I will also make the observation that much of the critical commentary toward the crew's action is focusing on their initial response to the event (stick shaker at rotation) before they ever got to the point where MCAS activated (flaps retracted). Up to that point, there was no significant difference between how a MAX responds vs an NG, so you can't really pin their initial response on Boeing's hash job with MCAS. The fact that these crews were struggling with basics of airmanship and compliance with a well known procedure is suggestive of an overall human factors issue.

If you look at the DFDR data, the indicated airspeeds were already diverging from liftoff, so no, they were not "normal".

The preliminary accident report contains a synopsis of the discussion between the Captain and the First Officer. At 400', the FO selected VNAV and the Captain called for "Command", but the A/P did not engage. The Captain called out for "Command" again. The A/P did not engage. He asked the FO to change to a new ATC frequency. The FO checked in with radar, received an ATC instruction, and responded. The Captain called for flap retraction, and the FO responded. The Captain asked the FO to request runway heading from ATC. The FO did so.

What is conspicuously absent from the dialogue is any attempt to confirm that their airspeeds were indeed "normal". They were not. Despite having multiple warnings that are explicitly stated in AD 2018-23-51 (the 737 AD published post Lion Air), this crew made no apparent attempt to determine this crucial fact. If they had done so, then performance of the Airspeed Unreliable non-normal would have been anything but discretionary.

Yes, to often that isn't the case any more. I think it would be clear to anyone that actually did read Fullwings post(very mature post in terms of how a preconceived problem identification could kill you) that he is a experienced thinking pilot.

Fly the damn airplane...
 

I think your analysis and continuum of blaming the pilots/Boeing and everyone else harmless to blame Boeing and everyone else/pilots harmless is a nice summary.

I will tell you where I sit on this, based upon 36 years of professional flying (31 airline, 5 military), type rated on B737-200/767/777/787 and Airbus A330/A340, and a Professional Engineer.

First, aircraft are amazing and complex machines. Aircraft do what man was never intended to do naturally so there is an inherent risk in that alone. But even the best designed and best maintained aircraft have components that break and the human onboard is the last line of defense in many of those instances. Hydraulic systems, electrical systems, pressurization systems, propulsion systems (engines) all fail and often there is not double or triple redundancy due to weight issues, cost issues, probability of failure and the impact of a failure. As an example, there is no level of redundancy that will offset an engine failure no matter how many engines the aircraft has; the adverse yaw, the reduced performance, etc cannot be compensated for by having more engines, although the more engines an aircraft has, the less effect a single engine failure has. Aircraft manufacturers, at the urging of their airline customers, like two engine aircraft for the economics (one engine versus two, reduced fuel, reduced weight, etc). So we accept that fact that an engine failure on a two engine aircraft will have big affect when one fails, more than a single engine failure on an eight engine aircraft.

So when an engine (or hydraulic, or electric, or pressurization) system fails a pilot is in the loop to manage the situation. That is why we are highly trained and, hopefully, well compensated financially for that knowledge and judgement. The level of training lies with the individual knowing their stuff including their emergencies, particularly emergencies that are memory items - that is a personal responsibility of any professional pilot. The airline and country CAA is responsible too for insuring that the pilots charged with responsibility for that aircraft and those lives in the aircraft are also trained properly initially and on a recurrent basis.

Where do I sit on your continuum? Somewhere between 2 and 3. MCAS is a required stall protection however it needs to be toned down a bit as it can move the stabilizer trim to very large aircraft nose down angles. Should pilots be aware of MCAS in their technical training on the MAX; sure, but it would not have affected the outcome regardless as, I will describe shortly, MCAS failure presents characteristics identical to a stabilizer trim runaway, an emergency checklist item that has been around since the original B737 fifty years ago.

I would expect that any pilot with a type rating for a Boeing 737, MAX included, should be able to identify an Unreliable Airspeed (UAS). That is basic stuff, stick shaker when the aircraft if flying normally, disparity between the indicated airspeeds, etc. Any professional pilot should be able to recognize this regardless of what aircraft they are flying as every aircraft in the world is subject to this problem.

Only 2 of the 3 non-US MCAS incidents saw the pilots recognize the UAS and do something about it. This was long before MCAS reared its ugly head so that begs the question: Why? Training and experience would be my answer. And, as it turned out, the crew that did execute the UAS drill were the ones that ultimately saved the aircraft. Bear in mind that UAS is a memory drill. To me, there is no excuse why this was not done as it was a textbook UAS; to me also, in the Ethiopian case, engaging the autopilot at 400 ft is a definite faux pas as it is contrary to the memory UAS drill and, if the aircraft was indeed stalled, another definite faux pas as one does not recover from a stall with the autopilot. This points to a training/experience issue too but it also points to an over reliance on the use of the autopilot at the expense of hand flying an airplane. This too is likely an airline/CAA issue as well as an individual pilot issue. And, not to imply that this is a "third world" country issue as I am seeing this more and more with the FO's I fly with; they are terrified to hand fly the aircraft and hence their hand flying skills begin to decline.

When all of these crews experienced MCAS, 0 of 3 were able to recognize a classic stab trim runaway; while manually flying the aircraft, the nose pitches down all by itself. You simply can't miss it, with or without seeing the stab trim wheel moving. In the Lion Air case where the aircraft was saved, it took a third pilot from a different airline to tell the crew what to do; in the case of the fatal Lion Air accident, the Captain handed control to the First Officer so he could go hunting through the checklist for something to do - this is a memory drill! I fault the individual pilots for not knowing their memory emergencies - harsh as that may sound, that is what we are paid for. (Personally, I do a study of the memory emergencies regularly as the memory isn't what it used to be and I like to think that most professionals pilots do the same).

I am not going to rehash the rest of the issues (flying to destination with this problem, not controlling the aircraft speed in the Ethiopian case, etc) as that would be covering old ground but there are lots of basic airmanship issues that are highly questionable.

In conclusion, all of these MCAS issues were recoverable situations by a trained and competent crew. I do not blame the individuals entirely as the airline, their training, their hand flying policies as well as the CAA overseeing them deserve scrutiny too however. Let me repeat that: These MCAS accidents were all recoverable. Boeing's mistake, in addition to what I noted earlier, is assuming that B737 type rated pilots would be able to do even the most basic, memory emergency drills. This, however, begs the question: If pilots cannot even do a simple UAS emergency, what hope do they have for a more complex one such as an engine failure at rotation or an engine failure followed by a cabin depressurization which that Southwest crew handled masterfully?

What this MCAS situation points to, in my estimation, is that the aviation industry has been "whistling past the graveyard" for a little too long and that the underlying problems quickly manifest themselves with even the slightest irregularity - the Turkish B737 accident at Schipol, the Korean 777 in SFO, etc.. A complete rethink of pilot training, basic flying skills and airmanship are in order as there is a finite limit to what Boeing, Airbus, Embraer or any other aircraft manufacture can do to design and build airplanes without human intervention being required a certain times by competent aviators.

A really great post but the conclusion that the answer lies only in training cannot be other than seriously debatable.

You led off with experience for a very good reason. Experience is profoundly important.

Training doesn't provide experience. It's only the entry point.

Automation denies experience.

As an overall improvement we need either reduced reliance on automation enabling greater experience, or improved automation.

We both know which one of those is more likely.

34 years aircraft maintenance.

Good post L39 - but MCAS is far from a stall protection (system). MCAS is a system that supplies an a feel to the control column, and the aircraft can fly perfectly without it.

But without MCAS the 737 MAX could not meet the certification requirements for Grandfather rights. Without the Grandfather rights the 737 MAX requires more certification and pilots will require extra training - this would make the MAX less desirable/competitive.

So the task was to incorporate a new system without calling it "new" and retain Grandfather rights. It seems to me that the priority was the "extra training" being minimal between the MAX and other 737's (you heavily touched training in your reply) and the result was a few hours of slides on an iPad that included nothing on the new system.

If the Grandfathering certification allowed for MCAS as a new system, and it was designed and implemented as a stand alone system with dedicated on/off switches and proper training - I doubt we would have had the crashes.

Even now MCAS seems to be strange! (I do not recall exact numbers) but MCAS was said to have a limit of 0.6 in certification. From memory that was not enough so it was increased to 2.5 to get the required "feel" to pass the test. Yet the "software update" will reduce this to either 0.6 or 0.9 ( I do not recall) so how then can it pass the test when 2.5 was required?

There also still seems to be far too much effort, to not have this aircraft require "extra" training, particularly in the sim. My view is that this "fix" will have put the "feel" somewhat less than what the original MAX had, this feel was to give the pilot an aid to prevent pulling back too far preventing a stall. So I believe that simulator training should be carried out on the MAX around the stall area in the parameters that MCAS is enabled. Obviously this training should be with and without MCAS engaged as there is included in the fix more ways the MCAS can be disengaged, and pilots left with a soft stick around the stall zone.

Further more, given it seems decisions at Boeing have been made heavily for financial reasons on the MAX project - I would hope a full review of the certification is/has been made external of Boeing.

Training - FAA and Boeing seem against it other than an iPad slide show, a few more craters till that changes.

The faults in the US reported of altitude change when AP selected - has this problem been identified and rectified?

We have an issue here that was fatal two times out of three. As far as we know.

Every airline in the world has contributed to an improving safety record over the years.

But people here are still pushing the "well, actually, had the crew had more basic flying skills, like us [adjective noun,unspoken] , it woulda been fine."

Would it?

A simply question - had these 3 crews been flying any other B737 predecessor, would the aircraft have crashed?

C/Would it?

Obviously!

Does the system have protection against using data from a sensor with internal or wiring issues?
Safety related angular encoders used in other industries are having a dual architecture inside, the labeling of signals SIN,COS, and COM on an conceptual drawing could indicate this. (Two independent channels: one analogue, and one digital)
This does however not protect against a sensor with a frozen, or damaged mechanical vane from providing false data.

Is that a obviously yes or a obviously no answer Capn Bloggs?

Keeping in mind, no other breed of the 737 would have put a nose down input to the stabiliser.

As a side, a run away stabiliser you would expect after a input was made and a stop of input expected, but did not happen. That is not what happened in these two cases as no nose down input should have happened (it was not an over run, but more like a 1958 red Plymouth Fury event)

Not in Boeing parlance.
Yes

"Fairly confident". Would you agree if the discrepancy at that stage was less than five knots there would be no such annunciation?

"apparent". ie nothing recorded on the CVR. You exclude the possibility of non-verbalised assessment?

It would only be a human factors issue, in the negative sense, if it could be shown that there was a significant deviation from the performance that could be expected of an average crew on an average day. Your use of words such as "clearly" and "basic" are argumentative strategies to support your point of view but are not justified.
You are not building a watertight case.

Sorry there BAL, I misread your question. If you're used to manually trimming ie hand-flying, you're in a much better position to detect when the trim has gone haywire, and much more likely to do something about it ie trim back to neutral. It goes without saying that if you are used to handflying and trimming, when you are forced to, it won't be a total shock to have to do so and you'll have more spare brainspace to think about what is occurring and take action (or instruct your mate) to resolve the situation.

Not a pilot. To paraphrase what you seem to be saying: When all the trim automation (including MCAS) fails, it hands you a manually controlled aircraft, in a mis-trimmed configuration, that a pilot without frequent practice is unfamiliar with. Sounds like ET302, or is the problem more widespread?