Ethiopian Airlines Crash; Another 737 Max

Nobody knows exactly what these guys were dealing with here. Probably a lot of confusing and conflicting information such as attitude and airspeed disagree, stick shaker, aural warnings, flashing lights and only a minute to figure it out.
But.. they're pilots, theoretically trained, with CRM, and the ability to still "fly the plane" when things go south.. that's what they're paid for.. to get you home safely when things go awry. Gimli glider, Air Transat, sully, Sioux City DC10, Baghdad DHL Airbus, etc., all these guys were dealing with scary and confusing situations.. but they're real aviators and figured it out as best they could. Plus, the single biggest fact to me that this was a human solvable issue was:
and we know that the flight before the Lion Air accident....was able to disable the system and safely fly the aircraft


Plus the order of magnitude more MAX that were safely flying around the US still leads me back to deficient airmanship

"they got overwhelmed, it's excusable" <- no, it's not. And the AD from Nov of 2018 basically spelled this failure out very neatly

..and yes, my perspective would be different if this were an AAL, UA, DL crash.. but the first plane (Lion Air) was crashed by an airline that has a garbage safety record, and both flights were operated by a crew who has culturally shown will fly perfectly good airplanes into seawalls on a VFR day, shut down the incorrect engine, etc., It seems the training is too rote, too systems focused, and not enough on actual skill based hand flying of aircraft
 
and we know that the flight before the Lion Air accident....was able to disable the system and safely fly the aircraft. The next crew....not so much.

Wasn’t it attributed to a dead heading pilot sitting in the jump seat? One that wasn’t in the middle of it and had time and knowledge to direct them to the correct procedure?

I don’t understand, from what I read no pilots were able to get this specific training because there was no actual simulator that was set up for this. “Fly the plane”, I don’t know airliners, and am just a student in GA, but generally when we have problems the plane isn’t actively fighting to do the wrong thing.

I see folks go into detail about this seeming to be a hack, instead of integrated correctly, and somewhat of a conflict of interest in that Boeing apparently sold this plane as low cost in that it didn’t require extra training, but it seems they should have. Just as a passenger I would want to know that the pilot on the plane has trained in the procedure for recovery. Who is that up to? I’m not sure, but between the airlines themselves and the FAA I would have expected this to be a requirement before a pilot got to fly it.
 
remember....these were not US operated airlines and have different requirements. US carriers got different equipment and probably different training.....and there was an AD issued "last" year.
 
There's a good article in the Seattle Times by their aviation writer. Among various tidbits:
  1. The 737 max has two AOA indicators, but MCAS uses only one
  2. The software does not ground check the AOA indicators, where they should read zero, nor does it compare them against each other
  3. The MCAS system will repeatedly apply nose down trim, which the pilots apparently trimmed out after each erroneous reaction to a bad AOA indication, until trim is fully nose down. Then your hands will be full all of a sudden.
All the above seems like a strange set of software engineering decisions for an important flight safety system. While pilots may be able to decipher the faulty AOA issue from the cryptic behavior of the MCAS system, it seems wholly unnecessary to have it be so sensitive to a single point failure mode that potentially leads to a serious out of trim condition that may not be so easy to recognize without specific training. How much would it really cost to ground check and cross check the AOA indicators and limit the assertiveness of the assisted control? Or annunciate the AOA discrepancy in flight or on the ground? For one of the two crash flights, it was known that the AOA indicators disagreed by 20 degrees on the ground. The "bad" one was attached to the MCAS system.

Additionally, it seems barely defensible not to own up to the fact that the system as originally designed was not as safe as it could have been. That's not a reassuring safety culture. It's the kind of behavior that blows up space shuttles. Blaming the pilots is an irrelevant distraction from the core issue.
 
Yesterday's (March 20) AW&ST article states that the MCAS's primary data source are the MAX's two AOA sensing vanes, one on either side of the nose, and that Boeing designed the MCAS to receive input from only one of the sensors during each flight. The left and right sensors alternate between flights, feeding AOA data to the FCC and the MCAS.

That could explain why the Ethiopian flight 302 airplane's previous flight was uneventful, but the ill-fated flight had trim problems immediately after takeoff.
 
Additionally, it seems barely defensible not to own up to the fact that the system as originally designed was not as safe as it could have been. That's not a reassuring safety culture. It's the kind of behavior that blows up space shuttles. Blaming the pilots is an irrelevant distraction from the core issue.
This.
 
I just read an article that said an AoA disagree warning is available on the MAX, but neither Lion Air or Ethiopian had purchased the optional feature.

Now Boeing is saying it will become a standard feature on future builds.

Again, this seems like a crazy omission on Boeing's part. Making it optional? Jeez.
 
....
  1. The MCAS system will repeatedly apply nose down trim, which the pilots apparently trimmed out after each erroneous reaction to a bad AOA indication, until trim is fully nose down. Then your hands will be full all of a sudden.

See, that right there doesn’t even make sense. If the pilots trim out the MCAS created trim each time how do you end up with full nose down trim? There’s something else to this story. I’ve worked avionics for a major airline for 30 years and we virtually have never had trouble with AOA sensors. I can’t recall ever replacing one for anything other than a time based call out. So that right there is odd.
 
Yesterday's (March 20) AW&ST article states that the MCAS's primary data source are the MAX's two AOA sensing vanes, one on either side of the nose, and that Boeing designed the MCAS to receive input from only one of the sensors during each flight. The left and right sensors alternate between flights, feeding AOA data to the FCC and the MCAS.

That could explain why the Ethiopian flight 302 airplane's previous flight was uneventful, but the ill-fated flight had trim problems immediately after takeoff.
I read an article recently that recounted the flight crew - 1 day prior to the Lion Air crash - experienced the excessive nose down trim and a 3rd pilot riding in the jump seat saved their bacon by recognizing the problem and got them to disengage the MCAS or some part of it. The next day crash was a different crew. Not sure if the crash flight was "the other AOA" or not.

I thought i also read that during one of the crashes the flight data recorded the pilot applying over 100lbs of back force in the stick prior to the crash...you cant do that very long!!!
 
They worked on the Lion Air airplane the night after that first crew experienced the trim issue and the jumpseater flipped the stab trim cutouts. It's unclear from the released crash investigation information which AOA sensor was providing input to the FCC and MCAS on the fatal flight.
 
Additionally, it seems barely defensible not to own up to the fact that the system as originally designed was not as safe as it could have been. That's not a reassuring safety culture. It's the kind of behavior that blows up space shuttles. Blaming the pilots is an irrelevant distraction from the core issue.
Seems to me that whether or not it would be reasonable to expect competent pilots to disengage a malfunctioning trim system IS the core issue. If it’s reasonable, Boeing and the FAA are off the hook. If it’s not reasonable, they’re culpable.

Not that it will be an either/or situation...I’m sure somewhere in between.
 
Last edited:
Since, as described, all one needs to do to override the MCAS (even if temporarily) is to thumb the trim switch on the yoke it would seem almost obvious that there is more to this than we are imagining. Especially considering that both of these catastrophes happened in clear daylight conditions. Something’s missing in this puzzle.
 
If these issues had been caused by a straight runaway trim, would the outcome have been the same? In other words, does the condition look, feel, or act as anything else but R/T?

I know the stick shaker is involved, but as Tantalum said earlier, when the rotating big wheel keeps moving the nose down trying the kill you, why hasn't the response been to stop that from happening?

Were these third world crews frozen in their tracks due to lack of training on a basic yet emergency fault, R/T?
 
Since, as described, all one needs to do to override the MCAS (even if temporarily) is to thumb the trim switch on the yoke it would seem almost obvious that there is more to this than we are imagining. Especially considering that both of these catastrophes happened in clear daylight conditions. Something’s missing in this puzzle.
lack of training....

but, that's always a poor excuse for a new design.From a human factors standpoint, training is not a good mitigation to a design issue. Humans are not very repeatable and reliable. Designing the problem out with automation is far more repeatable and reliable.
 
I know the stick shaker is involved, but as Tantalum said earlier, when the rotating big wheel keeps moving the nose down trying the kill you, why hasn't the response been to stop that from happening?

Were these third world crews frozen in their tracks due to lack of training on a basic yet emergency fault, R/T?

As others have said, it may not be this easy, Stick shaker, aural warnings, spinning wheel, increasing yoke pressure, no doubt numerous flashing lights, and probably more than a few error codes scrolling on the displays. So, they're drinking from the fire hose at this point, on an issue that they don't know about nor received any training.

I believe even the best of pilots might be saturated at this point.
 
Cliff Notes version: its unclear whether these pilots trained in an NG sim that didn't simulate MCAS failures, or a MAX simulator that doesn't simulate MCAS failures. :rolleyes:
That's what you get with a typical news outlet; I'm used to it by now. Still
There's a good article in the Seattle Times by their aviation writer. Among various tidbits:
  1. The 737 max has two AOA indicators, but MCAS uses only one
  2. The software does not ground check the AOA indicators, where they should read zero, nor does it compare them against each other
  3. The MCAS system will repeatedly apply nose down trim, which the pilots apparently trimmed out after each erroneous reaction to a bad AOA indication, until trim is fully nose down. Then your hands will be full all of a sudden.
All the above seems like a strange set of software engineering decisions for an important flight safety system. While pilots may be able to decipher the faulty AOA issue from the cryptic behavior of the MCAS system, it seems wholly unnecessary to have it be so sensitive to a single point failure mode that potentially leads to a serious out of trim condition that may not be so easy to recognize without specific training. How much would it really cost to ground check and cross check the AOA indicators and limit the assertiveness of the assisted control? Or annunciate the AOA discrepancy in flight or on the ground? For one of the two crash flights, it was known that the AOA indicators disagreed by 20 degrees on the ground. The "bad" one was attached to the MCAS system.

Additionally, it seems barely defensible not to own up to the fact that the system as originally designed was not as safe as it could have been. That's not a reassuring safety culture. It's the kind of behavior that blows up space shuttles. Blaming the pilots is an irrelevant distraction from the core issue.
I think we should eliminate the pilots if they fly an airplane with wonky AoA indicators, and then cannot figure out a problem before the aircraft crashes. All of my line pilot friends tell me that we can't do without them, but a good AI system would have landed Sully's plane at the airport.
Yes, I'm being snarky, and the system isn't as safe as it could have been, but there are plenty of systems on most planes that require proper pilot actions to work as designed.
 
A continuous cross check of captain and co-pilot side AoA indicators with a discrepancy lockout of this system is the minimum that should have been done. If the system is necessary for flight (which I'm pretty confident it is not), then triple-redundant systems should have been the baseline.
 
As others have said, it may not be this easy, Stick shaker, aural warnings, spinning wheel, increasing yoke pressure, no doubt numerous flashing lights, and probably more than a few error codes scrolling on the displays. So, they're drinking from the fire hose at this point, on an issue that they don't know about nor received any training.

I believe even the best of pilots might be saturated at this point.
That's why we're trained to 'fly the airplane' first.

Climb power and about 10deg ANU (aircraft-nose-up) and you know you'll fly. (Trim nose up) There are three attitude references in the airplane, which two agree? (Trim nose up) Now you have time to figure out what is going on. (Trim nose up) There are three airspeed indicators, which two agree? (Trim nose up) There are three altimeters, which two agree? (Trim nose up) There are two VSIs and three altimeters, do the majority of those indicate a climb? (Trim nose up) Hey! Why do I keep having to trim nose up? There's my problem!

The ability to 'fly the airplane' first is essential to safe air travel in everything from a Piper Cub to Concorde. The bigger, faster, airplanes just have the capability to present a lot more distractions.

Something else I've been thinking about but can't come up with a good answer.

A normal takeoff in a 737 involves climbing to 800' AFL (sometimes 3000' AFL), lowering the nose to accelerate then retracting flaps on schedule. (25 -> 15 -> 5 -> 1 -> Up. Most takeoffs are flaps 5 so you'd start there. Flaps 1 is minimum for takeoff)

MCAS is inhibited if flaps are not up.

We (still) don't have the data on the Ethiopian accident but know that on the Lion Air accident they had a stick shaker (Captain's side only) from liftoff. If the stall warning is going off, why would you retract the flaps? I wouldn't. There's no compelling reason to retract them. We don't fully retract the flaps on a go-around unless we expect an extended pattern to fit back into the flow at a busy airport. If you're coming right back around to land, leave the flaps at 5 or 1. That's not a rule, limitation, or procedure. It's just how I think most 737 pilots would fly.

Though they wouldn't have known it, if they did leave the flaps at 1 or 5, the MCAS never would have activated. Why did they retract flaps while the stall warning (stick shaker) was going off?
 
they had a stick shaker (Captain's side only) from liftoff
This would also be an immediate and obvious indication that something is gravely F***'d.. so even if you're after V1 I'd cut the power and slam on the brakes. Going off the end at 60 knots will be better than taking a very clearly defective plane into the sky
 
This would also be an immediate and obvious indication that something is gravely F***'d.. so even if you're after V1 I'd cut the power and slam on the brakes. Going off the end at 60 knots will be better than taking a very clearly defective plane into the sky
A false stick shaker activation is nowhere near a “clearly defective plane”...it’s still flyable, and if the plane is flyable, it’s normally far safer to take it into the air after V1.

I’ve made a couple of trips around the patch with activated stick shakers over the years...it isn’t a big deal.
 
A false stick shaker activation is nowhere near a “clearly defective plane”...it’s still flyable, and if the plane is flyable, it’s normally far safer to take it into the air after V1.

I’ve made a couple of trips around the patch with activated stick shakers over the years...it isn’t a big deal.
Exactly. I could tell you many cases where the shaker went off and it was no issue with the aircraft, it was an issue with the shaker and associated sensors. It is a "land asap" item, but not a "slam on the brakes and go into the ditch" item.
 
A false stick shaker activation is nowhere near a “clearly defective plane”
Exactly. I could tell you many cases where the shaker went off and it was no issue with the aircraft
Got it.. no experience with big iron.. with the "hindsight is 20/20" thing though these guys would have been better off. Obviously the whole "fly the plane first" .. even if that just means straight pitch and power with trim cut out.. was beyond these aviator's capabilities
 
a "slam on the brakes and go into the ditch" item
are there any such items? If you are after V1 is there any published thing that would cause you to stay on the ground, or is that left up to the pilot's ADM to judge how catastrophic the situation is
 
are there any such items? If you are after V1 is there any published thing that would cause you to stay on the ground, or is that left up to the pilot's ADM to judge how catastrophic the situation is

I posted this a little while ago:

https://www.pilotsofamerica.com/community/threads/ntsb-report-on-a-rwy-excursion.117495/

The Captain aborted takeoff after V1, his checkpilot gave him grief but didn't override.

From the NTSB report:

>>>
However, company guidance also stated that a high-speed rejected takeoff should be made only for safety of flight items, such as a condition where there is serious doubt that the airplane can safely fly. Boeing guidance also stated that rejecting the takeoff after V1 is not recommended unless the captain judges the airplane to be incapable of flight.
<<<
 
This would also be an immediate and obvious indication that something is gravely F***'d.. so even if you're after V1 I'd cut the power and slam on the brakes. Going off the end at 60 knots will be better than taking a very clearly defective plane into the sky
First, you're not on the runway anymore so you cut the power and you fall, very hard, back to the ground. If you try to do it more gracefully, you're unlikely to touchdown prior to the end of the runway.

TWA tried this in an L1011 at JFK in 1992. Luckily, everyone survived, although one passenger received injuries in the evacuation that eventually put him in a wheelchair. They were lucky that there were no fatalities due to smoke/fire as the aircraft was destroyed by the post-crash fire. As the NTSB found, the airplane was performing normally, except for the invalid stick shaker activation, and could have been flown safely to an emergency landing.

https://en.wikipedia.org/wiki/TWA_Flight_843
 
I've been seeing this talk of a warning light go around
https://www.cbsnews.com/news/boeing...d-light-warning-pilots-of-sensor-malfunction/

So when they say indicator light- does that mean a separate physical light on the panel? Wouldn't this same information be displayed on flight displays?

Also- $80,000 for an indicator light? Seriously? It takes $80k to run a wire from what would have to be a computer output pin to a little light bulb/LED on the panel?
 
That was after V1 but while still on the runway. The TWA accident was after liftoff (16' AGL). Quite different outcomes from these two accidents.
I just read that - ouch.

The more stuff on an airplane, the more stuff that can go wrong. That's why those two seats up front need some good wetware occupying them.
 
and could have been flown safely to an emergency landing
Thanks. That's a highly theoretical condition though.. chances are that both MAX planes that crashed could also have been flown safely.. (as was evidenced by the prior crew who cut out the trim and flew it safely)
 
So when they say indicator light- does that mean a separate physical light on the panel? Wouldn't this same information be displayed on flight displays?
I don't know. The 737, being a design from the 1960s, does not have an EICAS (or ECAM in Airbus terminology) which displays short text messages on the central display. It has old fashioned warning lights and a "Recall" system which alerts the pilots (Master Caution light) and directs them to the appropriate system on the overhead panel where the light is located (Recall Panel).

That said, the instrument warnings, such as "IAS DISAGREE" and "ALT DISAGREE" appear on their respective airspeed and altitude "tapes" on the pilots primary flight displays.

The AoA display and disagree are more associated with the HUD installation and its use for lower than standard approaches and departures (CAT II/III). I've never flown an airplane with a HUD (or an AoA indicator for that matter) but those airplanes use the HUD to hand-fly approaches and departures in lower than CAT I ILS or standard reduced visibility takeoff minimums (RVR 5/5/5). I believe the AoA indicator is required for those operations (can anyone confirm?) so the airlines that use the HUD for those operations would have them. Airlines that use autoland for CAT II/III operations likely would not have the AoA indicators.
 
Thanks. That's a highly theoretical condition though..
That was the NTSB's conclusion. They found nothing else wrong with the airplane and that the airplane was performing normally up to the point that the takeoff was aborted.

I would say that a successful emergency landing would have been the expected outcome of that TWA flight if the takeoff had been continued.

As you already know, the Lion Air accident airplane had the same failure on its previous flight. On that flight, the crew accomplished the runaway stabilizer procedure, after being prompted by their jumpseater???, then continued on to their destination for a safe landing using only manual stab trim. What may not be obvious to all is that the Captain's stick shaker was activating FOR THAT ENTIRE TWO-HOUR FLIGHT!

Why they would have continued to their destination with a constant, and unsilenceable, stick shaker activation is mystifying. They had lost their stall warning system, primary trim, and autopilot but continued the flight?
 
Why they would have continued to their destination with a constant, and unsilenceable, stick shaker activation is mystifying. They had lost their stall warning system, primary trim, and autopilot but continued the flight?
JFC. I didn't know about that. Even if you can deal with the stick shaker for 2 hrs.. you've lost a lot of key safety components

and why did they continue?..because they're a garbage airline with crap training that now Boeing is feeling the pain for
 
JFC. I didn't know about that. Even if you can deal with the stick shaker for 2 hrs.. you've lost a lot of key safety components

and why did they continue?..because they're a garbage airline with crap training that now Boeing is feeling the pain for
A single airplane with what seems like serious system problems, one crew that ignores it and continues a flight, and another crew that crashes.

Does NTSB get to issue a report on a US manufactured aircraft that belongs to a foreign carrier and crashed outside the US? I've heard NTSB is "assisting", but that may be more for technical help in gathering the data and not as part of the whole investigation. I'm not sure where jurisdictional lines are drawn.
 
A single airplane with what seems like serious system problems, one crew that ignores it and continues a flight, and another crew that crashes.

Does NTSB get to issue a report on a US manufactured aircraft that belongs to a foreign carrier and crashed outside the US? I've heard NTSB is "assisting", but that may be more for technical help in gathering the data and not as part of the whole investigation. I'm not sure where jurisdictional lines are drawn.
The NTSB has issued reports of foreign incidents in the past; I don't know if they are official, or just advisory.
 
Back
Top