4 ms·

The Boeing 737 Max 8 Crashes: The Case for Pilot Error

2 points by hsnewman 7y ago

PaulHoule 7y ago
Sure, the pilots did the wrong thing. But they weren't trained to do the right thing. It is normal to get flustered in a stressful situation, the answer is to prepare for it. If Boeing doesn't want pilots to make the same mistake it should insist that they get trained. My suspicion is that return-to-flight for the MAX is being delayed by Boeing trying to avoid a simulator training requirement for the MAX because the whole point of the MAX was that pilots did not need to be retrained.
salawat 7y ago
Oh, it's this song and dance again. Let me touch on a couple points: >On overspend and throttle management The article asserts: >The Pilots left the throttles at 94% power (normal ascent power) for the entirety of the emergency. >this exaggerated aerodynamic load forces on the horizontal stabilizer, rendering manual wheel retrimming infeasible >This occurred due to lack of pilot training First off; not managing the throttles at that time is completely understandable. The AoA sensor being bad causes absolute havoc with the data available to the pilots since the AoA is a value utilized to calibrate several key data points required for successfully flying the plane, the most important being; airspeed. As soon as that AoA sensor was rendered out of whack, they were essentially flying in an airspeed unreliable state, to which the remedy is... You guessed it: advance throttles to appropriate position for desired flight regime, move on to resolving next problem. Now; there's not actually any documentation or training material that specifically mentions that remedying an airspeed unreliable could be contraindicated in the presence of some other problems. If you do both have an explicit call out for this sort of thing, you never even incorporate the possibility into a mental fault diagnostic tree; you just try to correct the seeming most present issue. Also note: there's an immediately observable effect to throttle retardation in engine underslung aircraft. Every pilot would know this. No pilot in their right mind would naturally make the problem worse until they absolutely exhausted their checklist items. It's simply not good form. So, into Stab Trim Runaway we go! >On the trim oops The monitoring pilot had about 200 hours; the pilot in command had about 8000 hours, alarms and glaciers going off, and a painful amount of pressure having to be held on the controls to try to keep things from getting worse. They had both seen the emergency bulletin from Boeing, and we're therefore primed for stab trim runaway. In that bulletin, airspeed gets a minor callout at the end. Almost an afterthought really. Somewhat lost in retrospect to the list of symptoms that may arise due to AoA sensor malfunction, and the news of this MCAS thing that might try to kill you. Now: They cutout the trim switches. I have a question; how many, dear reader, have, when faced with something highly resistive to turning, naturally, without thinking tried turning the thing in question the other way? It takes 40 turns per unit of trim if I recall cotrectly; this means no more than 8 turns of the wheel (which would smoothly actuate in the nose down direction since the screw jack in back is being assisted by the aerodynamic load) accounts for the change from 2.1 to 2.3 units. This strikes me as just about the right amount of time for the human eye to discern movement on an analog gauge if that is what is used, or to filter out from a chaotic interface that the trim problem is being exacerbated. I can hear the protest from everyone in the back, but the fact remains, that when under pressure, and in the absence of a visual cue, it is more than possible that either pilot fell victim to this most human of heuristics before quickly realizing their mistake. >on departing from the checklist Having run through everything they had from an authoritative source, the pilot's found themselves in a position where the choice was improvise or die. They improvised. They reactivated the electric trim motor and slipped the thumb switch to make sure the system could respond. Unfortunately, this also reset MCAS to activate 5 seconds later, before they could hit the trim again. They had no clue as to the coupling between the trim switches, or disabling of the air-trim override switches in the yokes. They did the best they knew how with the cards they were dealt. >Conclusion: Did pilot error doom those flights? In the sense that lithobreaking with the ground is the physical cause of death, yes; but we aren't discussing the physics, we're discussing culpability. The primary point of interest in questions of culpability is: Were they negligent? Did they unreasonably endanger the plane by doing anything other than a best faith demonstration of mechanical and piloting aptitude in the face of incomplete information and malfunctioning equipment? I assert that no, the pilot's acted in good faith, and to the extent of their training and beyond as well as any human being trying to operate a highly complex and unknowingly dangerous piece of equipment can be expected to under the circumstances. Was Boeing negligent? Did they in good faith do everything they could in order to ensure the safety of these planes while wholeheartedly complying with aerospace design regulations, free of any form of malicious compliance, or free of any willingness to unreasonably sacrifice the soundness of the finished product in the name of greed/dedication to maintaining market share? I assert so. This unfortunate series of reactions could only have come to fruition through the combined negligence in creating an unsafe design, patching over this danger through poorly evaluated or extrapolated code, removing all mention of the danger from the training material, and refusing to perform their duty as a maker of complex equipment to make sure pilots knew enough to safely handle the plane. All with the intent of maintaining market share, and shareholder value creation at the expensive of the quality of the product, and the process by which said product is handed off to its users. No pilot should stand by and let these men be disparaged. No professional should either. This company, on whose auspices and assurances they placed their lives and the lives of their passengers willingly abrogated their responsibility in the name of financial expediency. "The customer already expects the box to be of high quality; we must focus on shareholder value", should clearly illustrate where Boeing's priorities were at the time. That in trying to recover from this negligence on Boeing's part does not justify or release Boeing from ultimate culpability. If we as a people accept the logic whereby these men must carry ultimate culpability, we accept that it is fine for an organization of people to be negligent in the execution of their duty, just so long as the poor sod who pays the price for it makes a mistake. Please don't be that group of people.
- PaulHoule 7y ago
  Charles Perrow, in his book "Normal Accidents" points out that the people killed in an accident like that can't speak out on their own behalf.
  - salawat 7y ago
    Hence my attempts to be so vocal on the subject. Speaking of which, in the paragraph starting "Was Boeing negligent?" in case it was confusing to parse, the "I assert so" was in answer to that question. Not to any of the subclasses. Wish I would have caught that before the editing period expired.
- mannykannot 7y ago
  When you write "Also note: there's an immediately observable effect to throttle retardation in engine underslung aircraft", are you referring to the airplane pitching downwards, on account of the thrust line being below the center of drag?
  - salawat 7y ago
    Yeppers. That is what I'm referencing.
mannykannot 7y ago
The authors of this article repeatedly ask why Ethiopian Airways did not train the pilots for the situation they faced, ignoring the fact that their training was in accordance with the requirements set out by Boeing and the FAA. Prior to the Max variants, trim runaway was normally stopped by the pilot resisting it with an opposing force on the yoke; this caused the system driving the trim change to stop. consequently, trim runaways rarely resulted in the airplane being much out of trim. This is also the case for the Max variants, except for MCAS-commanded changes, as this would defeat its purpose. Training for trim runaway in a simulator goes through the checklist memory items, ending with the toggling of the cutout switches. Any use of the trim wheels, let alone in the 'roller-coaster' maneuver for when they are hard to move, was dropped from training a long time ago [1]. The 737 AFM mentions the possible difficulty of using manual trim when the jack screw is under high load, and outlines the roller-coaster maneuver, but the checklist for trim runaway makes no mention of it and does not say anything about slowing down. Furthermore, the AD, issued after the cause of the Lion Air crash was found, says nothing about these issues. Nor does there seem to be any consideration that the roller-coaster maneuver is infeasible at low altitudes, or of the complication that, as mentioned by Salawat, reducing thrust results in the airplane pitching downwards, on account of the low-slung engines. So, MCAS raises the possibility of trim runaway being a qualitatively different experience than before, and significantly more difficult to handle, yet Boeing's and the FAA's response was to merely reiterate the existing procedures. Most significantly, they did not require, or even propose, any additional simulator training, even after the Lion Air crash. I think it is likely that no-one in these organizations connected the dots to anticipate the problem, but we should also consider the possibility that someone did, but it was downplayed to avoid the question of additional MCAS-specific training. As I have written elsewhere, the more you argue that the pilots made mistakes, the more you simultaneously argue that Boeing's and the FAA's response to the Lion Air crash was inadequate. The authors of this hatchet job say that "with hindsight", the situation could have been saved, and maybe they are right, but the one thing no-one has when they need it is hindsight. Boeing and the FAA had a responsibility to substitute training for hindsight, and if they intentionally defaulted on this, and especially if they did so in the name of sales, they are culpable. [1] https://www.youtube.com/watch?v=n4qDLR4s45U&t=296s https://www.youtube.com/watch?v=n4qDLR4s45U&t=296s

2 points by hsnewman 7y ago

Sure, the pilots did the wrong thing. But they weren't trained to do the right thing. It is normal to get flustered in a stressful situation, the answer is to prepare for it. If Boeing doesn't want pilots to make the same mistake it should insist that they get trained. My suspicion is that return-to-flight for the MAX is being delayed by Boeing trying to avoid a simulator training requirement for the MAX because the whole point of the MAX was that pilots did not need to be retrained.

salawat 7y ago

Oh, it's this song and dance again. Let me touch on a couple points: >On overspend and throttle management The article asserts: >The Pilots left the throttles at 94% power (normal ascent power) for the entirety of the emergency. >this exaggerated aerodynamic load forces on the horizontal stabilizer, rendering manual wheel retrimming infeasible >This occurred due to lack of pilot training First off; not managing the throttles at that time is completely understandable. The AoA sensor being bad causes absolute havoc with the data available to the pilots since the AoA is a value utilized to calibrate several key data points required for successfully flying the plane, the most important being; airspeed. As soon as that AoA sensor was rendered out of whack, they were essentially flying in an airspeed unreliable state, to which the remedy is... You guessed it: advance throttles to appropriate position for desired flight regime, move on to resolving next problem. Now; there's not actually any documentation or training material that specifically mentions that remedying an airspeed unreliable could be contraindicated in the presence of some other problems. If you do both have an explicit call out for this sort of thing, you never even incorporate the possibility into a mental fault diagnostic tree; you just try to correct the seeming most present issue. Also note: there's an immediately observable effect to throttle retardation in engine underslung aircraft. Every pilot would know this. No pilot in their right mind would naturally make the problem worse until they absolutely exhausted their checklist items. It's simply not good form. So, into Stab Trim Runaway we go! >On the trim oops The monitoring pilot had about 200 hours; the pilot in command had about 8000 hours, alarms and glaciers going off, and a painful amount of pressure having to be held on the controls to try to keep things from getting worse. They had both seen the emergency bulletin from Boeing, and we're therefore primed for stab trim runaway. In that bulletin, airspeed gets a minor callout at the end. Almost an afterthought really. Somewhat lost in retrospect to the list of symptoms that may arise due to AoA sensor malfunction, and the news of this MCAS thing that might try to kill you. Now: They cutout the trim switches. I have a question; how many, dear reader, have, when faced with something highly resistive to turning, naturally, without thinking tried turning the thing in question the other way? It takes 40 turns per unit of trim if I recall cotrectly; this means no more than 8 turns of the wheel (which would smoothly actuate in the nose down direction since the screw jack in back is being assisted by the aerodynamic load) accounts for the change from 2.1 to 2.3 units. This strikes me as just about the right amount of time for the human eye to discern movement on an analog gauge if that is what is used, or to filter out from a chaotic interface that the trim problem is being exacerbated. I can hear the protest from everyone in the back, but the fact remains, that when under pressure, and in the absence of a visual cue, it is more than possible that either pilot fell victim to this most human of heuristics before quickly realizing their mistake. >on departing from the checklist Having run through everything they had from an authoritative source, the pilot's found themselves in a position where the choice was improvise or die. They improvised. They reactivated the electric trim motor and slipped the thumb switch to make sure the system could respond. Unfortunately, this also reset MCAS to activate 5 seconds later, before they could hit the trim again. They had no clue as to the coupling between the trim switches, or disabling of the air-trim override switches in the yokes. They did the best they knew how with the cards they were dealt. >Conclusion: Did pilot error doom those flights? In the sense that lithobreaking with the ground is the physical cause of death, yes; but we aren't discussing the physics, we're discussing culpability. The primary point of interest in questions of culpability is: Were they negligent? Did they unreasonably endanger the plane by doing anything other than a best faith demonstration of mechanical and piloting aptitude in the face of incomplete information and malfunctioning equipment? I assert that no, the pilot's acted in good faith, and to the extent of their training and beyond as well as any human being trying to operate a highly complex and unknowingly dangerous piece of equipment can be expected to under the circumstances. Was Boeing negligent? Did they in good faith do everything they could in order to ensure the safety of these planes while wholeheartedly complying with aerospace design regulations, free of any form of malicious compliance, or free of any willingness to unreasonably sacrifice the soundness of the finished product in the name of greed/dedication to maintaining market share? I assert so. This unfortunate series of reactions could only have come to fruition through the combined negligence in creating an unsafe design, patching over this danger through poorly evaluated or extrapolated code, removing all mention of the danger from the training material, and refusing to perform their duty as a maker of complex equipment to make sure pilots knew enough to safely handle the plane. All with the intent of maintaining market share, and shareholder value creation at the expensive of the quality of the product, and the process by which said product is handed off to its users. No pilot should stand by and let these men be disparaged. No professional should either. This company, on whose auspices and assurances they placed their lives and the lives of their passengers willingly abrogated their responsibility in the name of financial expediency. "The customer already expects the box to be of high quality; we must focus on shareholder value", should clearly illustrate where Boeing's priorities were at the time. That in trying to recover from this negligence on Boeing's part does not justify or release Boeing from ultimate culpability. If we as a people accept the logic whereby these men must carry ultimate culpability, we accept that it is fine for an organization of people to be negligent in the execution of their duty, just so long as the poor sod who pays the price for it makes a mistake. Please don't be that group of people.

PaulHoule 7y ago
Charles Perrow, in his book "Normal Accidents" points out that the people killed in an accident like that can't speak out on their own behalf.
- salawat 7y ago
  Hence my attempts to be so vocal on the subject. Speaking of which, in the paragraph starting "Was Boeing negligent?" in case it was confusing to parse, the "I assert so" was in answer to that question. Not to any of the subclasses. Wish I would have caught that before the editing period expired.
mannykannot 7y ago
When you write "Also note: there's an immediately observable effect to throttle retardation in engine underslung aircraft", are you referring to the airplane pitching downwards, on account of the thrust line being below the center of drag?
- salawat 7y ago
  Yeppers. That is what I'm referencing.

mannykannot 7y ago

The authors of this article repeatedly ask why Ethiopian Airways did not train the pilots for the situation they faced, ignoring the fact that their training was in accordance with the requirements set out by Boeing and the FAA. Prior to the Max variants, trim runaway was normally stopped by the pilot resisting it with an opposing force on the yoke; this caused the system driving the trim change to stop. consequently, trim runaways rarely resulted in the airplane being much out of trim. This is also the case for the Max variants, except for MCAS-commanded changes, as this would defeat its purpose. Training for trim runaway in a simulator goes through the checklist memory items, ending with the toggling of the cutout switches. Any use of the trim wheels, let alone in the 'roller-coaster' maneuver for when they are hard to move, was dropped from training a long time ago [1]. The 737 AFM mentions the possible difficulty of using manual trim when the jack screw is under high load, and outlines the roller-coaster maneuver, but the checklist for trim runaway makes no mention of it and does not say anything about slowing down. Furthermore, the AD, issued after the cause of the Lion Air crash was found, says nothing about these issues. Nor does there seem to be any consideration that the roller-coaster maneuver is infeasible at low altitudes, or of the complication that, as mentioned by Salawat, reducing thrust results in the airplane pitching downwards, on account of the low-slung engines. So, MCAS raises the possibility of trim runaway being a qualitatively different experience than before, and significantly more difficult to handle, yet Boeing's and the FAA's response was to merely reiterate the existing procedures. Most significantly, they did not require, or even propose, any additional simulator training, even after the Lion Air crash. I think it is likely that no-one in these organizations connected the dots to anticipate the problem, but we should also consider the possibility that someone did, but it was downplayed to avoid the question of additional MCAS-specific training. As I have written elsewhere, the more you argue that the pilots made mistakes, the more you simultaneously argue that Boeing's and the FAA's response to the Lion Air crash was inadequate. The authors of this hatchet job say that "with hindsight", the situation could have been saved, and maybe they are right, but the one thing no-one has when they need it is hindsight. Boeing and the FAA had a responsibility to substitute training for hindsight, and if they intentionally defaulted on this, and especially if they did so in the name of sales, they are culpable. [1] https://www.youtube.com/watch?v=n4qDLR4s45U&t=296s https://www.youtube.com/watch?v=n4qDLR4s45U&t=296s