Lion Air crash: Jakarta Boeing 737 'had prior instrument error'

[Marco]

Lets say they used the thumb sticks to pull trim back and then took out the checklist because they assume they have corrected the problem for now and have some breathing room, so now they are distracted and then the system hits trim again after the time out. It's just such a fucking mess.

IMO they need a big red button to turn off autopilot, but really, actually, completely, not partially.

[sokoloff]

It's a two-crew airplane for a reason. Pilot Flying (PF) does the memory items and calls for the checklist. Pilot Not Flying (PNF) pulls out the Quick Reference Handbook (QRH) and reads the checklist items in challenge form. PF responds with the response. The whole time PF keeps flying the airplane and monitoring its performance, including attitude and airspeed.

You don't have two people going heads-down into the books. (It happens, most notably, tragically, and avoidable in something in Eastern 401, where 101 people died because a cockpit crew of three mismanaged a burned out landing gear position indicator bulb. Since then, CRM classes have taught the division of duties to avoid future "two researchers, zero pilots" situations.)

[daddylonglegs]

If there is disagreement, it obviously doesn't matter which one is wrong to detect there is a malfunction. This case would warrant the disconnection of any automatic system that could rely on or even be influenced by those sensors, and issue a warning. First thing that happens is usually the disconnection of the autopilot, and I'm pretty sure that happens when there is a faulty pitot, at least on Airbus planes. This is the first thing that happened on the AF447 for instance, AFAIK. Of course, this could be different in this case. Maybe.

In the AF447 scenario, the computer decided it did not have reliable air data, so it not only disconnected the autopilot, it dropped into alternate law. Basically, the computer just punted and said "your airplane!" This might seem like reasonable behavior, but one might imagine that handing a degraded airplane to a pilot at the very worst moment is not a satisfactory failure mode, and in the AF447 scenario the outcome was obviously bad.

  ISTM that Airbus have made fairly clear and consistent design choices. The flight envelope protection will not allow the pilots to stall or overspeed the aircraft unless the sensors start playing up (in a detectable way) or the pilots deliberately disable this system (which takes a single action). Not everyone likes the choices but they are reasonable. I agree with you that it is difficult to make changes that are unambiguously better.

  Judging by the AD, the accretion of changes to the 737 has produced a system that is:
A - Badly documented (supposedly a deliberate choice to save time and money during training).
B - Does not completely handover control when the sensors disagree (even warning the pilots that the sensors are playing up is an optional extra! [1]).
C - Requires two actions, both disabling the autopilot and cutting out the trimming system.

  I think Boeing have been caught out by the compromises involve in updating a very old design[2], marketing pressures[3] and - possibly - by their own propaganda regarding how traditional the flight controls on a Boeing are. My predictions are that:
A - Training will now heavily cover the auto trim systems.
B - The sensor disagreement warning will not be optional in practice (everyone will buy it)
C - Boeing may have to change the MCAS system by adding a cut-out when the sensors disagree (should have done that from the start IMHO).

  I'm surprised at what is hidden from pilots both in the manuals and by making the sensor warnings optional. If pilots get a stall warning without an explicit indication that the sensors disagree[4] they may well believe it. Particularly if they have already been surprised by changes in the aircraft's motion (driven by the automated system's intervention) and so lost their 'feel' for whether the aircraft might plausibly be close to stalling. A curious and sad inversion of the AF447 crash if so.

  All of this is my ill-informed opinion. Both Airbii and Boeings are flown very safely by sensible airlines. The Lion Air crash may well be something completely different, etc.

[1] Maybe Boeing could consider in-appflight purchases in this case? "We've noticed that your subscription does not include information helpful to your continued survival. Would you like to upgrade? (T&C apply. Always make sure you have the permission of the credit card holder. Offer not valid if you're going to sue us.)"

[2] First flight in 1967, 51 years ago and 64 years after the Wright brothers. Dates from wonkypedia, maths from my sligtly tired brain.

[3] Honestly, I get infuriated by manufacturers making an oscilloscope's actual capabilities an optional (i.e. expensive) upgrade and now Boeing's marketing department gets in on the game? And then nobbles the flight manual/ conversion course to boot?

[4] Do any of the checklists (memory or written) call for comparing the instruments?

[sokoloff]

C - Requires two actions, both disabling the autopilot and cutting out the trimming system.

This is standard. My 3600-pound airplane has the same basic response for an AP or electric trim malfunction. My previous 2800-pound airplane didn't have electric trim installed, but if it did, it would have been the same.

[4] Do any of the checklists (memory or written) call for comparing the instruments?

This is the fundamental skill of all instrument-rated pilots and a primary focus of instrument flight instruction.

[raptor1956]

I don't think there's much in this video that is different from what has been discussed on this thread so far. For now, I'm sticking with my "prediction" that the pilots and Boeing will end up with some of the responsibility. There's nothing here to imply that the aircraft could not be flown. It's definitely not okay that Boeing put in a new flight control system and made no mention of it in the  true-up training for the MAX. On the other hand, if that system started moving trim according to rules unknown to them, that would seem like the very definition of runaway trim to pilots who didn't know about MCAS, and if they followed the procedure for that, they would have been fine.

Umm, no...

If this runaway trim happened at 5000 feet and they spent even a few seconds fighting the stick by the time they did turn things off the plane would be at 3000 feet and 30 degrees down -- you won't recover from that even with the trim system turned off.

The idea that pilots will automatically and with zero delay jump to a seldom used (or never used) emergency procedure is delusional -- that's not how humans respond.  They will attempt to pull back on the stick and then, maybe, realize they need to turn the trim off but by then...


Brian

[floobydust]

You have to manually crank the stabilizers back to center, after shutting off power to the motors.
Can you imagine how hard that is, it's two BLDC with screw-jack and you don't have much time.

[djacobow]

If this runaway trim happened at 5000 feet and they spent even a few seconds fighting the stick by the time they did turn things off the plane would be at 300 feet and 30 degrees down -- you won't recover from that even with the trim system turned off.

The idea that pilots will automatically and with zero delay jump to a seldom used (or never used) emergency procedure is delusional -- that's not how humans respond.  They will attempt to pull back on the stick and then, maybe, realize they need to turn the trim off but by then...

I think neither of us are airline pilots, so we'll just have to disagree about this and find out what's in the reports when they come out. My sense as a pilot (not an airline pilot) is that this is EXACTLY the kind of thing that is -- or should be -- on a pilot's mind, particularly when climbing out at low altitude. As I've said before, these trim changes, though fast, are not instantaneous. They can be felt. Furthermore, a sudden trim change doesn't result in an instantaneous attitude change. (The plane has a moment of inertia and it takes time for the nose to rotate down). And an instantaneous attitude change does not result in an instantaneous altitude change. Some of these delays work against you as you try to arrest a descent just as they work for you at the start of the descent, but I'm nearly certain that proficient, attentive pilots can recovery from a runaway trim situation that starts at 5000 feet.

Yes, if you get into a 6000 fpm minute descent then you are probably screwed, but you'd really have to let the plane get into a seriously unusual attitude to get that kind of descent.

It may seem as if a few armchair pilots are being perhaps a bit too harsh on the pilots in this scenario -- and I'm willing to concede that maybe we are. But you have to understand that pilots are always supposed to be thinking of what can go wrong in each phase of flight and what they'd do about it. It's the essence of piloting. Flying a working airplane is not particularly hard. Managing a "normal" flight takes up a surprisingly small fraction of overall flight training. Most flight training is about handling abnormal situations, and dealing with those is the crux of a pilot's job. Cross checking instruments (including checking them against the huge non-artificial horizon available in VMC) and avoid and recovering from unusual attitudes are definitely in this category.

[rx8pilot]

You have to manually crank the stabilizers back to center, after shutting off power to the motors.
Can you imagine how hard that is, it's two BLDC with screw-jack and you don't have much time.

Pretty sure it is somewhere between 'very hard' and 'impossible' once the system has gone full nose heavy. In a manual operation, it is a LOT of turns and needs considerable effort to move. All of that has to be accomplished in the middle of a very confusing and scary emergency event that is very rapidly getting worse. There is little doubt that both pilots knew it was a deadly situation early on which would make rational decision making that much harder.

The speculation will slowly give way to actual facts.....but it is looking like the pilots would need to be superhuman to have avoided the outcome.

[sokoloff]

Pretty sure it is somewhere between 'very hard' and 'impossible' once the system has gone full nose heavy. In a manual operation, it is a LOT of turns and needs considerable effort to move

This is true, but they struggled with the airplane for quite some number of minutes. It's during this time, before the stab trim runs full nose-down, that was the time to interrupt the power to the trim system.

They were dealt a crap hand to be sure. I'm much less sure that a majority of 737 pilots would fall victim to that crap hand. I think the CVR will be enlightening to understand what was going on in their heads as they fought the airplane. The FDR and wreckage can tell us a lot of the WHAT; the CVR will help with the WHY.

[raptor1956]

As is common in major tragedies the end result stems from more than one problem.  If the stab ran-away as a singular event I'd guess they would have responded as they should and they would still be alive.  However, going back over time to the 3 previous flights, which the pilots must have been aware of, the thing of concern appears to have been BOTH a pitot/static problem and an AOA problem.  So, on the day of the crash there first indications may well have been in line with what the previous pilots reported and they were very likely working that problem when the stab trim system threw them for a loop -- a stab trim system they did not fully understand because Boeing neglected to provide that information.

When you're working one or two known problems and a third rears its head you can bet there was a period of seconds at least where they tried to correlate the new problem with the existing ones.  At 5000 feet they don't have that time.  And, as other suggest, once the system threw the plane into a nose down attitude with, potentially, substantial down trim, they would have to manually rotate the trim wheels which, again, at 5000 feet they would have very little time.

This has the makings of a horror story!


Brian

[raptor1956]

The video from the Mentor pilot about the MCAS system provides some interesting info about what may have necessitated the MCAS system.  The 737MAX uses larger more fuel efficient engines but do to the low ground clearance of the basic 737 design they need to mount the engine further forward and up to gain some clearance.  Doing that would have changed the balance of the AC meaning that all things being equal it would need the stabilizer to be trimmed up more than previous 737 AC for a given flight condition.  It's not clear to me if there were any other changes to compensate for the changed balance or if the stabilizer was changed in some way, but his mention of the larger engine and the need to alter the mounting location certainly raises some questions in my mind.  In a perfect world such an alteration would necessitate other changes to return the balance -- perhaps moving some gear further aft, perhaps moving some of the fuel tanks further aft, perhaps altering the stabilizer to account for what would otherwise be a nose heavy tendency. 

We are likely to see an interim report within a couple weeks but the FAA and Boeing probably know already exactly what happened.  Hopefully the interim report provides some closure on this.  OTH, if the engine change started a chain of other changes that lead to this Boeing could have bigger problems than the legal one they are no doubt going to face.


Brian

[djacobow]

I think folks here are ignoring some key facts. First, that the emergency bulletin from Boeing does nothing but explain MCAS behavior and amplify existing procedures. Second, that the -MAX aircraft are still in service.

This strongly suggests that the aircraft are believed safe by Boeing and the FAA and that at least for the interim, the only extra training pilots need is a reminder not to let the plane get way out of trim.

I find the points that the pilots were screwed once the plane was way out of trim to be mostly irrelevant. Yes, there may not be time to recover by manually re-trimming a plane that is totally nose heavy at low altitude. That's why you don't let the aircraft into that condition. There are many conditions that you can get into that are not recoverable. You just don't let the aircraft get there.

[raptor1956]

I think folks here are ignoring some key facts. First, that the emergency bulletin from Boeing does nothing but explain MCAS behavior and amplify existing procedures. Second, that the -MAX aircraft are still in service.

This strongly suggests that the aircraft are believed safe by Boeing and the FAA and that at least for the interim, the only extra training pilots need is a reminder not to let the plane get way out of trim.

I find the points that the pilots were screwed once the plane was way out of trim to be mostly irrelevant. Yes, there may not be time to recover by manually re-trimming a plane that is totally nose heavy at low altitude. That's why you don't let the aircraft into that condition. There are many conditions that you can get into that are not recoverable. You just don't let the aircraft get there.

You are presupposing that the pilots chose to remain at low altitude when in a normal climb-out they would have been much higher at 13 minutes.  The flight profile shows a lot of erratic up and down which I should point out is not typical of a normal climb-out. 

So, it looks to me that the plane was at a dangerously low altitude and when the final push happened, or whatever the final event was, they were too low to recover even if they had the control authority to do so which is highly suspect at this point.  The emergency bulletin from the FAA and the technical bulletin from Boeing would hardly be considered business as usual.


Brian

[Nusa]

Actually, such alerts probably are business as usual when pilots point out a safety/training deficiency.

Until we get more info, it's just speculation that it had anything to do with the crash.

[raptor1956]

Actually, such alerts probably are business as usual when pilots point out a safety/training deficiency.

Until we get more info, it's just speculation that it had anything to do with the crash.

The FAA seldom uses the word "Emergency" to issue nothing out of the ordinary messages. 

We should see the interim report in a few days and have a better idea of what is fact and what has been speculation.


Brian

[Nusa]

Actually, such alerts probably are business as usual when pilots point out a safety/training deficiency.

Until we get more info, it's just speculation that it had anything to do with the crash.

The FAA seldom uses the word "Emergency" to issue nothing out of the ordinary messages. 

We should see the interim report in a few days and have a better idea of what is fact and what has been speculation.


Brian

You missed my point. Issuing Airworthiness Directives is simply one of the things the FAA does. That includes emergency ones, of which you'll find at least a handful every year, sometimes more. They're just doing their job. It's not unusual when deciding what qualifies as an emergency is normal procedure, so it's business as usual.

[djacobow]

I think folks here are ignoring some key facts. First, that the emergency bulletin from Boeing does nothing but explain MCAS behavior and amplify existing procedures. Second, that the -MAX aircraft are still in service.

This strongly suggests that the aircraft are believed safe by Boeing and the FAA and that at least for the interim, the only extra training pilots need is a reminder not to let the plane get way out of trim.

I find the points that the pilots were screwed once the plane was way out of trim to be mostly irrelevant. Yes, there may not be time to recover by manually re-trimming a plane that is totally nose heavy at low altitude. That's why you don't let the aircraft into that condition. There are many conditions that you can get into that are not recoverable. You just don't let the aircraft get there.

You are presupposing that the pilots chose to remain at low altitude when in a normal climb-out they would have been much higher at 13 minutes.  The flight profile shows a lot of erratic up and down which I should point out is not typical of a normal climb-out. 

So, it looks to me that the plane was at a dangerously low altitude and when the final push happened, or whatever the final event was, they were too low to recover even if they had the control authority to do so which is highly suspect at this point.  The emergency bulletin from the FAA and the technical bulletin from Boeing would hardly be considered business as usual.


Brian

I'm trying to reason based on clearly stated suppositions -- none of which require that the pilots "chose to remain at low altitude."

- It is a fact that the aircraft are still in service.
- It is arguably a fact that Boeing and the FAA are extremely risk-averse. Certainly, the repercussions for another crash due to the same or a related problem would be severe.

This strongly suggests that the model is safely flyable, including under all but the most improbably failure circumstances. It also suggests that a "design fault" is not the proximate cause of this accident.

If that is true, then either:

 - the accident aircraft was "broke" in some very unlikely way
 OR
 - the operator of the accident aircraft conducted maintenance malpractice and sent an unairworthy aircraft into service
 OR
 - the pilots screwed up
 OR
 - some combination of the above


As for why the pilots maneuvered at low altitude, we simply do not know.

As others have stated before, accidents do have multiple causes, and I expect this one will, too.

[floobydust]

The root cause, the AOA or air speed sensor issue I don't think has been figured out. Another airline, Southwest had replaced a couple of AOA sensors on their 737 MAX 8's.

The problem looks like it was intermittent, on and off. Previous flights had the same problem, sensor was replaced etc.
Air Asia Crash Report : Cracked Solder Joint on Rudder Limit Circuit Board https://www.eevblog.com/forum/chat/air-asia-crash-report-cracked-solder-joint-on-rudder-limit-circuit-board/

How MCAS resolves a discrepancy when you only have two sensors that can (normally) read different, it's like the redundancy engineering is shit (=not there) and the pilot is expected to cover for it.

Boeing has a backlog of 4,783 orders for 737 MAX. for multi-billions $. I hope they don't put all this on the flight crew.

[sokoloff]

I hope they don't put all this on the flight crew.

I hope they put the blame on wherever the right place for the blame to go is.
If that's on the crew, so be it...

[raptor1956]

Actually, such alerts probably are business as usual when pilots point out a safety/training deficiency.

Until we get more info, it's just speculation that it had anything to do with the crash.

The FAA seldom uses the word "Emergency" to issue nothing out of the ordinary messages. 

We should see the interim report in a few days and have a better idea of what is fact and what has been speculation.


Brian

You missed my point. Issuing Airworthiness Directives is simply one of the things the FAA does. That includes emergency ones, of which you'll find at least a handful every year, sometimes more. They're just doing their job. It's not unusual when deciding what qualifies as an emergency is normal procedure, so it's business as usual.

Well of course the FAA issues emergency reports with some frequency, there are hundreds of different AC each with issues that rare there head from time to time.  FDR, like all US presidents, addresses Congress and on December 8th 1941 he did so again -- I would not describe his speech on that day as 'business as usual' even though numerous other presidents have made similar speeches.  The circumstances that preceded his speech made it something of an emergency.

When the FAA issues an emergency directive its because something time critical makes it an emergency so postponing the directive isn't acceptable.  When FDR declared war on Japan it wasn't business as usual even though doing so was the job of the president.

The terms 'business as usual' and 'emergency' or not synonymous -- they are in fact incongruous.


Brian

[Nusa]

Actually, such alerts probably are business as usual when pilots point out a safety/training deficiency.

Until we get more info, it's just speculation that it had anything to do with the crash.

The FAA seldom uses the word "Emergency" to issue nothing out of the ordinary messages. 

We should see the interim report in a few days and have a better idea of what is fact and what has been speculation.


Brian

You missed my point. Issuing Airworthiness Directives is simply one of the things the FAA does. That includes emergency ones, of which you'll find at least a handful every year, sometimes more. They're just doing their job. It's not unusual when deciding what qualifies as an emergency is normal procedure, so it's business as usual.

Well of course the FAA issues emergency reports with some frequency, there are hundreds of different AC each with issues that rare there head from time to time.  FDR, like all US presidents, addresses Congress and on December 8th 1941 he did so again -- I would not describe his speech on that day as 'business as usual' even though numerous other presidents have made similar speeches.  The circumstances that preceded his speech made it something of an emergency.

When the FAA issues an emergency directive its because something time critical makes it an emergency so postponing the directive isn't acceptable.  When FDR declared war on Japan it wasn't business as usual even though doing so was the job of the president.

The terms 'business as usual' and 'emergency' or not synonymous -- they are in fact incongruous.


Brian

Tell that to a 911 dispatch service, where every call is an emergency, by definition, until determined otherwise. They decide which calls are real emergencys many times an hour, and also the type of emergency....and it's still business as usual FOR THEM.

This FAA directive does not compare to an FDR declaration of war at all, which I do agree was a fairly unique event. It doesn't even describe anything wrong with an aircraft. All it demands are a few updates to training manuals in reaction to information put out by Boeing. I still call that business as usual FOR THE FAA. They determined the information was sufficiently urgent and reacted like they're supposed to.

[djacobow]

From https://www.nytimes.com/2018/11/27/world/asia/indonesia-lion-air-crash-.html

"""
The information from the flight data recorder, contained in a preliminary report prepared by Indonesian crash investigators and scheduled to be released Wednesday, documents a fatal tug-of-war between man and machine, with the plane’s nose forced dangerously downward over two dozen times during the 11-minute flight. The pilots managed to pull the nose back up over and over until finally losing control, leaving the plane, Lion Air Flight 610, to plummet into the ocean at 450 miles per hour, killing all 189 people on board.
"""

Eleven Minutes.

Two dozen attempts.

Sorry, there's no way these pilots did not screw up. Again, that's not to say that the undocumented MCAS system and potentially erratic behavior from potentially bad air data are not factors. But these pilots most definitely screwed up.

Also:

"""
Despite Boeing’s insistence that the proper procedures were in the handbook, also called the emergency checklist, pilots have said since the accident that Boeing had not been clear about one potentially vital difference between the system on the new 737s and the older models. In the older versions, pilots could help address the problem of the nose being forced down improperly — a situation known as “runaway stabilizer trim” — by pulling back on the control column in front of them, the pilots say.

Family members grieving after police handed over the remains of their relatives who had been aboard Lion Air Flight 610.CreditEd Wray/Getty Images
In the latest 737 generation, called the Max, that measure does not work, they said, citing information they have received since the crash. The pilots on Lion Air Flight 610 appear to have forcefully pulled back on their control columns to no avail, before the final dive, according to the information from the flight data recorder.

Capt. Dennis Tajer, spokesman for the American Airlines pilot union and a 737 pilot, said he could not comment on any aspect of the investigation. But, he said, “in the previous model of the 737, pulling back on the control column, Boeing says will stop a stabilizer runaway.”

Information provided to American Airlines from Boeing since the crash, Captain Tajer said, “specifically says that pulling back on the control column in the Max will not stop the runaway if M.C.A.S. is triggered. That is an important difference to know.”

Boeing said in its statement on Tuesday that the existing procedures covered the latest 737 model.
"""

This strongly hints at something I suggested earlier, that the written procedures, when followed, work, but that 737 pilots may have been accustomed to another procedure (pulling back on the control column) that was never the official book procedure, but which also worked on all 737's until the -MAX.

[floobydust]

Two dozen attempts -
It's an automatic system switching in and out, doing retries on its own, and possibly operating with bad noisy sensor data.

You have at least four oscillators. The MCAS software, the sensor(s), the pilot, the co-pilot.
The airplane or the pilot as the ultimate authority is yet another conflict.

Investigators are saying the plane was not airworthy. It should have been grounded as soon as problems surfaced on the previous flight.
One AOA sensor had been replaced, but still a 20 degree discrepancy.

Really need the CVR to figure out all that happened.

[Berni]

Yeah it seams to me like the pilots could have easily prevented the crash if they knew what to do, even if they spent a while to diagnose the problem.

But that's not to say they are completely at fault. The finger pointing should instead be focused more towards the training the pilots ware given. Clearly they should have known about such a significant system on the aircraft.

And this doesn't just mean hammering a bunch of step by step procedures into there head on what to do if something goes wrong. They should understand why each step of the procedure is on the list and how it interacts with the problem. Yes there need to be defined procedures for dealing with faults, but in the event that the procedure does not work (Such as a fault they did not foresee) or there are multiple interacting faults that render the procedure useless they shouldn't just be repeating the procedure over and over because that's all they know to do. They should instead start thinking about why the procedure did not help and try other things.

Tho in this case the pilots surely should have noticed the big noisy trim wheels spinning all the time. If the aircraft is pitching down and trim wheels are moving that should have been an instant association of it being a trim problem. You don't need to look trough a manual for a troubleshooting procedure if you can see where the problem is. Fix it quickly and then try to figure out why it happened so that you can make a decision if its safe to continue flying to the destination.

[Gyro]

Well it looks as if there is some progress in not making the poor pilots the scapegoats....

Indonesian investigators have said the Lion Air plane that crashed last month killing 189 people was not airworthy and should have been grounded.

https://www.bbc.co.uk/news/world-asia-46121127