chatting with my partner. I finished eating and needed to pee and throw away the fast food container. I walked straight to the bathroom, raised the toilet lid and threw the fast food container right into the toilet.
As I get older, I do some similar stuff, way more than past, even it is just once per month. And I guess way more when sugar is high than not. Don't know your age or medical profile and I am not a doctor, just keep an eye.
Because there's no difference in actions needed to do so. A similar mistake is throwing away a useful item while holding onto a piece of trash. The action is the same, it's just the item in question that's different.
This is not what happened here at all. The actions needed to activate the fuel cutoff switches are not similar to any other action a pilot would want to make during takeoff.
How about actual switch covers (and switches that are not located right in the same area as stuff you are using routinely) instead of a glorified detent? Though I suspect this would also succumb to muscle memory
What about up on the overhead panel where the other engine start controls are?
Or (at the cost of complexity) you could interlock with the throttle lever so that you can't flip the cutoff if the lever isn't at idle
Also the fire suppression system is a different activation (covered pull handles I think)
> How about actual switch covers (and switches that are not located right in the same area as stuff you are using routinely) instead of a glorified detent? Though I suspect this would also succumb to muscle memory
The switches are already pretty distinct - but that only reduces failures, it can never eliminate them entirely.
> Or (at the cost of complexity) you could interlock with the throttle lever so that you can't flip the cutoff if the lever isn't at idle
More complexity also means more failure modes. You don't want it to be impossible to shut down the engines due to a broken throttle sensor.
And a gun that doesn't let you point it at your face. And a knife that doesn't let you cut yourself. And a car that doesn't let you accelerate into a static object. And...
Hey my car won’t let me accelerate into a static object. It’s so good it will even slam on the brakes when driving 5mph in a parking garage because it thinks parked cars are oncoming traffic.
"Sent into terrain in seconds by flipping a switch" is both too inaccurate and feels too cursory to take as impetus for serious design criticism, especially when the extensive preliminary report explicitly does not recommend any design changes with the current information.
Hilarious how Hacker News routinely bashes software managers who don’t understand a problem space and give vague and impossible goals. But somehow “just don’t let an aircraft fly itself into the ground” is a reasonable statement.
If we are going to pretend that Hacker News is a single person who should have consistent opinions on different topics then we might as well pretend that we are in a room together.
Those are caused by operating the same lever in a slightly different manner. Not comparable to two completely differently designed levers placed far apart.
Same goes for accidental acceleration instead of braking. Two of the same kind of lever right next to each other.
Accidental acceleration while intending to turn on the wipers would be a fitting example, I don't think that happens though.
Think of the action as a stored function. Maybe they’ve always recalled the function as part of a certain list. It can be a case where the lists get confused rather than the modality of input (lever etc)
Landing gear controls are nothing like the fuel shutoffs. And they are in completely different locations. Landing gear controls are in front of the throttle, fuel shutoffs are aft of the throttles.
Is that "nothing like" though? You are saying they are in different places, ok, but are they similar in other ways? Are both controls the same shape? size? colour? texture?
Respectfully, it's not up to other people to disprove your toy theory. The question you're asking here can very easily be answered with a quick Google search.
The short answer is that they are _very_ different controls, that looks and operate in a completely different way, located in a different place, and it's completely unrealistic to think a pilot could have mistaken one for the other.
Different controls with different t shapes, operated in different ways, of different number, different size, and very different positions. One is down almost on the floor, and well rearward, the other is at stomach height and well forward.
The fuel cutoffs also require pulling the control out and over a guard.
Is there a video feed of the cockpit inside the black box?
If not there should be one as even my simple home wifi camera can record hours of hd video on the small sd card. And If there is, wouldn't that help to instantly identify such things?
No neither black box stores video. One stores audio on flash memory and the other stores flight details, sensors etc.
I don’t think video is a bad idea. I assume there is a reason why it wasn’t done. Data wise black boxes actually store very little data (maybe a 100mbs), I don’t know if that is due to how old they are, or the requirements of withstanding extremes.
This isn’t true. This was a 787. It does not use a separate recorder for voice and data (CVR, FDR).
(Most media outlets also got this wrong and were slow to make corrections. )
Rather, it uses a EAFR (Enhanced airborne flight recorder) which basically combines the functions. They’re also more advanced than older systems and can record for longer. The 787 has two of them - the forward one has its own power supply too.
There should be video as well, but I’m not sure what was recovered. Not necessarily part of the flight data recording, but there are other video systems.
That's really interesting. From reading air crash reports there's a lot of times I've seen."Nothing is known about the last 30 seconds because the damage broke the connection to the flight recorders in the tail"
In the US, the NTSB has been recommending it for over 20 years. The pilot unions have been blocking it, due to privacy and other things.
I'm not in aviation. But my between-the-lines straightforward reading is that unions see it as something with downsides (legal liability) but not much upside. It could be that there are a million tiny regulations that are known by everyone to be nonsensical, perhaps contradictory or just not in line with reality and it's basically impossible to be impeccably perfect if HD high fps video observation is done on them 24/7. Think about your own job and your boss's job or your home renovation work etc.
Theoretically they could say, ok, but the footage can only be used in case the plane crashes or something serious happens. Can't use it to detect minor deviations in the tiniest details. But we know that once the camera is there, there will be a push to scrutinize it all the time for everything. Next time there will be AI monitoring systems that check for alertness. Next time it will be checking for "psychological issues". Next time they will record and store it all and then when something happens, they will in hindsight point out some moment and sue the airline for not detecting that psychological cue and ban the pilot. It's a mess. If there's no footage, there's no such mess.
The truth is, you can't bring down the danger from human factors to absolute zero. It's exceedingly rare to have sabotage. In every human interaction, this can happen. The answer cannot be 24/7 full-blown totalitarian surveillance state on everyone. You'd have to prove that the danger from pilot is bigger than from any other occupation group. Should we also put bodycam on all medical doctors and record all surgeries and all interactions? It would help with malpractice cases. How about all teachers in school? To prevent child abuse. Etc. Etc.
Regulation is always in balance and in context of evidence possibilities and jurisprudence "reasonableness". If the interpretation is always to the letter and there is perfect surveillance, you need to adjust the rules to be actually realistic. If observation is hard and courts use common sense, rules can be more strict and stupid because "it looks good on paper".
You also have to think about potential abuses of footage. It would be an avenue for aircraft manufacturers, airlines, FAA, etc to push more blame on the pilots, because their side becomes more provable but the manufacturing side is not as much. You could then mandate camera video evidence for every maintenance task like with door plugs.
I wonder how the introduction of police body cam footage changed regulations of how police has to act. Along the lines of "hm, stuff on this footage is technically illegal but is clearly necessary, let's update the rules".
If you work in a job where the lives of hundreds could be ended in seconds due to an error or intentional action then there is no excuse to not have critical control surfaces recorded at all times. Non-commercial/private flights/flight instructors and trainees have cameras, trains have camera, stores have cameras, casinos have cameras, buses have cameras, workers who work for ride hailing services have cameras as do millions of other people who just drive.
Hopefully other countries will start deploying recording systems or start forcing manufacturers of planes to have these integrated into cockpits.
> The answer cannot be 24/7 full-blown totalitarian surveillance state on everyone.
Surveillance is actually pretty common in many high-risk environments. And piloting is very much not just any other job but an exceedingly rare situation where the lives of hundreds of people are in the hands of only two people without anyone else being able to do anything to influence the outcome.
That pilot unions don't want surveillance is to be expected (the union is there to act in the pilots interest) but ultimately it isn't just up to them.
> Should we also put bodycam on all medical doctors and record all surgeries and all interactions?
Yes. We are finally starting to do so for police. These are all situations where an individual or very small team has direct control over the life of others who can't defend themselves.
In fact, you could add some AI to it, even, as an embedded system with a decent GPU can be bought for under $2000. It could help prevent issues from happening in the first place. Of course airgapped from the actual control system. But an AI can be very helpful in detecting and diagnosing problems.
If you shut off the engines a couple of dozen meters above ground shouldn’t every alarm be blaring or there should be some sort of additional lever you have to pull way out of the way to enable shutting off the engine that close to the ground.
If you read through the boeing procedures, if an engine fails just after take off you delay cutting throttle or hitting the cutoff until you have positive climb and pass a certain altitude. Specifically because a mistake here would be so incredibly catastrophic. The following number of steps and verbal cross checks for then shutting down the engine are quite daunting. Not something applicable here, but still interesting to learn about
That’s absolutely applicable here.
It means that an engine cutoff shouldn’t be allowed at all during certain parts of flight. It’s not crazy to think that a design fix would be to prevent those engagements during certain parts of takeoff (a certain window). It’s fly by wire anyway so it could presumably be done programmatically.
MCAS was basically made to prevent user input that would send the plane into a dangerous angle. The computer overrode the inputs. So there’s precedent for something like it.
I just remember that it happened commonly enough to US pilots. American pilots always recovered quickly enough that it didn't make the news before the fatal crashes.
which has detailed information, but I don't trust the NYT. The article also erroneously reports that the MCAS trim rate is twice the speed of manual trim. The trim rate for both is 0.27 degrees per second.
MCAS autonomously adjusts trim downward. The trim switches override MCAS, but when released, MCAS can resume trimming down again. The trim adjustments don't "override" the pilot's elevator inputs (MCAS has no direct control over the elevators), but they can make the controls so heavy that it's impossible to pull up.
If MCAS is running the trim, the thumb switches override it.
MCAS affects the stabilizer, the thumb switches affect the stabilizer, the cutoff switch affects the stabilizer.
The elevators are controlled by the control column and the autopilot.
> The trim switches override MCAS, but when released, MCAS can resume trimming down again.
That is correct. That is why the procedure is to return the trim to normal with the thumb switches, then turn off the trim system. That's it. That's all there is to it.
> but they can make the controls so heavy that it's impossible to pull up.
Almost right - the trim has more authority than the elevators. The trim's ability to travel far is to provide great ability to get out of trouble. I don't really know what factors the aerodynamics guys used to calculate the max travel required. I do know there is a travel limiter on it (as I worked on that, too!) which reduces the max travel at higher speeds, because otherwise it can rip the tail section off, which is a big no-no.
There are sooo many constraints on the design of an airplane I sometimes wonder how anyone manages to make one that works at all. The Wright Bros calculated that their machine would fly, and it did, barely. Their contemporaries did seat of the pants design, which is why they failed.
Thank you. I've had commercial pilots email me telling me I was correct and to keep on the good fight :-)
The Wrights did a lot more than that to be successful. Their innovations were:
1. using a wind tunnel to correctly get lift and drag coefficients for various wing sections (as you wrote)
2. first aviation engine (high power/weight ratio) (as you wrote)
3. first propeller theory, enabling 90% efficiency (other aviation propellers were 50% efficient)
4. first 3-axis control system
5. identified and solved adverse yaw problem
6. first research and development program, where problems were identified in advance, and a machine was developed to solve each problem, then the solutions were put together to make the 1903 Wright Flyer
7. kept meticulous notes on all their work and preserved the evidence of their success, such as photographs and notebooks. Exacting replicas have been built, and their flight characteristics match the Wright's results
Would it matter in this case since you would crash either ways. I’m talking about protection in a very specific situation where you make it harder to shut off both engines when you’re very close to the ground.
All I’m saying is in those situations it should involve another toggle or pedal that needs to be pushed to cut off the engines so it’s outside the realm of muscle memory.
I don't think the theory is that the muscle memory sequences resemble each other.
Instead, it's that because muscle memory allows you to do things without thinking about it, you can get mixed up about which action you meant to perform and go through the whole process without realizing it.
They're pilots, they do hundreds of stops each year. In case of domestic pilots, even thousands. And with years of experience, switching off fuel control switches is basically muscle memory at this time now.
That's a ridiculous analogy. The pilots aren't sitting in front of a uniform set of keys that they need to press in a specific order with a specific timing.
The mistake equivalent to what the pilot supposedly did would be if the pianist accidentally stuck a finger up his nose instead of playing the notes or something.
Quite, but the point is that even after doing something correctly a thousand times, someone can make a mistake that seems unbelievable.
The cutoff switches are operated every flight so the muscle memory is there, ready to be triggered at the wrong time.
All we know is that something went wrong in the pilot's head in at least a single moment that caused him to perform a ground action during takeoff.
Depressive murder-suicide is one possible explanation. Altered mental state is another: insomnia, illness, drugs/medications could all explain an extreme brain fart. Perhaps he just had food poisoning? It's India after all.
I keep reading "muscle memory" but the theory that one pilot shut down the engines instead of performing another action has nothing to do with muscle memory.
Muscle memory allows you to perform both actions effectively but doesn't make you confuse them. Especially when the corresponding sequence of callouts and actions is practiced and repeated over and over.
All of us have muscle memory for activating the left blinker in our car and pulling the handbrake, but has anyone pulled the handbrake when they wanted to signal left?
i have several passwords i type all the time. sometimes i get them confused and type the wrong one to the wrong prompt. i type them by muscle memory, but i also think about them while typing and i think thoughts like "time to reach up and to the left on the keyboard for this password". I couldn't tell you the letter i'm trying to type, i just know to do that.
not all my passwords are up and to the left, some are down and to the right, but when i type the wrong one into the wrong place, i type it accurately, i'm just not supposed to be typing it.
"time to do that thing i've practiced, reach to the left". shuts two engines off by muscle memory.
> "time to do that thing i've practiced, reach to the left". shuts two engines off by muscle memory.
If that were true, pilots would perform arbitrary motions all the time. Same with car drivers.
Typing something on a keyboard, especially when it's always in the same context, is always essentially the same physical action. The context of a password prompt is the same, the letters on the keyboard feel the same and are right next to each other.
Not comparable to pressing two very different buttons placed far apart, in a context when you'd never ever reach for them.
One time I almost drove to work instead of the shopping mall, because driving to work was basically subconscious. But my car doesn’t have safety features designed to stop me doing that. And I never did it twice in a row.
that makes it less likely, not impossible, we're trying to match against the data we have. I think distracted muscle memory is more likely than suicide and sounding innocent while lying about it
Not possible. Two fuel cutoffs. Two engines. Two intentional acts in rapid succession. Plane would have survived one cutoff. It is what it appears. Captain crashed the plane.
