When is The New York Times going to gather the views of experienced pilots rather than alleged, self-promoting “experts”? The comments from pilots tell the real story.
Prior to reading this article, I emailed a long colleague who is a 737 captain in Australia. He replied along the same lines as the experienced pilots in the comments section: in situations like this, disengage the autopilot and any other automated systems and fly the airplane manually, like in the old days.
I spent my entire working career in the airline industry, and it disturbs me greatly to see this sloppy reporting from the Times.
2
>>> ". . . in situations like this, disengage the autopilot and any other automated systems . . ."
There have not been (known) situations sufficiently like this for such cavalier comments to be useful or even meaningful.
On the 737 MAX, the only aircraft fitted with the MCAS system, MCAS is *only* active in manual flight. The autopilot is not involved.
The situation that some pilots, and others, are identifying as "like this" is runaway stabilizer. Because the effects of the MCAS system are intermittent, when temporarily overridden by the column trim switches, it is entirely likely that the flight crew (inundated by warnings, the stick shaker, etc.) was unable to recognize the problem in time to correct it -- especially since they had no idea that MCAS existed and were flying at low altitude.
2
Did experienced pilots write this software? If not, maybe there should be pilot input on any code written so crucial to flight safety.
4
The article says, "Angle-of-attack sensors are crucial in determining if a plane is stalling."
If that is true, why is it I flew in the USAF for seven years, Pan Am for twenty years, and UAL for ten years and never had an angle-of-attack indicator?
It is neither crucial nor required. Again, this is an example of how reporters who lack expertise in aviation get the story wrong.
There is nothing wrong with the 737 or with Boeing. The procedure for dealing with the problem is in the manual. The pilots on the previous flight used the procedure and landed safely. The pilots on the crashed flight didn't.
5
It's like this, Captain... the AoA SENSOR is an integral part of the stick shaker system of any aircraft you flew at Pan Am and at United. It is usually located on the fuselage on airliners. You probably saw it many times on your walk-arounds. This SENSOR, apparently, is also integral to the M.C.A.S. system that was automatically trimming the horizontal stabilizer, forcing the nose down on Lion Air 610, and doing so erroneously due to a faulty AoA SENSOR. The AoA INDICATOR you cite, that you never saw before in your career, is a cockpit-located instrument that is a repeater of the information derived from the AoA SENSOR, and this instrument is found in all tactical Navy jet aircraft cockpits, aiding the Naval Aviator in Aerial Combat Maneuvers (ACM) and in landing aboard the carrier, while maintaining heads up. Hope this helps.
7
I have a trip on this type of plane scheduled for January, albeit on a large legacy US carrier. I am a nervous flyer to begin with - this accident makes me extremely concerned, especially since it appears that the airline is pointing to Boeing and Boeing is pointing to the crew.
4
I'm not a pilot, but I'm a usability consultant. If a new feature is added to an aircraft that has potential for catastrophe in the event of failure, the procedures to remedy that should not be buried in a manual for others to find. Something of this nature needs to be called out up front - it should be the first thing pilots see about the new plane, in all caps and red letters.
7
In Nov. 1961 I went through USAF KC-135 (Boeing 707) flight training at Walker AFB in New Mexico. One of the first Emergency Procedures covered was how to cope with runaway stabilizer trim.
Here are the steps as I remember them.
1. Grab and hold the stab trim wheel.
2. Switch off the electric trim and leave off..
3. Switch off the stabilizer axis of the autopilot and leave off.
4. Extend the handle on the stab trim handle and trim the aircraft manually.
5. Land ASAP.
6. Steps 1-4 took no more than 3-4 seconds to accomplish.
Note that due to difference in size no amount of "up elevator" will counteract a significant amount of nose down stab trim. We learned this lesson when we hd a 135 crash on landing at Loring AFB, ME in 1957.
Coll Gene Cirillo, USAF (Ret)
3
A simple flip of a switch would have prevented this crash. The switch is there to prevent this exact situation. instead the pilots continued to fight the autopilot for no reason. One would think after manual trimming 25 times to maintain level flight they would have realized the autopilot trim was not functioning correctly. Pilots are suppose to memorize the runaway trim procedure and they also have a quick reference handbook close by where the runaway trim procedure is on the first page for flight controls.
3
I'm a commercial rated pilot in the US, and from all the data presented so far this is clearly a crew mistake. The issue of 'run-away' trim is incredibly common on even the smallest two-seater aircraft with a basic autopilot. This article is not well sourced and very misleading. There are lots of causes for run-away trim and the emergency procedure is documented in every POH. It's not a "complex series of steps". The solution is always the same: disable electric trim controls. Assuming you could overpower a servo is sometimes true - but you would never use that approach to deal with this issue. Why would they fight the trim for 10 minutes using their yoke control, either with the electric trim switch or by pulling back on yoke? That's insane. As soon as you notice unwanted trim input, cut the electric trim and fly the airplane. The real story here is the lack of proper training and a flight culture where people are so use to automation they don't have basic stick and rudder skills, or how to deal with a simple error that would commonly occur in a Cessna 150.
4
We are seeing similar phenomena in the self-driving car accidents.
It's like HAL is trying to save the mission by killing the humans.
3
There is a myth that the human brain can learn how to do things by reading them in manuals. Science has repeatedly shown this to be incorrect. The only way the pilots would have been ready for this situation would be if they had been trained to handle a similar situation in a simulation - and preferably in a simulation in which they were not given any clue what was going on and had to work it out for themselves.
The manual is important for checking and reference. In an emergency, it might still help. But just imagine for a moment that you were one of these unfortunate pilots. Do you think you'd have found time to read the manual? Do you think you'd even be able to tell, from the crazy things that were happening, even what section of the manual to look in?
So the fact that the new manual contained the right procedure is almost irrelevant unless the pilots have actively gone through a simulation in which they practiced for this exact scenario.
2
"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.”"
Whether this is the sole cause of the crash awaits further investigation, but if it's even one part, it's a dreadful error. In the pilot training manuals, this sort of change should not simply be included, but headlined as just that, a significant change with life-and death consequences, particularly on take off and landing
3
Maybe the software engineers who wrote the MCAS system should consider selling hot dogs on the beach as their next career move. Can you believe someone would release aviation software that aims the aircraft into the ground and does not respond to the pilot's attempt to correct that? Can you believe someone would create code that aims the aircraft into the ground when it was at 5000 feet? Congress needs to get involved here to fully investigate the incredible disregard for aircraft safety as demonstrated so clearly here by Boeing.
4
This seems like a case of very deficient computer software. The computer had access to a myriad of valid data that would have contradicted the bad data from the single faulty angle of attack sensor. The second angle of attack sensor was providing valid data so why did the software take drastic action when the two sensors were in disagreement ? Why was all the other data ignored from a wide range of instruments that would have indicated no need to force down the nose against pilot commands ? What about the gyroscopic pitch indicators, the terrain/sea warning system, GPS data, altimeter, vertical speed, air speed and ground speed data to name just a few ? Why did the software ignore an abundance of data that would have contradicted a single bad sensor. In my opinion, both as a computer programmer and recreational pilot, the software deserves much more blame than the pilots for this tragic incident.
5
I wonder if all this automation has been carried too far. I for one feel more comfortable if a skilled pilot is at the controls than a computer. Pilots seem to lose their skills when just sitting there. Remember the Asianic Airlines crash at SFO when the pilot took over but clipped a seawall? Maybe it's time to take stock and rely on automation only when absolutely needed.
And if this happens in jetliners, where cost is no object, what will happen with automated cars? Again, there is an optimum amount of automation that should not be exceeded.
3
Those wishing to learn more about the fly-by-wire safety systems in modern commercial aircraft would do well to read William Langewiesche's "Fly by Wire: The Geese, the Glide, the Miracle on the Hudson." Obviously, it is the story of the heroic efforts of Sully and Jeff Skiles, but it also relates the story of how and - more importantly - WHY these electronic safety systems have become so important and dominant in modern airliner design.
There are several groups who have disparaged these systems for years, and initially Boeing itself was one of those. There are good and sound reasons for implementing them, and for training or re-training pilots to work with these systems. But of course, a system that makes it more difficult to pilot a plane is not a safety system.
But as the Boeing pilot Peter G stated, there is no single cause for this accident. While there is credible evidence to suggest that Boeing has been negligent in its redesign and education of its customers, there will be many contributing factors that also must be addressed before we feel safe with this airline and this airliner.
2
This horrific accident which needn't have happened seems to be due more to financial considerations than anything else. No airline should ever knowingly fly an airplane that has serious problems. This is not the first time it's happened and probably will not be the last. But passengers and airline personnel deserve better than to be on a plane that is a potential accident due to poor maintenance or the lack, AI systems that don't work properly or a combination of the two at any time.
Rather than automating everything and putting bells and whistles in for every alarm perhaps we should occasionally consider the human factor and human judgement. There are times when information is the problem or when "intuitive" is not. The former can be referred to as information overload, the latter is a design issue. Yes, we want human pilots not to have too many opportunities to make mistakes. But we don't want them forced into mistakes because the onboard systems are not working correctly.
It's not too much to ask to be able to board a plane as a passenger or airline personnel and expect that the flight will not end in disaster because of the AI or a lack of maintenance. These are things that are preventable.
3
Problem in modern aviation is that Pilots have forgotten to fly stick and throttle, it seems that most recent crashes had to do with the pilots fighting the auto pilot systems instead of just disconnecting and manually flying the plane.
1
I wonder if we'll have to mandate training for passengers in autonomous cars to regain manual control of the car - in case critical malfunction is detected, and car systems fail to recover or switch to safe state automatically.
4
I'm a glider pilot. A portion of commercial aviation mishaps occur when, for whatever reason, automation meant to assist the pilot ironically causes a loss of control. In many cases pilots don't understand that these systems are the culprit.
May I suggest a master automation disconnect so that the aircraft returns to full manual control and trust in your instruments?
Then, develop new automation, completely independent of other systems, perhaps gyro based, whose sole purpose is to return the aircraft to straight and level. An "emergency trim" device, if you will.
3
So after reading all these comments runaway trim is a comon thing on 737's ? Or does this happen on all modern aircraft? Sounds like to me there needs to be NO automatic trim whatsoever! Hand only.
1
“It’s in the manual” clearly what is in the manual did not work! These are lives not data points!
2
@Barbarra
Point taken however I think the intent is missed.
A Captain and co-pilot on the very same plane on the inbound flight the day before had the same problem as the incident crew did. They flipped the trim to manual, pulled the breaker, and used the trim wheel per the checklist/manual and I assume their training. I believe all that took 5 minutes or so before they resumed normal flight. Their passengers arrived safely on that flight on that very same plane.
The incident crew took over the next day and, facing the same issue, failed after a longer period to diagnose and bypass the problem, ultimately flying the plane into the ocean at high speed on a perfectly clear day.
Why the two separate outcomes? Same plane. Same Airline. Same problem. Yet one crew arrived safely, and the other didn't.
Not saying MCAS and the manual are great, just that this is about training and experience of the crew as much or more than the checklists.
Not to mention that the Lion Air ground crew put the aircraft back into active service when it was not properly repaired and the air crew seems to have been unaware of that.
3
I have nearly 3 million miles on commercial airlines as passenger. I read most comments of experienced pilots. A plane went down with two experienced Boeing pilots (new to the MAX model) in the cockpit. They tried to save but failed: 189 people are dead. Human training or machine design something went awfully wrong. I see who gets blamed has huge financial implications but for me what is important now is if it is safe to fly MAX model. All I want to hear therefore when such errors can be prevented 99% of time.
2
The “automated” system was based on the idea that computers are smarter than people (pilots in this case). That assumption is also why the technocrats are so excited about taking away the steering wheel, brake, and accelerator from car drivers. Autonomous airplanes anybody?
4
Smoking gun might the Consistent 20 degree variability between the two sensors. The MACS algorithm system design needs to fixed as well. I hear Boeing is working on that well.
2
From my reading of the article there was a variance of 20 degrees between the two angle of attack sensors on the plane. I’m not a pilot but I assume a 20 degree variance should have triggered alarms when the plane powered on at the terminal. I suppose we’ll know more when they find the black box if it survived the impact.
1
@Joseph Finsterwald
If the fail existed on the ground, the plane would have rejected take off. Would not have allowed the plane to progress beyond V1 without aborting
Looking at the climb out data and overlaying previous cycles of the same aircraft, it appears that the variance and error was picked up between 15 and 25 seconds after V2 when the plane was airborne. Thats seems like when it went into an anomalous flight pattern.
the FDR has been recovered and would show the AOA position while on ground. Its the CVR that is still missing.
Occasionally hiring a competent team of lawyers will always be less expensive than perennially hiring teams of competent pilots, training instructors, ground crews and safety engineers. Corporate malfeasance rarely, if ever, translates into jail time. Perhaps we should rethink that?
9
As a layman (to the art of piloting and flying aircraft) I'm struck by the sense that in the attempt by industry to automate as highly as possible their aviation machines - all with the noble goal of enhancing safety - they have inadvertently created massively complex Rube Goldberg flying contraptions that are more unsafe than not. One with humans sitting at the center. Those humans are supposedly in control, but in truth it seems not much more than a monkey in a can throwing switches and watching for results.
How is it ever possible for them to always react appropriately - oft-times in seconds - when something unexpected goes wrong with the machine they are flying, eh? I daresay this is not so much a training (of the human) issue as it is a screaming violation of a basic engineering precept. The K.I.S.S. principle. Keep It Simple, Stupid.
This should be engraved in the minds of anyone designing such, or any such, machine. From planes to trains to automobiles (heh), a return to that basic principle seems in order doesn't it? Boeing's "ooops here's a change noted on page 327 of the manual" client broadcast update certainly doesn't cut it.
John~
American Net'Zen
7
I wiil refuse to fly the B737 Max plane until the investigation is complete and a simpler pilot action can override the MCAS rather than this current complex multi-step procedure that Boeing just now revealed. I'll fly older versions of the B737 and Airbus planes. Delta is the only large US carrier that did not order the B737 Max. Hello Delta.
4
The automated system that cause the nose to want to stay down is a computer. The computer caused the plane to crash along with pilot error for not disabled the computer. When fly by wire was first introduced this was always the greatest fear and has come true unfortunately.
4
I read another report that pointed to pior flights of this same aircraft that had similar issues that were not properly fitted during maintenance of a sensor that positions the airplanes nose in a stall condition.
The other report mentioned the aircraft should have been taken on a test flight without passengers before flying regular scheduled flights. Sounded like Lion Air had major safety issues.
7
It seems to me that Boeing is completely missing the point and is being tone deaf on this issue. The MCAS should be not receiving and reacting to false positives. If they knew of this problem, and from their reaction it seems they did, why did they not fix it?
12
This incidence forces us to think of how close are humans to the future partnering with artificial intelligence (A.I.) that we ultimately may not be able to control?
I recall seeing the 1968 classic science fiction master piece “2001: A Space Odyssey" and the chilling feeling of the Movie audiences hearing the calmly intoned and ominous words , spoken by a spaceship's intelligent computer named HAL 9000 confirming that it could think for itself, and that it was prepared to terminate the astronauts who were planning to deactivate it.
HAL 9000 learned from observing its environment, watching and analyzing the words, facial expressions and movements of the human astronauts on the spaceship. It was responsible for performing rote functions such as maintaining the spaceship, but as a "thinking" computer, HAL also was capable of responding conversationally to the astronauts.
When the mission goes awry and the astronauts decide to shut HAL down, the AI discovers their plot by lip-reading. HAL arrives at a new conclusion that wasn't part of its original programming, deciding to save itself by systematically killing off the people onboard.
The prospect of AI doing more harm than good may not be that farfetched. Experts suggest that weaponized AI could play a big part in future global conflicts, and the late physicist Stephen Hawking suggested that humanity might soon find AI to be the biggest threat to our survival.
9
In all of my years of flying both commercial and military aircraft I have seen well trained crews in both communities crash because they simply misunderstood the cues the aircraft was telling them or it could be a design problem that was only unique to this particular model of aircraft. If this aircraft's trim system had previous writes ups or had the angle of attack vanes swapped out the crew would have known about this as part of their pre-flight duties.
Finally, I have seen many examples where despite the synergy of having two pilots in the cockpit the ability to analyze and solve the problem as a crew was greatly impaired by many things fatigue, fear, and task saturation. Bear in mind, that in 1908 the first passenger killed in a powered flight, Lt. Selfridge flew with the most experienced pilot on the planet at the time, Orville Wright.
My condolences to the families and friends of those who were lost on Lion Air Flight 610.
3
Isn't AI wonderful?
3
This grusome tragedy felt weirdly familiar as I read of the aircraft's automatic "trimming" system going amok, but I couldn't place where I'd experienced something vaguely similar.
Then I suddenly realized: Boeing, like many contemporary computer softwares, has abandoned intuitiveness as a priority, instead expecting users to re-condition themselves to their own artificial logic & "think Boeing." Like Windows 10, much of Microsoft Office, & other systems, the user is supposed to learn them as one would a language (or riding a bike, or playing an instrument), a process of constant reinforcement & repetition till one - or at least a substantial part of one's brain - is rewired to process tasks according to designers' schemes. Especially with the dozens or hundreds of inscrutable icons...
When these instrumentalities expect your brain re-wired to their specs, this can be efficient for people willing & able to give their brains over to the systems' architects. But for others, these systems seem deliberately, almost spitefully opaque, as if designed to punish those who fail to fully assimilate to the machinery.
In the Boeing case, rather than making the process of disengaging the automatic trimming obvious, it required subtle knowledge that had to be trained. In a similar vein, I was on Amazon.com on another's computer, buying something. When I wanted to logout afterward, the command was nowhere to be found, & required scrolling down menus. No mystery why: They like us captive.
13
Why, for example, was the "smart" trimming system not programmed to detect the plane's violent convulsions and either shut itself down on its own, or in some thoroughly obvious way signal the pilot to shut the process down (including a single simple step to do so) and adjust/trim manually??? That would be too simple, that's why-- boringly straightforward and skill-less, insufficiently "technical" for the pilots.
4
Deregulation is the answer. We need less health, safety, environmental and financial protection so that our corporations and their plutocratic owners can unilaterally decide what is safe and what is fair, and who lives and who dies.
8
How can any system force a nosedive of a plane that’s only 3000 feet high (when the MCAS first kicked in) when there is literally no reason for doing so, since there was only one faulty sensor, easily identifiable as such for any computer, since the angle of attack reading was already off when the plane taxied to the runway and at least when the plane was travelling at 3000 feet with 450 mph.
Even a well trained pilot would have trouble to manually turn the trimweel 50! rounds (from max nosedown) to level the plane in time when MCAS kicks in at such low heights.
I don’t know if the airbus approach is any better, but at least they give the command back to the pilot (alternate law), if there is any invalid sensor data, which is common practice for any automation. Here we have the complete opposite, which for a manufacturer is obviously a strange move, since the liability with this design always stays with him.
13
Look at the photo. The switches to disable the stab trim are directly forward of the number two engine fire bottle switch on the center pedestal. They couldn't possibly be easier to locate and operate. The pilots never selected off, they spent 11 minutes fighting the aircraft instead. Literally all they had to do was throw two switches. No, they cannot reset themselves.
2
As someone that maintained transport category aircraft for major airlines (Braniff, American) for 40 years the level of ignorance in most of these comments is amazing.
6
Good design anticipates what a normal human being would do, and in the case of pilots, what they have been doing for decades. There needs to be a very simple single button that deactivates all autopilot features. Boeing needs to get some Apple designers to simplify the interface between man and the machine.
9
Boeing is not doing itself any favours by saying it's in the manual. Of course, the pilot should have read the manual, he obviously didn't or panicked and now 189 souls are lost. This story also needs some quotes from experienced pilots to help ease public fear if Boeing is going to be pedantic. 737 is a workhorse of a plane and many fly it or have flown in it. We need to know Hal is not in charge.
7
Every properly trained Boeing pilot knows how to remedy the problem the Lion Air pilots encountered. For any comptent pilot, the first step is intuitive: simply flick the trim switches that sit on the control wheel under your thumb. It is intuitive because you use these switches whenever pressure builds up on the control wheel. That intuitive step temporarily stops the unwanted movement of the stabilizer. The second step is to turn off the power so that unwanted movement cannot return. Manually reposition the stabilizer if desired.
This procedure is nothing new. I learned the procedure back in 1967 when in training to fly Boeing 707s at Pan Am.
An airline that doesn't train its pilots to perform this basic intervention cannot legitimatlly call itself an airline. But instead of owning up to its failure, the airline is trying to swift-boat Boeing.
Unfortunately, reporters covering the story are not experienced enough to recognize it whe they are being deceived and used.
13
@Capt Tom Bunn LCSW
This doesn't explain why the onboard systems can detect (erroneously in this case) a 20-degree angle of attack and decide to push the nose down, but they are unable to detect that they have now pushed the plane into a severe dive and it is about to crash! An altimeter spinning down towards zero should be a clue, if nothing else seems to be working!
Blaming it on inadequate training won't cut it. This system seems to be grossly misdesigned. Trying to avoid an imaginary catastrophe by causing a real one does not sound like a good design principle.
11
@Capt Tom Bunn LCSW
The 707 was not fly by wire old timer, you had cables telling the flight control surface what to do. I’d love to hear some more stories about the good old days.
2
@Mike The 737 (even the 737 Max) is not really fly by wire. Unlike on an Airbus A320 or 777 (both true fly by wire), the 737 has mechanical connections between the flight controls and surfaces that even in a complete loss of power can move the surfaces. Usually the hydraulic pumps and computers assist with moving the surfaces, but if all fails, but you can muscle a 737. This is also why the 737 does not have or need a ram air turbine for emergency power.
1
Cockpit voice recorder will reveal the following msg from the flight management computer:
"I'm sorry Dave, I'm afraid I can't do that"
7
I, for one, am leery about welcoming our robot overlords.
11
I read the article about the MCAS system ten days ago, and it seemed to have flown under the radar. There were only 55 comments. When I read it, I was horrified to learn the pilots may not have been trained in the proper response to runaway trim. In any case, people seem to be picking up on the story. This article has some insight from experienced pros. I had wanted to ask an old C-130 pilot acquaintance what he thought, but I think I get it... The pilots weren't and aren't trained. And lack of training among people who hold the lives of hundreds of people makes commercial flying a less salient prospect, of a sudden.
6
What is the point of a system to prevent stalls, which are most dangerous near the ground, that cannot be easily overridden in time to prevent an unintended impact?
12
The plane has had the same problem and they still let it take off with passengers. This tragedy is on the airline more than the manufacture.
10
I am an Electronic Technician and the first thing I thought was, why they didn't kill electrical power, sure enough as I read further, one of the troubleshooting steps was to kill electrical power. I guess, I might just know a little how Electronics work.
8
M.C.A.S. - they wanted to call it HAL but the name was taken.
I don't fly. Waiting for those self-driving cars here down on land.
6
Could the recently-replaced sensor been a counterfeit unit, not working properly? Perhaps this accident will finally bring to light this huge problem in the aircraft maintenance supply chain.
4
Boeing - The Best There Is.
The best technology company on the planet - real tech not digital tech - could not provide a fool proof intelligent control system. It became vulnerable to catastrophe.
Now does this mean?
It could mean that our tech companies should just try harder. And companies such as Boeing will get it right next time. Of course they will - next time.
How about now?
Right now - it means any intelligent machine is vulnerable to catastrophic failure. The best company on planet earth could not prevent it.
5
The pilots fought honorably for the lives of everyone. No shame of any sort is due them.
11
Actually, they were doing the wrong procedure rather than the correct one that was already in the manual for earlier systems regardless of the new system. If they knew the manual, they would’ve known how to fix the problem.
7
Chalk it up to (very) bad pilot training. Shame on both the airline AND Boeing for not insisting on this.
6
Re the Boeing bashing: this airplane should not have been in the air. It should have been in the hanger with technicians crawling all over it to make it airworthy.
15
The assertion here is that the pilots "fought" the autopilot system for 11 minutes without ever turning it off and flying the airplane manually.
Were the people in the front of the airplane pilots or passengers?
2
They were *not* fighting the autopilot. The MCAS system operates *only* in manual flight mode.
5
I am 26 year veteran airline pilot with thousands of hours experience flying Boeing jets. This accident resulted from multiple failures from mechanics to designers to pilots. Boiling this down to MCAS is oversimplification. NTSB always determines contributing factors and it seems likely the MCAS will be one of those. But also you must consider the confusion as the airplane warns pilots it stalling when in fact it is not. Stall warnings involve stick shakers, audible alerts and visual lights and displays. The demand for your attention is overwhelming and the task saturation is high. Tunnel vision restricts your awareness of additional problems. The aircraft behaved inconsistently when flaps were raised and lowered. What was consistent however was the trim system that was moving without pilot input. That runway trim should have been stopped intuitively not by pulling circuit breakers but by switching off the electric trim motors using the 2 cutout switches on the pedestal or physically grabbing the trim wheel by your knee (not ankle). Doing this would have stopped the MCAS from wrongly trimming the airplane to lower the nose. The pilots could then have manually trimmed the airplane to execute a safe landing just as those who flew the plane on the prior flight. Minimal training and experience, cost cutting by management, and reliance on automation will all be additional contributing factors. Ultimately we must wait for the full investigation to know what happened.
191
@Peter G:
All of your comment makes solid sense, Peter. I would point out, though, that the crew of the accident flight may well have had trouble recognizing a runaway stabilizer condition, because it occurred in manual flight and because using the column trim switches would have interrupted the MCAS operation, which would have then reinitiated a few seconds later. Also, of course, as you note, the startle factor and task overloading must have been major contributors to crew confusion.
As for manual trimming after operating the cutout switches, I'd guess that success would depend upon how far nose-down the airplane was trimmed, and what its actual attitude and airspeed were, when when manual trimming was initiated. That task could require a lot of spinning, against significant loads -- at only 5K feet.
23
@Peter G Peter in your opinion was this a case of the pilots failing to continually keep on top of /familiarize themselves, hiring pilots without the deep experience and knowledge you clearly have or boeing burying in important change in an appendix not realizing the possibly for disaster
5
@Peter G Curious about one thing...the article says this tech was new to the 737, but were the presumably large trim wheels right by your knee also present in the older version of 737? And if not, then wouldn't this new, very intrusive (?) item in the cockpit have made pilots fully aware of this new technology, regardless of where in the manual the new procedures were mentioned?
3
I am an avionics systems engineer. This MCAS system is just the sort of thing my coworkers and I develop. We pride ourselves on our work because we truly take the safety of the flying public to heart in developing our systems. To us, we realize that "the blue screen of death" really could live up to its name, so we rigorously test our software. So much so, that the IT industry should adopt our methods and standards, such as RTCA DO-178 and DO-254.
Yet, my coworkers and I are upset about this. (We, personally, did not develop this MCAS system.) But now I am angry at discovering the root cause is ultimately management. Some MCAS managers somewhere made decisions to go cheap on documentation and communication to save money and look good. Lion Air management is equally culpable for their criminally negligent homicidal behavior.
Instead of saving money, we have (again!) deaths, lawsuits, and ruined careers.
My definition of management for about 30 years:
"Management is what gets space shuttles blown-up".
... and now aircraft.
Instead of obsessing about program engineering metrics such as cost risk and schedule risk, maybe every MBA should be required to study reliability and safety and read the Roger's Commission Report "Appendix F - Personal observations on the reliability of the Shuttle" by Richard Feynman which ends with "For a successful technology, reality must take precedence over public relations, for nature cannot be fooled."
28
Boeing should just get its checkbook out now, save the time, energy and trouble of a thousand lawsuits.
7
Nope. The majority of the negligence will be found to be the airline management. It is a lying there that would, will and should go bankrupt
1
why couldn't the pilots immediately regain manual control with the easy click of one button on the yoke? As used to be the case with complex multi-engine aircraft (thumb activation, auto pilot off, etc...)
Seems like too many complex steps between automated and manual flying. Set up for failure....
9
I recall that part of Mr. Manafort’s plea agreement included interviews without the presence of legal representation. Does that then imply that Mr. Manafort could not discuss any facet of the interview with his lawyers? And that if he did he would be violating the terms of their agreement? If not, then why include that stipulation if it was well known that Mr. Manafort was a part of the defense pact?
Wrong story...
14
Avoid the 737 MAX until a clear explanation from Boeing is out.
14
The only relevant question is why an airliner with a malfunctioning navigation system was allowed to fly. The tragedy would not have happened if Lion Air was a responsible airline that put safety first!!!
9
Like Putin trolls, there seem to be Boeing trolls trying to obfuscate the issue here.
The real issue here is that this plane has a system that can allegedly detect when it's angle of attack is too high and thus about to cause a stall, but it's unable to detect that it's in a dangerous dive and about to crash!
The pilots apparently tried more than a dozen times to pull the nose up, while this system kept insisting a vertical dive was the correct path! The Boeing engineers here can hem and haw and try to avoid responsibility, but they're guilty of at least negligent homicide.
33
@JB yup, just google "arstechnica spacex op-ed campaign" and you'll find the "PR company" where most of these trolls work
2
Regarding pilot error or poor Lion Air training and maintenance, soon after this accident was first reported, news reports included comments from some current AA and Southwest 737 pilots who said they were unaware of the new MCAS system and wondered what other important changes Boeing put in the 737 MAX.
In any case, regardless of pilot error or not, there would seem to be a basic problem with a new "safety system" that thinks the airplane is about to stall while going 400+ mph.
20
>>> '. . . a new "safety system" that thinks the airplane is about to stall while going 400+ mph.'
Yes. You *might* be able to stall the airliner at such high speeds, but you'd probably more likely rip off the wings.
3
My skepticism of AI and ‘autonomous’ vehicles confirmed.
12
What I gleaned from an airline pilots' chat site today is bizarre. They've had sight of the black box raw data, too.
Yes, this aircraft had three eventful flights in the days leading up to the accident.
The Lion Air maintenance engineers appear to have replaced the angle of attack (AoA) sensor two days before. It's possible that the replacement sensor - which was a 'reconditioned' not a brand new unit - may ALSO have been faulty, OR, more likely, the flight issues were not caused by an AoA sensor fault at all.
Most odd of all is that this aircraft had developed an identical control problem the day before but that the issue was successfully handled by the (different) flight crew. It turns out they too were ignorant of the details of the MCAS; somehow they just muddled through and nobody seems to have thought too deeply about what happened.
It's too early to start pointing the finger of blame. I can however imagine Boeing's unease, although the inquiry is being assisted by professional accident investigators from Singapore, Australia and the US, the final, last word accident report - and the one to used for legal purposes, if any - will be written by the Indonesian TSB.
4
This 'safety system' is right up there with the air bags that kill you instead of saving you. Software has become the problem instead of an aid like we hoped it would be. Doctors and health workers spend more time with the 'system' then with patients, at the detriment to the patient's health. If the health worker doesn't complete the entries in the system, and don't forget the correct codes and parameters required, they will be penalized. We have to acknowledge the part that software and social media play in the quality of our lives, let's start thinking about what oversight of these systems looks like.
8
Follow the money. It doesn’t matter if it actually works as long as it has been “improved.” Lots of people get paid to implement new systems, often with limited input from the people who will actually be using it or little benefit to the end user. Things are sometimes rushed. Execs still get bonuses.
3
Horrible! That Boeing would not inform pilots about this change is astoundingly bad engineering management. Boeing will owe hundreds of billions for extreme negligence.
5
Ironic how a system designed to prevent another Air France 447 created another Air France 447. The law of unintended consequences. Too bad so many people had to pay with their lives.
11
the nosedive feature (where the plane dips if the nose tilts too high) was not disclosed to pilots; it follows then that they werent trained in what to do if the nose dived WITHOUT plane nose going too high).
5
Yep lets make sure we keep going on self driving cars and planes. What a joke. The next move will be those producing and selling these non fail safe devices to get
Congress to exempt them from any liability, you know, like doctors and hospitals in TX.
They will claim the machine is responsible, not themselves...........
4
It appears that undocumented, unannounced changes in automation doomed this flight.
There should be a single button that disengages all flight automation - autopilot, autoFADEC, everything - so that a pilot can immediately fly manually when needed. It's currently too complicated to return to manual flight.
14
And elites are all excited about self driving cars and drone deliveries. I don't care what anyone claims. Computers are not smart and never will be. Past claims about IBM's Watson revolutionizing the medical diagnostic business, especially cancer diagnostics and treatment, has been anything but.
4
There appear to be a lot many Boeing employees on the comments thread.
4
Design error. The AI should never fight a living breathing pilot with human lives at stake. The AI should have disenaged as soon as the pilot took control. Isn't that how most airplanes work? If they don't I am not going to fly again.
7
>>> "Isn't that how most airplanes work?"
It isn't that simple in modern airliners. In Boeings, manual control isn't too many steps away from computer-controlled flight, although the operation of the MCAS system is a significant departure from that company's traditional design philosophy -- MCAS is *only* operative in manual flight.
Airbus planes are much more "fly by wire" and full manual control is much harder to establish.
2
Interesting to see the number of 'expert engineers with pilots licenses' trashing Lion Air, the pilots and everything except Boeing. One wonders if Boeing is employing a PR firm that is using trolls to lead the comments chain.
Lets see. The article mentions that several of the sensors were faulty and had to be replaced.
Why is Boeing making so many faulty sensors?
Also here is the kicker. The Boeing fly by wire automatic thingamajig is so advanced but unfortunately, it cannot distinguish between an ascent and a descent.
And for those 'experts' with flying hours and Engineering degrees, where there is software, there are bugs. I know because I have written and am writing code for control systems like these for over 23 years.
Anyone want to take a guess that Boeing knew how bad Lion Air was in how it maintained its aircraft and trained its pilots? And yet it sold its planes to Lion Air.
Someone might get a couple of research papers on ethical business practices on this.
Finally, it seems this 'runaway trim' problem seems to be pretty well known. In cases like this why not add an emergency voice command to the pilots telling them how to exactly kill that stupid M.C.A.S system?
That this kind of thing should happen in the age of 'OK Google' is absurd.
14
The root-cause of the crash is steadily implicating Boeing's new MCAS system.
There may be concomitant causes, such as failed/poorly maintained sensors, however the primary (root) cause points to MCAS system design - which continuously overrode a human pilot's correct judgement. When a pilot has to fight against a plane "bucking" and countering his every effort, he cannot hit pause and begin reading the flight operations manual.
Boeing's assertion that flight operations manuals describe how to disable MCAS, is insufficient. If this system is capable of causing such a serious crash, Boeing should have better communicated the risks.
Boeing did insufficient risk-mitigation for such a critical system designed to over-rule a human pilot. Risk mitigation is a standard technique used in high-risk systems. Risks of sensor failure (single or multiple) were not addressed. Why was not the system designed to show a warning message (in a large red flashing box) informing the pilot of WHAT was occurring and HOW to fix it?
Automation is a good thing - it increases capabilities, safety, efficiency, etc. However in safety-critical systems, automation must be tested fully and must undergo full risk assessment/mitigation. Risks must be reduced using EVERY method available - a simple warning on a display would have averted this disaster!
1) Boeing did not use "fault-tolerant" design methods - as regulations require.
2) They did not alert airlines of the risks inherent in their new system.
8
I am not booking any flights on a 737 until the issue is thoroughly resolved and hope all flyers do the same.
4
I'm sorry, but all the focus on the MCAS is, in my view, ridiculous. That this plane flew with the same alarming issues on previous flights - and was repeatedly 'fixed' and put back into service without a proper test flight is the root cause. The previous crew flew the entire leg with the captain's stick shaker activated! Instead of grounding this airplane and sorting it out they kept on going until it was put in the hands of a crew that failed to turn off two prominently placed switches on the control pedestal. Bad training, lack of experience - both pilots and technicians- and above all bad management killed these people. Very sad chain of events.
5
OK then. This aircraft should then have been "smart" enough to notify Boeing of the past days' incidents, and Boeing should then have INSISTED that the craft be grounded until fully examined and test-flown by their experts.
If Lion Air had refused, Boeing would have a paper trail with which to defend its actions.
Bottom line: pilot re-education and software redesign remain indicated by this entire ordeal. And the airline remains at least partially culpable.
1
I share comments expressed by some engineers and flight professionals - that in general, automated systems are important - often crucial - in preventing accidents.
While nothing remotely similar - the recent NASA success on Mars was a testament to this auto mode. Under no circumstances, a human could have landed that InSight.
I worked in the oil industry - all my life - over 40 years and I am a Caltech graduate.
Even in our industry where brawns often are supposed to triumph over brains - our latest technkogy uses a lot of auto mode. For example, keeping a deepwater vessel operating in 2 miles of depth - pinpoint location - regardless of seas, waves etc.
None of our highly trained marine folks could humanly do this.
For that matter - the Macondo incident also pointed to human failures.
These well meaning but wrongly acting men - some brainy and some just brawny - drew wrong conclusions from data they were receiving. And when push came to shove - disengaged or over rode an auto system to disconnect doomed Deepwater Horizon.
As an engineer - it's hard for me to believe that Boeing would design a system that actually leads to a complete loss of aircraft - not to mention human lives.
I wish this report had told us attempts to retrieve these Angle of attack sensors. They may tell a whole different story.
2
>>> "I wish this report had told us attempts to retrieve these Angle of attack sensors."
It is very unlikely that the AOA sensors could have survived impact in a condition that would make adequate testing possible. They are, necessarily, very exposed and the impact with the water, at that speed and angle, was massively destructive.
The FDR should have plenty of data from multiple sensors to establish an AOA sensor failure. And the behavior of the MCAS system makes it extremely likely that there was such a failure.
4
From my training as an aeronautical engineer, Navy attack pilot (20 yrs) and 21 year airline pilot, I have a few pertinent observations. Gone are the days when the pilot was in sole control of an aircraft. Computer massaging of digital flight control systems have replaced skill as the prime guarantor of flight safety. While this sounds like progress, it has many downsides. The interaction of flight envelope protections, failure modes, and crew response has changed profoundly in the past 10 years. In many cases the response to abnormal system failures has changed from "aviate, navigate, communicate" to a head-down computer entry/switchology drill. There is SO much data from so many different sources that the crew must correctly interpret to effect a successful outcome. Mistakes, however minor, have dire consequences. I'm NOT suggesting we revert back to the past. I AM suggesting that ALL computer flight management systems be disconnectable with the push of ONE button on the control yoke (or stick). When in doubt, ONE button turns it back into conventional aircraft.
13
@James Cooper
Exactly! Thank-You
6
>>> "I AM suggesting that ALL computer flight management systems be disconnectable with the push of ONE button on the control yoke (or stick). When in doubt, ONE button turns it back into conventional aircraft."
It's hard to think of a better prescription that that, at least with respect to control of thrust and the control surfaces. We need to add, however, that flight crews need to be proficient and practiced in hand-flying their airplanes. As an increasing number of crashes and incidents have been telling us, far too many commercial aviators are not.
7
Not a pilot, but this "enhanced" system apparently took away the ability to pull up on the stick to override the auto-trim. Based on years of accident evaluations (in the US) we all learned that this behavior was intuitive for trained pilots, and historically consistent. We then were tutored over "wild shear" warnings and pilot training, $$ for doppler radar. Airbus "fly by wire" designs overrode pilot control and killed. Failures in systems that prevent or countermand universal fail-safe manual flight control demands regulator scrutiny. Pilot & crew training plus maintenance and airline acceptance required before anyone put at risk. This 1M passenger thinks any first go-to step that requires manual leg operated trim wheels (without powered assist?) bizarre and risky: pilots should stay home. Likewise requiring breaker pulls disabling the entire system. Flying with hundreds of lives at risk should never involve automated systems that fight pilots trying to save lives. No pilot will confront identical emergencies but these pros weren't "afraid to land leaving autopilot to drive," they were fighting a system, even if better understood, that killed everyone on that aircraft. Ground these worldwide until we know more.
84
@marksjc In 2017 there were no fatal passenger jet crashes anywhere in the world. Before the current batch of automation, planes used to crash as often as several times a year, just in the US. Automation works very well and saves thousands of lives every year.
5
What you say is true, John, *and* it is increasingly clear that the MCAS system on the 737 MAX is a threat to passenger and crew safety and is much too little understood by the crews flying those airplanes.
It would be perfectly reasonable to decline to fly the aircraft until the flight crew have a thorough understanding of MCAS and its potential failure modes and, and have has adequate experience dealing with them in the simulators (which very likely will require software upgrades from the manufacturer to permit that training).
10
@marksjc "Flying with hundreds of lives at risk should never involve automated systems that fight pilots trying to save lives." It's not nearly that simple. Sometimes the automated systems prevent pilots from doing something stupid, or they try to. AF447 was lost because a temporary icing of the pitot tubes caused the flight computer to turn the airplane over to the pilots, with all flight envelope protections removed. The PF, apparently in a panic, tried to climb from 35,000 feet in warm air and stalled the airplane. Had the protections been in place, his crazy inputs to the system would have been disallowed. The Dash-8 crash in Buffalo (actually Clarence) was similar. Pilot inattention during the landing sequence led to a near stall. The stall protection system went off and pushed the nose down to gain airspeed. The PF panicked and over-rode it by pulling back on the yoke, exactly the wrong thing. The co-pilot exacerbated the situation by retracting the flaps, making stall avoidance more difficult.
But you certainly raise a point of contention in the industry and a difference in philosophy between Boeing and Airbus.
10
This seems to be a ridiculous system design. And from many reports the counter-intuitive response to this type of failure was not communicated to the pilots of the new 737.
What sort of auto-pilot system built to prevent stalling does not take into account repeated pilot maneuvers to reverse the computer-generated nose dive? And why doesn't the software calculate that the plane is approaching zero altitude at 450 mph and let the pilots fly the plane?
The plane should not have taken off to begin with if the sensors were giving faulty data. But this is a fact of life: sensors can develop faults in flight, they could report erroneous data or simply fail.
This was an avoidable tragedy and I suspect blame lies with the airline, the pilots (for taking off to begin with), and with Boeing.
4
far be it from me to defend Boeing and esp new systems but this is clearly pilot error and not just bc they werent trained on the new system or they didnt read (or understand) what the POH said.
I refuse to believe Boeing designed a robo system that takes control from a pilot based on a single input, here nose pitch. I'm familiar with design of a particular "expert system" used in F-16s that, while used for something else i cant elaborate, makes sure that before control surfaces or weapons systems are allowed to make certain decisions, they must take input from 2+ sensors to assess false positives/true negatives etc.
and esp if pilots dont know how to override the system. and it will be exposed at trial exactly what the software asks.
these systems are just automated methods to do what any pilot does when flying instruments; constant cross checking to ensure the aircraft is doing what you think bc you're flying w/o looking outside the cockpit with visual flight rules.
FAA isnt shy about putting out emergency bulletins requiring every pilot so type-rated to reread the specific part of POH dealing with this and could even require sim training before turning them loose again.
flying is safer than taking a bath. lets keep it that way.....
2
Firs do not harm
@Why not Yeah but those oaks, watch out!
3
@Atleon says that MCAS was documented in the manual. @Douglas responds that it was not. This a factual dispute and the NYT should be able to to tell us which is the case.
4
The crucial link in this chain of events seems to be the pilots not cutting power to the automatic trim system, which is a simple flick of a switch. That alone would have saved the plane (and my guess is that it's exactly what the crews on previous flights in this aircraft had done, who reportedly experienced similar problems).
The big unresolved question is why they didn't do that. Boeing says it's in the manual, but flight manuals are a more complex thing than the layperson would think. Every airline is responsible for creating its own operating manuals, and pilots don't read through manuals during flight emergencies - they perform memory items based on the indications they're receiving from the aircraft (tasks they've been trained to do from memory in specific situations) and then read and accomplish the appropriate emergency checklist. These are also prepared by the airline.
So pilots need to be properly trained on memory items and checklists, and those checklists need to be correct and up to date. Moreover, aircraft systems need to be taught to pilots in advance so they know how to even identify which memory items and checklists to follow. If a pilot misidentifies a problem, he/she will follow the wrong checklist, like a surgeon operating on the wrong organ.
So the question as to where the fault ultimately lies is unresolved. We know that even if MCAS played a role, it was possible to save the plane. We don't know why that didn't happen, or who is responsible for it.
2
This is the most pertinent comment and an extremely important question: "Much remains unknown about the doomed flight, including why a plane that had encountered problems with the sensors was permitted to fly in the first place."
15
In the last twenty plus years of aviation history, several air disasters have happened when pilots have been bombarded with incorrect information from artificial intelligence. In some cases, the pitot tubes have been frozen of inadvertently covered in maintenance. The computer only responds to binary data, not the reality that the pilots see.
AI has assisted in many ways, but this incident reveals the ethics of AI. Any life threatening AI must have a mechanism to disengage immediately with a one button hardware switch and or simple one word voice command.
Furthermore, those designing AI need to get a wider array of potential users to test the AI in a war of human perception and AI mandates to determine when AI should voluntarily surrender to human will.
Could ground radar have provided mili-second data and change in height to the ground that would have told AI it was wrong and instead surrender to human input, especially when human input clearly and repeatedly conflicted with the AI mandates?
Can AI determine in simulation who is more frequently correct in conflicting scenarios to make surrender to human will immediate?
These seem to be ethics questions that should be put in the design of any system before being put in place. To ask the question now as to who is right and who is wrong will most likely favor the corporate interest of keeping a flawed plane flying and the default conclusion of pilot error.
122
@Ben
It is not necessarily "AI", just an automated decision and control process that is run by computers. The kind that existed for many decades before Siri.
The problem is dumb over-reliance on automated decision making. The kind that assumes sensor inputs are always correct. A lot of the accidents that were made worse by automation involved some faulty sensor.
A pilot's "natural" intelligence is the only factor that can identify failures that were simply not taken into account in the design of the automated system. Overriding the pilot is a bad, bad idea.
10
@Ben. Just part of the bigger question of how much autonomy and decision making humans are going to relinquish to machines. The trend is toward more not less.
1
@Ben — Sensors aren't AI. If a human leaves a pitot tube covered, the pilot will get the same bad data that the flight computer will.
And plenty of pilots have trusted their own reality and flown a perfectly functional plane straight into a mountain.
2
the replacement sensor was either put upside down, or the leads installed incorrectly , so the on board computer was given information that down was up.
1
There are two sensors; the data from one did not match the data from the other by a wide margin. Yet, it does not appear from the reporting that the discrepancy triggered an alert or an automatic cutoff of a critical system. Had either been available, the outcome might have been better.
4
@ANDREW
You would think the software would detect this disparity and fault. Maybe it does?
Let this be a warning to all that endorse Artificial Intelligence, Robotics, and self-driving cars as our ‘immediate future’. While the promise is there, the technology is not.
While the NYT has reported on problems with this technology, I only hope you apply vigor to how well suited it is for widespread deployment. Boeing is an easy visible target and they deserve to be scrutinized over this. Please apply the same vigor to the other applications.
5
Actually, the technology IS there in many ways and is markedly safer than the typical human. Based on the article, this appears to be a multi-faceted failure involving training, documentation, and perhaps maintenance. You know, the "boring" stuff..which turns out to be pretty important.
Every post by a pilot or some other knowledgeable person in here that lays blame on boeing, which really is the case as malfunctioning software brought that plane down , there are hordes of replies that attempt to deflect blame from Boeing. Very Putin-bot-esque, I says.
7
All our appliances are idiot-proof but not an airliner? Especially if some combination of failures and obviously terribly written code can lead the plane to essentially commit suicide in the most relentlessly considered fashion, in which case it would need to be idiot-proof in panic mode?
2
It seems there was one computer system, at least, that 'knew' things were going wrong: the black box.
I foresee lawsuits against Boeing over this 'accident'.
5
Mechanical hydraulic issue with elevator. Senses working fine thus explains the continuous over correcting.... Automative system and manual inputs could never resolve.
Just a armchair theory...
Does not add up
If the sensors were an issue, why did the plane fly?
Seems like a red herring blaming Boeing with an airline that has a terrible safety record and files criminal complaints against pilots who highlight issues
As a pilot I find it absolutely mind boggling that a proven simple procedure to disable the stall system was not communicated and practiced as part of the qualification to fly the new Boeing. Unconscionable. Why would this not be part of simulator training? How did Boeing not evaluate and identify the severity of a sensor failure with such a system especially at a low altitude?
10
I believe it boils down to two questions: 1. Were the pilots of the plane aware of the override and knew how to use it?
2. If they knew how to override the MCAS, did they try to do so, and, if yes, why didn't the override work?
The answers to these questions, which I am sure are currently being pursued, will tell us, and especially those who lost loved ones in this tragedy, who was at fault.
Related, but separate from this, an additional feature in any type of MACS would be desirable, if it doesn't already exist. Specifically, the unit should record any occurrence in which its action were being overruled by the human crew, so that the exact circumstances of those situations are available for study. Lastly, this disaster shows once again that decision making (machine or human) is utterly dependent on the quality of the relevant information - garbage sensor reading in means garbage decision out.
2
The pilots couldn't turn the automatic system off? Or didn't know how to turn it off? Sensors aren't always going to work properly.
1
All the focus seems to be on Boeing. If this were a US crash, we would wait 9 months for the NTSB to complete their investigation before we start pointing the blame cannon.
I want to know where the replacement AOA sensor came from. I read one story that Lion air could fly a jet for 10 days with a broken weather radar, and on the tenth day they would swap the (broken) weather radar into another plane. 10-days to repair secondary flight instrmenets wasn't meant too allow floating bad parts around the fleet!!!
If Lion's maintenance was really that bad, I want to know:
1) Why did a 3-month old AOA sensor fail on a new 737 (or did they swap one to fix another aircraft).
2) Where did the replacement unit come from? Who certified it as flight worthy (Hint: it wasn't the FAA).
3) The early reports talked about a failed air-speed indicator. How was that fixed?
4) The flight controls on the 737-MAX should clearly be able to manage a single failed sensor. If one AOA was 20 degrees out of sync with the passenger side, the flight controls should use other inputs (like air speed that maybe was also faulted) to figure out which sensor failed.
I suspect some combination of "borrowing" good parts from a new plane, replacing parts with bad parts, and maintenance changing the wrong part will be the majority of the causal factors. The MCAS is the final trigger in a much larger causal chain.
3
I've only piloted a Cessna but the trim in that plane is only a correction to slightly adjust a tendency for a plane to fly nose up or down, which is easily counteracted by the larger movements of the yoke. It's all mechanical control. Maybe the investigators are afraid to publicly state a theory that the plane's flight control computer was hacked or compromised.
1
The automatic control relied on a _single_ sensor input. This is a serious oversight in design with no redundancy. When that single sensor failed, the automatic system brought down the airplane.
Human pilots rely on a multitude of sensors and inputs. There is redundancy and therefore safety. This is why there are two pilots - redundancy = safety.
All automatic flight systems should have some form of redundancy designed in. When two (or more) sensors disagree, warn the pilot(s), then gracefully (and safely) cease automatic operation.
3
I'm not a pilot. I do financial spreadsheets for a living. And I always include crosschecks between various values, with bold red error messages in bright yellow boxes that show up when there are hidden errors. In this way I'm aiming to avoid causing financial harm to my clients.
Apparently it didn't occur to anyone at Boeing to similarly crosscheck the data output of the 2 sensors, with an obvious alert to the pilots and a clear method to immediately take corrective action when there is a problem.
Unbelievable.
7
The procedure that Boeing specified for disabling the MCAS system is dangerously complicated, and much more complicated than the previous directive in the pilot's handbook. At the point where the system fails, pilots are going to be in an emergency situation and need a simple way to regain control. Clearly, this was a problem that had been anticipated and should have been engineered better.
9
On one hand, Boeing made an automated system that is designed to prevent disaster caused by human error, but fails to create an automated system that is design to prevent disaster caused system error (while counting on the human to do this instead). There is a lesson in design to be learned here.
On the other hand, there is a system override and the crew had some time to engage it but failed to do so. Given the importance of such a function, I cannot imagine that it requires any kind of demanding physical acrobatics to achieve if the pilots are well aware of it. If the pilots did not try to do this or did not know where to find the override, then it's likely an issue with training.
I am sure why we are so eager to pick a side on this issue. It's not a human vs. machine question. At least right now, both are clearly needed to make things work well.
1
Doth Boeing protest too much, instead of owning up? Oh yeah, protect the share holders.....
" . . .Boeing has said that the proper steps for pulling out of an incorrect activation of the system were already in flight manuals, so there was no need to detail this specific system in the new 737 jet. In a statement on Tuesday, Boeing said it could not discuss .... "
Could not or would not discuss ?.....
6
A test flight without passengers is required when repairs are made to flight control systems.
3
Required by whom?
The fighter jets I worked in had two independent systems. Pitot tubes are known to be unreliable or to become blocked. A jet must be engineered to a worst case scenario, not a best case.
3
The article refers to a flight control system that "automatically pushes down the nose of an aircraft that is approaching an aerodynamic stall . . ." In actuality it should read the system "automatically pushes down the nose of an aircraft that A COMPUTER THINKS is approaching an aerodynamic stall . . ." Basic flight training includes pilots recognizing an impending stall and recovering from it, training a computer doesn't have. A computer responds to sensor inputs and acts IAW its programming. If the sensor data is wrong the flight control computer mindlessly commands the autopilot to maneuver incorrectly. Pilots, at least in my experience, can instantly recognize when an autopilot malfunctions by being discordant with the flight instruments and immediate disconnect it. I guess when you rely on a computer to control the autopilot it's not so simple.
4
It sounds like the MCAS is an add-on system that is not integrated with the rest of the airplane's instrumentation and controls. If the pilots can recognize that something is forcing the airplane into a dangerous dive, why can't the airplane itself recognize that? And, furthermore, recognize more quickly than the pilots that the MCAS is doing it? It sounds like the airplane has been "told" to assume the MCAS is infallible.
13
In reading the comments, I have to agree with those that speak of the "improved"or changed procedures to solve a flight problem as possibly complicating the situation rather then helping it. If it's not broken don't fix it.
I myself have experienced wind shear while my plane was trying to land at Kennedy and bucked like a bronco, the nose of the plane trying to be forced down as the pilots kept pulling up, it was just terrifying. They finally aborted the landing and went back up. I thank my lucky stars everything turned out okay, and I'm sure that there are plenty of these types of stories we never hear about because nothing tragic happened, that in the vast majority of times in a flight emergency the corresponding corrective procedures work!
3
Yep, we will soon have self driving cars!
5
When possible, I try to check what planes I would be flying when booking flights, to avoid models who have had this sort of automation problem. The new 737 will be added to my list.
3
same here! I've advised friends and loved ones to do the same, and sent out the world rankings for airlines.
1
That a 'safety' system can command a plane to do something fatal is wrong. Does my home computer not have a bug in it? Does my truck always start? No. I know of know system that is always and invariably perfect, mechanical or software.
Once the AOA system was disconnected from the yoke, it became wrong. It became a killer.
It is not more training of pilots that is required. It is more training of aircraft engineers that is required. Engineers are not pilots. This is a bad system. I for one shall not fly on any 737 Max. Period
5
Also not a pilot... Apparently the new automatic controls to away the ability of override
1
Just say no to autonomous vehicles.
3
I am not a pilot or an engineer, but I ask a question. The automatic response by the aircraft is designed to counter human error to avoid a stall. Why not have multiple independent sensors that feed into several computers? If the data from the different sensors and the interpretation by the different computers doesn’t match then the system needs to back off and communicate that to the pilot who can then choose what set of data to believe rather than becoming an rogue force.
Also would continuous satellite monitoring of position , speed and altitude help give trustworthy data to the pilot and computer? Can a pilot call on a satellite to track it when needed if it’s unrealistic to track so many flights all the time?
2
@Mickey
Hello Mickey, I like your idea that differing data,, a self check system, should be an alert. Maybe better,, inconsistent data,, the system should auto-shut down.
Your other idea of a remote electronic sensor,,, likely does not need to track multiples,,, just a system that uses GPS data to give a readout to the pilot. Aircraft-centric. It could be done today,, my reluctance would only be that with GPS,, a remote source,, in orbit,, any data you receive is old news,, it tells you where you were, by 200 or so meters.
The system on all 737s prior to this newest Max was always,, yank back on the yoke hard enough and the system would disengage,,, which if the plane is tryting to make an unplanned dive,, you'd be yanking back pretty durn hard.
1
@Mickey -you are dead on. know that these systems have existed for 30 years in many fighter jets weapons systems targeting packages and prob nav systems too. plus you obviously train pilots how to override any automated system and they werent it seems.
the idea is in a dogfight, a minimum threshold for a weapon to successfully hit a target must be met by the system assessing several (not one) measures of the enemy's aircraft's actions then feeding them to the targeting system before it allows arming. the "expert systems" help ensure weapons arent wasted as can happen in heat of the moment if the pilot bases his decision on a single incorrect in put as you correctly state.
because if one (or say both) of his limited supply of weapons is wasted, the probability of surviving the day goes down drastically.
It’s hard to change people’s behavior, it’s the designer’s job to ensure that systems respond the way users expect them to.
7
The comments on this are fascinating. Many are from experts and I say that without putting air quotes around the word. It's astonishing (and frightening) how much disagreement there is among them. Piloting an aircraft shouldn't be like medicine—a combination of art and science. It should be strictly a science. Humans went from powered flight to landing on the moon in just 66 years. You'd think that in the 49 years since that accomplishment a short hop in the morning would be free of risk.
8
it seems that the responses are from knowledgable people.
As a non-non pilot and one who has since my youth marveled at the fact that tons of steel and passengers can be airborne, I recall a flight that went down on an approach to Buffalo International a few winters ago due to icing on the wings. Young pilots with insufficient experience.
2
Common sense says that this is Boeing's fault. What I find remarkable is that almost all comments stating this fact are rebutted. I can well imagine that Boeing has their own crew monitoring this comment section.
9
@Wolf Bein
Common sense and history would indicate there's a series of failures involved here, some causational, some contributing.
The incident airframe experienced the same fault on the inbound flight. That crew seemed to have the training and experience to deal with it in 5 to 6 minutes and deliver plane and passengers in one piece. No problem.
Next crew on the same plane with same fault crashed it. Why? Why could one crew fly though it and the next one crashed?
Why was a plane with a known problem allowed back into the air after supposedly being fixed by ground crews? Was it tested? Properly?
Why didn't the next crew know about the airworthiness issues from the previous flight?
Why did a 4 to 6 month old sensor fail?
None of these things have anything to do with Boeing.
MCAS isn't the first trim control system. Experienced and trained pilots seem to know how to shut off systems. Hit the switch, pull the breaker, and hand fly the thing. Previous crew knew that on the inbound leg. Fatally, this one didn't.
I have no dog in this fight, but I can just about gurantee that the causational factors will be a malfunctioning AoA, followed by innefective ground service/repair and improperly/inadequately trained pilots. MCAS may be a contributor, but very much doubt it will be causational.
Overall causational point will be the rapid growth of the industry in far east. Not enough experienced pilots, not enough good ground crews, not enough safety culture
Nice to have heard from pilots, private and commercial (they claim), that exclaim about the engineering shortcomings.
Best hold your comments, pilots, until you hear from Boeing engineering. Otherwise you will appear foolish.
It is nearly always pilot error - nearly always. You know it and I know it.
1
Baloney. I worked on a few investigations while I was in the Air Force. They were usually electro-mechanical. I'm trying to remember who said that jets were made up of hundreds of thousands of pieces and they all have to work. This aircraft should have been able to be flown manually, if needed.
And usually flight critical systems are designed to have redundant capacity.
7
This has some parallels to the 1989 Midlands Flight 92 crash. When faced with an in flight emergency, the pilots executed procedures that would have been correct for the previous version of the 737, but were exactly the opposite of what they should have done in the new version. Investigators found that Boeing over-promised on the simplicity of pilot conversion to airlines eager to minimize costs. Crews were given brief classroom instruction on the relevant procedure, but never practiced it in a full-motion simulator. That changed after the fatal crash. I agree with other commenters who point out that Boeing needs to do more than just point to a multi-step procedure buried deep in a flight manual.
15
This is the danger and what happens if a pilot cannot disable the automatic control systems controlled by a computer and fly the plane manually. It should be a warning sign that these type of systems need an easy over ride. The fact that instructions are contained in flight manuals is of no moment. Pilots only have, in some cases, seconds to correct a situation. They do not have time to consult a flight manual.
10
We need to think, not react.
The plane thought it was in a stall and forced the nose down. The pilots couldn't override the autopilot system, and the plane crashed. Should it be easier to override?
Hmm: in the Air France 447 incident, the autopilot had been disconnected by the aircraft. The plane really WAS in a stall, and a junior pilot, in a panic, pulled the nose up when he should have been pushing it down. The plane crashed.
It's easy to say, "Push a big red button to regain manual control," but manual control was the direct cause of the crash of the Air France flight. Yes, yes: faulty sensors. But pilot error was what put them into the ocean.
Obviously, Boeing messed up, but I don't think the solution is as easy as a big red OVERRIDE button. We need to remember that air travel is much safer now thanks to these automated systems.
Maybe it's time for designers to sit down and really think about how to improve flight systems, accounting for human judgment and potential system errors.
I don't know if airlines have QA people who spend their days in a flight simulator a) trying to crash the plane and b) reacting to all sorts of possible errors. I would hope so. Do simulators have error options, like "make pitot tubes report inaccurately?" or "set off all alarms at once and let user find problem?"
Do pilots sit in simulators once a set of risky scenarios have been identified? Regular pilots will be taken by surprise more than the QA people. Etc.
6
This is extremely sad and seems on first impressions preventable before the flight and during the flight.
1
I wonder, why there is no routine in place, that automatically disables the MCAS and alerts the crew to it, if there is there is a discrepancy in the reading of 2(or more) sensors that are redundant by design.
What's the whole point of redundancy, if that basic "feature" is not implemented?
Any pilots or engineers here who can enlighten us?
5
Regardless of what procedures the pilot could use to counteract the issue, a system that does not have enough redundancy to prevent faulty readings of this type is a poorly designed system that should not be up in the air.
4
Software upgrades with their annoying glitches and counterintuitive approaches to problem solving may prove the end of civilization as we know it. But software engineers got to engineer and everyone else has to promptly adapt or find themselves banished from cyberspace forever. This reader's heart aches for those poor pilots - from this report it appears they certainly could have flown that plane safely if only it did what they told it to do. Self driving cars, anyone?
7
"Boeing has said that the proper steps for pulling out of an incorrect activation of the system were already in flight manuals, so there was no need to detail this specific system in the new 737 jet."
Such cold, detachable arrogance. This information is of no comfort or relief to the families, loved ones or friends of the 189 souls aboard Flight 610. For the latest generation of Boeing 737s to have an automatic system which received incorrect sensor readings should be justification for grounding all of these planes until the problem can be resolved completely and thoroughly. I can only imagine the terror and fear the pilots, crew and passengers felt and endured during those 11 minutes.
Profound condolences to who knew and loved these individuals.
16
Has anyone checked the passenger list to see who might have been key to a corrupt government or business? An entity with access to hackers?
Is it true that pilots couldn't flip one switch and cut off all computer-controlled flight systems and fly the plane themselves? Frightening.
2
From the very beginning, Boeing’s reaction to the incident has been dragging its feet. When earlier evidence points to the new system problem, it said little because “it is under investigation”. Now more evidence and analysis by experts further blames the failure of that particular system, Boeing still offers little and says because “it is under investigation”. I wonder if passengers now want to get on a new 737 MAX. I travel a lot and I know I will refuse to get on a 737 MAX for now without further assurance from Boeing or FAA.
3
Why didn't the pilot entering Jakarta (barely surviving then) not raise the alarm and demanding an instant stop to using that plane.
I can't help but think of the passengers, who, for 11 terrifying minutes, must have experienced the plane going down, then up, down, up, etc. etc. 11 minutes is a looong time in such a situation. My heart goes out for all of them.
16
Boeing has always engineered around the concept that no one system failure will endanger the aircraft allowing the pilots to overcome the failure and control the plane. However with the new MAX 737, the engineers have added a system, MCAS, that has an automatic activation feature not found on legacy models. Obviously, the new operators of this type were unaware of this change, which degrades the margin of safety. Boeing has a lot of explaining to do and they will have to change this system once the full implications of the crash is revealed.
12
Modern pilots are lulled into reliance on the automation even when it malfunctions. Pilots need to be reminded that stripping away ALL of the automation and hand flying the jet as if it were a Piper Cub is suitable, doable and sometimes the best method for survival.
7
Any automatic in flight system should look at all the variables to decide how to adjust the control surfaces of the plane. The fact that the plane was at 5000 feet (low) was presumably being input into the computer. In such a situation it is a mystery how a system could issue the command to nose down (dive) for almost any reason. Also, automatic systems are designed to fill in for lack of pilot input and not to countermand that input without clear warning messages. It looks like this automated system became a demon with the loss of a single sensor and refused to respond to the inputs of both the pilot and co-pilot, inputs which should have been given a higher priority than a single sensor. No matter what Boeing claims, this looks like a design flaw and not pilot error. Pilots are not hired because they can memorize manuals, they are hired because they can fly. Any plane that does not let them do that is flawed.
35
Why do we keep adding so many safety features that appear to not be needed? This adds complexity to an already complex aircraft. How often do we have crashes because a pilot pulled too hard on the control stick?
Is Boeing simply making solutions to problems that barely or don't exist in order to claim their aircraft are more "modern"?
This reminds me of my new Honda vehicle I bought last year. It has a couple of safety gadgets that are worthless and dangerous.
There is the lane departure feature: If I am on a highway that splits into two different directions, the lane departure feature thinks I am veering off the road and abruptly shakes the steering column.
There is the collision sensor: It is supposed to detect frontal collisions. If I am driving on a two way winding road and a car is coming around the curve, it thinks we are going to collide because it is front of me (line of sight). The sensor thinks we are about to collide so the vehicle flashes "BRAKE" on the dashboard and sometimes will apply the brakes abruptly.
Fortunately, I am not going 400-550+ MPH and flying 20k-30k in the sky like the folks in the doomed plane were.
17
I have only had a few hours flying an old Cessna but my experience with technology is that it is helpful but that there is a point at which it becomes too helpful. It becomes too helpful when it removes the user's final authority and I have seen this time and time again. The computerized systems should not have been able to fight with the pilots and win , period, or to require that the pilots perform obscure and unintuitive maneuvers to convince the computer to let go as it is sending the plane to its death. This is not the first time that there has been a poor interface between human and computer . Boeing gave too much power to the computer and it abused it. No error codes, no sensing unequal angle of attack sensors, no surrender to the pilot, just the pig-headedness of a machine
26
Had this airplane been owned by a reputable, top-tier, safety-first airline, would this have happened? No. They would have identified the (glaring) problem and verified the fix before rushing it back into service.
2
It is way too early to start pointing fingers and blame. Multiple causes are possible, and most likely the final report will attribute blame to several causes, not just one - whether it is Boeings fault for its implementation of the MCAS system, maintenance for failing to correct the faulty sensor, pilot error, poor pilot CRM, a lack of simulator training in a brand new model, poor safety culture, or any other factor that has not been discussed.
Ultimately, automated safety systems save lives, especially since many modern air crashes are due partially or wholly to pilot error. The people calling for decreased automation in the cockpit should look at Air France 447, where the pilots' inputs pushed the plane even further into a stall and ultimately a complete loss of life.
73
It seems to me that some expert could read the training manual and report what if anything is said about this type of incident. Yes it is early in the process but the question of whether 737 pilots have been given clear instructions is an answer we need sooner rather than later.
7
@Michael
Then you ought to volunteer to sacrifice yourself
so machines can get better faster..........
Are you a human or just another algorithm?
1
If the Boeing emergency manual dictates that there are cut off switches to disable the system and return to manual control, it would be a training issue with the flight crew. Boeing obviously added this flight characteristic to the planes for a specific reason and also documented how and what it could do after their own test flights. Wondering how many airlines are not going through the entire aircraft's training publication to understand this system.
107
>>> "Boeing obviously added this flight characteristic to the planes for a specific reason and also documented how and what it could do after their own test flights."
Boeing DID NOT document the MCAS system in the Flight Crew Operations Manual or other documentation provided to operators of the 737 MAX.
20
@Douglas SWA pilot I talk with on a regular basis says differently. Its a training issue in his opinion and simply TURNING OFF THE SYSTEM takes it out of the equation. How come no other airline is having the problem? Because they TRAINED their pilots for just this scenario...among many others.
15
According to Peter Lemme, the satellite communications expert and former Boeing engineer from the blog cited in the article, different pilots on the same plane from its penultimate flight were able to cut electrical to the stabilizers and manually control them via the wheels by their ankles. This leads me to think that perhaps Boeing's instructions for this situation were communicated but that the pilots on the accident flight may not have had enough flight hours or experience to execute this maneuver. I can't speak for the pilots but wrestling with a plane gone awry for 11 minutes is a long time.
61
@Bun Mam
I would agree with you, there are two switches, each under a hood, which will effectively disable the feature.
@Bun Mam
What astounds me is that one the plane's penultimate flight its pilots engaged in a literal fight for their and their passenger's lives. Shouldn't this hair-raising experience had resulted in some reflection that maybe--just maybe--this plane should be removed from service and throughly evaluated and tested before being returned to operations?
19
Just reading the first few comments, it's clear there is a profound misunderstanding of what happened here. Pilots don't need to know the nuts and bolts of a system and how it works. They need to know what it does and what to do if it malfunctions. Runaway trim, which is what happened here, is known to absolutely every pilot who has flown a 737 since the first one left the ground. Since at least the 737-400, which started coming off the production lines in 1984, if you have a runaway trim situation in a 737, you pull the circuit breaker. End of story. The reason no specific mention or training was given regarding this new MCAS system is that none is needed. If the stabilizer trim goes haywire, step 1 is to pull the breaker. You then operate the trim manually via the trim wheel. This is not rocket science.
This is a very basic failure on the part of a clearly unqualified and untrained crew. The fact that they faced these pitch deviations for 11 minutes without pulling the trim breaker is beyond comprehension. In fact, if you put two well-pilots in a random aircraft and simulated this exact situation, without knowing anything about the airplane , their first question would be, “Where is the breaker?” It’s not the most obvious first step when dealing with runaway trim, it’s the ONLY first step when dealing with runaway trim.
If these guys DID pull the breaker and that didn’t resolve the problem, it’s a whole different story. But it won’t be.
249
@MR:
I'm afraid you are operating under your own misunderstanding of what happened on the accident flight. I won't attempt to explain here, but it might be a good idea for you to head over to pprune.org and read the relevant threads there.
And by the way:
>>> "You then operate the trim manually via the trim wheel."
If you can spin fast enough, at 5,000 feet.
31
@MR
There is "a profound misunderstanding of what happened here" because everybody is guessing what happened. The release of the information from the flight data recorder will be a big help.
I find it disconcerting that they have not yet found the cockpit voice recorder (CVR). Evidently it's still on the ocean floor, and it's 'pinger' has quit. If and when it's found it should provide more information about how the pilots handled this emergency.
13
@MR I am a complete know-nothing when it comes to flying a 737, but from what I read in a previous New York Times article about this (below), the steps are as you describe except that instead of pulling breakers the procedure for disabling the MCAS system is even easier: flip two covered switches down. The switches, appropriately, are called stabilizer trim cutout switches and are prominently located in the central console at knee level. Among the myriad of switches, levers, and buttons in an airplane cockpit, these seem relatively prominent, suggesting their importance. Then, as you say, they would control the flaps manually using the stabilizer trim wheels. In fact, pilots on a previous flight of the plane that crashed did exactly that.
It would seem that this procedure could have been relayed to the pilots over the course of 11 minutes of struggling with the plane since it apparently did not occur to them. There was also an obvious lack of communication between pilots of the previous flight that encountered the problem and pilots who later flew the plane.
https://www.nytimes.com/interactive/2018/11/16/world/asia/lion-air-crash-cockpit.html
16
I find an astonishing level of ignorance in so many of the comments on this article. I wonder if half the writers even read the article.
I'm an electrical engineer who specialized in automatic control. I also hold a commercial pilots and was a flight instructor for years. And, no, I have never worked for Boeing.
Automated flight controls are far safer than human-only systems. But things, especially humans, do fail. Boeing built in a system to take charge if the attitude of the aircraft gets dangerously close to a stall. That would save the passengers in the event that a pilot brought the nose up too high. If that system fails, the pilot must disable it. The procedure, as stated in the article, is to turn it off at the yoke and trim manually. EVERY pilot has done this maneuver in basic training.
The reason so many people are dead is because the pilots mishandled the emergency. The procedure to shut down the MCAS system is in the manual but the pilots apparently never read it. The airline, not Boeing, is responsible for training its pilots, and obviously the company didn't adequately fulfill this responsibility.
When new systems are introduced, especially if they depart from the behavior of earlier systems, pilots must be trained to deal with the inevitable system failures. That's what simulators are for. I wonder if Lion Air ever spent the money on them?
When the parent company won't spend the time and money to mitigate these factors, accidents like this one will occur.
182
WRT the use of the manual trim wheels: If MCAS/STS/whatever has moved the horizontal stabilizer to its full nose-down position, 50 revolutions of the wheels would be required to restore it to a level flight position (per the Boeing FCOM). And the amount of resistance would depend upon the air load on the stab.
23
@Bill McGrath What is unclear to me is whether the 737 Max manual had a warning before the MCAS failure emergency procedure that noted that the 737 Max procedure differed from the procedure for other 737-family aircraft in that control column inputs alone could be inadequate for counteracting a runaway stabilizer.
If this article is accurate, the pilots should have had adequate time to diagnose and executed the proper procedure, but it would seem crucial for Boeing to highlight key differences in how different versions of 737 aircraft would handle in an emergency.
27
@Bill McGrath
Suggesting that the flying public is ignorant is not helpful. They are the ones who put themselves, their families and their trust on these planes. I too am an electrical engineer who specializes in automated controls. But I have found that a good dose of paranoia is good thing when it comes to airplane safety. Always remember the old adage: if the tray tables are dirty, the plane hasn't been maintained.
53
IF the MCAS system is at fault and the pilots were unaware of how to bypass it and manually trim the aircraft ( a chore that would increase their workload and would warrant a return to base), then we must see why Boeing thought this system would be working in the background and the pilots would not be aware of it. The MCAS system was installed to reduce pilot trim inputs due to the changes that were made in the MAX 737. Larger more fuel efficient engines installed further forward and higher off the ground, extended height nose landing gear assembly gained fuel savings but the new engines nacelle shape caused increased nose up pitch under certain conditions of weight and balance, airspeed and angle of attack. MCAS handled that without pilot input. What we have is one more step away from "seat of the pants" flying. Since the advent of hydraulic or electro servo control systems this has become more and more entrenched. I'm not saying this is all bad, it is the price we pay for larger and more efficient aircraft. Do not forget that Boeing, for all intents and purposes, invented the aircraft cockpit checklist and pushed hard for cockpit resource management procedures for flight crews. Huge advances in safety. Let the investigation play out but beware of depending on technology that is not fully understood by those who use it. If you want to get there fast and fly you are tacitly accepting ALL of the incumbent risks, those that are known and those that are unknown.
37
@Roscoe
Mostly well said, mostly agree.
I am baffled however as a jammed AoA sensor would be part of Sim training right? Would be right there with ASI disagree, gyro issues, etc.
If it is, why didn't the crew seem to know about it? Others did.
@416 Yes, one would think that would be routine in any training. At this point we simply do not know what was offered as training to the Lion Air crew. Normally, purchase of a new airframe would include factory training for a certain number of crews. It could be that Boeings training was deficient in this area. That would be a surprise but we will have to wait and see.
This is not a surprise. Sounds like Boeing really screwed up. Check to see if it was hacked.
1
I’m just a passenger who likes his captains with a head full of grey. To me this is a nothing burger — if a pilot and first officer don’t know how to override the automatic systems and fly the plane manually then they shouldn’t be in the cockpit. Not Boeing nor Lion Air but pilots have the last say in how they fly.
4
"Pilots for two U.S. airlines flying Boeing's 737 MAX weren't trained about a key change to an automatic system that's been linked to the fatal crash of a Lion Air jet last month, according to pilot representatives at both airlines." Article Seattle Times-
Dominic Gates 11/12/2018
Take the time to read this. Also check out Lauda Air 004 crash and the experience Mr. Lauda had with Boeing.
and be sure to "report suspicious behavior".
4
Boeing thought they knew what was best for their customers (the pilots) -- even better than they knew themselves. Boeing was terribly wrong and must be held accountable.
7
How does one explain to those families that a new software was introduced to the Boeing 737s and Boeing forgot to train the pilots using the planes?
7
@Harold Rosenbaum I don't think the manufacturer trains any pilots, not that the system is good either.
1
@Harold Rosenbaum
It’s up to the Arline operators to train the crew. Not Boeing.
Apparently Boeing's latest 737 MAX 8 M.C.A.S. should be considered a widow maker? The misdirections in the manual and awkward pilot instructions to deactivate the M.C.A.S. anti-stall system. Watch as Boeing directs responsibility by blaming the pilot for Boeing' mistakes.
6
I think the Times owes Lion Air an apology for the article right after the incident that focused almost 100% on their safety history rather than Boeing.
7
No one owes Lion Air an apology for reporting on its terrible safety record. It's really terrible.
Also, despite Boeing's unforgivable failure to document the MCAS system in the 737 MAX, it remains the case that Lion Air put this aircraft back in the air, on four separate revenue flights, without properly addressing a critical maintenance issue.
3
So it turns out that, after all of those articles trashing LionAir, it was in fact Boeing's design flaws and misrepresentations that were responsible for this disaster.
7
The 377 model should be grounded by the FAA until Boeing is able to assure the flying public and those officials entrusted to protect them that what appears to have been a faulty electronic system can be overridden quickly and easily by the pilots.
7
This seems like a design flaw to me. The aircraft should have 3 AOA sensors that used "majority logic" (i.e. 2 sensors would have had to agree that the AOA was too high before the automated system could be activated).
With only 2 AOA sensors and a disagreement between them, the system should have only told the pilots that there was a AOA problem and not activated the stall warning system.
Just my $0.02
10
Boeing has several contracts with overseas outsourcing firms who write software code, perhaps the code that failed here.
I am sure it is no accident that the proving ground for the new aircraft, with lots of code written by low cost outsourcing contractors, is: Asia.
Boeing specifically avoided launching in the USA where the NTSB and the FAA and an army of liability lawyers would figure out the root cause.
and that root cause is-->American management selling out American software developers for bigger bonuses by using low cost overseas software sweat houses where quality is, well, where crashing is common.
9
@Michael
Maybe, maybe not. But, we'll have to get used to it.
Bearing in mind that outsourced code development costs about 1/4 of the US and European rates, companies are going to keep on placing their contracts in the cheapest places. You can hire a good PhD in India for under $20,000 per annum.
Remember also that modern aircraft have become so complex as to be beyond the capacity of aviation regulators to understand. That's why the FAA now allows companies like Boeing to 'self-certify' their aircraft systems for airworthiness purposes - because it would take too long and cost too much for the FAA to do it.
@nolongeradoc
Let me understand:
Because the software and hardware design is complex, we need to get used to lower cost, less capable engineers performing critical life and death delivery to Boeing so Boeing management can get larger bonuses?
Because, that is what you are saying.
1
This is eerily similar to a series of crashes in Airbus planes in the early 1990s, where the pilot automation would be in land mode, and insisted on going down no matter what the pilot did to try to override it. There were several crashes and hundreds of deaths before they got the software worked out.
4
@Patriot
There is a fundamental difference between Airbus and Boeing aircraft. Airbus planes are all fly by wire. There is no direct connection between the flight controls in the cockpit and the control surfaces that govern the attitude of the aircraft. A computer, or series of computers are flying the plane based on complex software and pilot inputs. Ultimately it is the computer that decides what the plane is going to do. There are redundancies and sophisticated safeguards to insure the plane can operate safely, but history has shown that due to combinations of pilot error and mechanical failures, Airbuses have crashed.
Boeing, even though it is advancing into the arena of computer controlled flight, still gives the pilot the option of taking full control of the airplane in the case of a failure. There are still cables and hydraulics that run form the "stick and rudder" to the wings and tail.
A modern jet airliner is an extremely complex machine that takes hundred of hours to master. Sophisticated flight simulators have been responsible for saving thousands of lives as pilots get to experience real world flight problems without the threat of a real world crash.
There is nothing more important than proper training and review and more training and more review. Pilots have to know every aspect of the aircraft they fly and how to best react to problems when they arise. The reaction should be as close to automatic as possible.
5
@Patriot
Thanks for setting the record straight! We feel better knowing that Airbus killed a few hundred more than Boeing. We can now just wait patiently for a few more hundred deaths before Boeing fixes it.
2
@Ronald Weinstein
I wasn't defending Boeing or being critical of Airbus. I fly on both without giving it a second thought.
The majority of air crashes, no matter the manufacturer, are due to piloting errors. Sometimes, downright stupid mistakes. Today's commercial aircraft are a technological marvel and safer than ever.
While important to understand the 'why,' so to be able to correct for future -- this only compounds the heartbreak for victims' families, as this was certainly a totally avoidable tragedy.
The utter horror that passengers and crew went through ...
I will keep them in prayers.
~ 9/11 family member
3
The world is down to, essentially, two commercial passenger aircraft builders, Boeing and Airbus. While I sure that neither wants to kill us, knowing that we have fewer choices does nothing to enhance their dedication to our safety.
I would be hard pressed to avoid Boeing planes if I came to the conclusion that their safety record is unsupportable.
Our economy was founded upon competition, and relies on competition to better serve our needs. Rampant monopolization has lowered our quality of life in many ways, including, as evidenced here, our safety.
7
@Justin
Sorry, have to disagree with you where aircraft are concerned.
The only area where there are 2 Western manufacturers (not a monopoly) are the larger airframes and long haul types.
Mid size/mid range and smaller has lots of competition.
That aside, both major brands have a lot to lose if their planes don't keep up with modern demands or are deemed unsafe.
Both have an impressive track record for safety, almost unbelievable when you look at airframes in service, operating cycles, hours flown, PAX transported. By any measure it is truely impressive.
Take the 777. In service since 1995, 1500 or so active hulls. 6 or 7 hulls lost. 1 crash fatality, thats it. One. Asiana 777 that crashed at SFO. Even that person survived the crash but was run over by a first responder vehicle.
The rest of the fatilities were Malaysia Air 370 - lost in the pacific - cause unknown, and Malaysia air 17 shot down by Russian military.
Note - Not one Western developed nation airline has lost a passenger on a 777. Mighty.
Other airframes, including the 737 , have similar statistical safety records, at least in Western airspace.
While the sentiment about monopolies is correct, the aircraft industry is not one where your theory applies. Both majors do an incredible job.
Its usually the operators where the problems occur, and there's lots of competition there.
1
As someone who lost a family member to another Aviation tragedy, this does little to help my confidence that airlines consider passenger safety as their highest priority.
11
This incident highlights why it is unlikely that we will own fully-autonomous cars anytime soon. These highly complicated machines require a high level of maintenance to property function. If you can't trust an airline to run an adequate maintenance program, how can you expect millions of car owners to diligently replace faulty sensors?
28
@Pete
The two are not comparable systems. And there’ll be autonomous cars and trucks on the roads within five years.
California is already restriping its freeways and making other necessary changes to the roadways to accommodate them.
As to “owners replacing faulty sensors”? Uh, they don’t do that now. Most can’t even replace their wiper-blades!
1
So basically it seems like technology prevented the pilots from safely flying the plane. Perhaps we have become too dependent on magical technologies, that have the potential to fail or lull people into acceptance or inaction. Perhaps those who profit from implementing unnecessary technology should not be making strategic decisions.
19
@RC Amen! And to think self driving cars are coming and will keep us safe. Doubtful.
3
Simply put, I'm not booking a 737 max trip.
26
@cfc
Actually, it may be safer now since this accident, rather crash, has made just about everyone involved aware of the flaw that allegedly caused it and now pilots are trained on this.
Note to self: double check all flights before booking to make sure it is not on one of these planes.
20
Weather, maintenance, and myriad variables can put a different aircraft into play. The Boeing 737 Max is as safe as any modern aircraft.
@BrooklynDogGeek,
I did double check all the flights, but then the airline replaced the equipment at the last minute....
Great article but there's one clarification.
Despite popular parlance, the person you refer to as "Pilot" is in fact the Captain. The person you refer to as "Co-pilot" is usually called the First Officer. They are both pilots, although there can only be one "Pilot in Command" at any given time, it can be either of the two.
7
@Impatient Traveller
And we all get to be souls on board.
1
to those interested, airliners.net showed those technical memos confirmed here by American airlines, 1-2w ago.
And we're going to have millenials "driving" self-driving cars! That's on a par with legalized "medical" and "recreational" marijuana in a country (and a generation) flooded with opium addicts!
3
The connection you are making is outrageous. No country with legalized pot suffers from opioid epidemics like the US. It's a working health care system, that offers proper therapies for reasonable costs instead of pain killers being available at every supermarket.
And don't you think, that an all-time high on pot millenial generation is better off with not driving themselves? /sarcasmoff
This cant be blamed on maintenance crews, this is a brand new plane, only a couple of months old. It’s like you buy a brand new car and then blame the lack of maintenance because it killed your family. How can you expect a crew to know of a manufacturer fault? The manual Boeing claims the instructions are written in is 900 pages long! Do you expect somebody to read 900 pages in 11 minutes? 189 people died because of Boeing and you can’t blame the pilots for it. They did what they could. Any pilot on that plane would die.
8
I would expect them to RTFM long before they attempted to fly the plane! Not wait for a situation to come up and then decide to read it.
1
@Chris You absolutely can blame the pilots. Because the pilots of the same plane on the previous flight had the same problem and took the correct action to disengage the automatic trim control. This was a runaway trim situation, which on a 737 means the large trim wheels to the right and left of the throttle would be spinning constantly providing a visual cue of the problem. The pilots should have cut power (turned off) the automatic trim functions. The switches to do so are right there below the throttles. For some reason, they did not.
1
@Chris
You actually think they have a manual in the cockpit that they were looking through?
See Don and Adam’s comments, the manuals are read during training. And yes, the pilots are to blame not Boeing.
189 Dead.
Boeing: But it's in the manual.
33
No experienced aviation person jumps to assign blame for an accident. It looks like there are at least three causes:
* Boeing's design, which did not readily allow the pilots to over-ride the MCAS
* Lion Air maintenance, which a faulty airplane to fly
* Training (Airline, Boeing, FAA, etc.)
8
@Mike1 Actually, as MR points out above, there is a way to override the MCAS (which is essentially a runaway trim situation) -- pull the breaker. The switched are right there between the stabilizer wheels.
This is terrifying! As a [ mostly clueless] occasional airline passenger, I always assumed there existed an easy mechanism to fully disengage automated systems and proceed in manual mode in cases of systemic malfunction. Scary and sad that there is not!
9
Why bother putting in a manual override or manual shutdown switch if the computer is simply going to reverse the pilot's decision instantaneously? It appears that Boeing believes a computer to be superior to a human when push comes to shove.
26
Computer/sensor system fails in one vehicle, and people are killed.
Yet we want to put millions of members of the general public in automated vehicles on the highways.
10
@John Richardson: Computer control systems and sensors are what makes air travel so safe. That there are errors is unfortunate. But there is no question that highways full of autonomous vehicles would be safer than what we currently have.
3
Fail Safe!
I can't wait until computers are driving everyone's cars and trucks.
10
I'm disappointed in Boeing, for not being more thorough in ensuring that operators of the new 737 MAX were aware that the MCAS system, newly installed in this version of the 737, had specific instructions to follow, should the automation errantly point the nose of the plane down, because of malfunctioning pitch control sensors. Blaming lack of training on Lion Air is not an acceptable answer to me, especially when I read in the comments that American trained pilots flying USA Airline planes, were not given specific training on the MCAS system! Pulling back on the yoke would bring the nose of the plane back up, and keep it there, on the previous 737 models, but that does not work on the 737 MAX. Instead, the MCAS system comes back and overrides when the pilots do that, and brings the nose back down. This is evidenced in the black box report, of the Lion Air's fatal crash. Also, it seems like the Lion Air pilots did not know the procedures for complete disengagement of the MCAS system, in order to fly the plane manually, so at least they could manually fly the aircraft. Obviously, it was complicated and they didn't have time or the knowledge to save the plane, and the lives of all the souls aboard. If it is true that Boeing did not include specific up-to-date procedures on how to override the MCAS system, and fly the plane manually, and that Boeing and Airline Companies, don't require specific training for new programs in new planes for pilots, then this needs to be fixed now!
26
It's incomprehensible how this plane could even be allowed to leave the ground with a 20 degree reading difference between the two sensors.
14
My mother, a Phi Beta Kappa UW class of '39, was an inspector in the Boeing factory during the war. She's rolling over in her grave.
10
I am astonished and hugely disappointed by Boeing, if what I read is correct. In the past I worked on software control algorithms and we had one thing called “plausibility check” - that is, if the data from a sensor caused a physically implausible mismatch with other sets of measurements, the sensor would self-diagnose its own failure and shut down or reset.
I cannot believe that on a sophisticated passenger jet built in 2018 Boeing would install a system so “stupid” that it would plunge the aircraft to its doom without ever cross-checking plausibility with the plentiful available flight data - airspeed, altitude, air pressure, horizon line, etc.) all which would have been screaming “IMPLAUSIBLE” to the faulty anti-stall system. This fault was fully self-diagnosable and fully avoidable. In the age of advanced AI, this is a shameful failure by Boeing.
97
@Andy.... Well you, unfortunately, had better "believe". It costs money to go through all those system backup checks, as well as costing $$$ to write fully descriptive manuals, and train pilots accordingly. Also, keep in mind that, newer, more powerful engines- that enhance the 'pitch-up' tendencies-are the reason for installing MCAS, AND ALSO, such a system, presumably, would [help] prevent a repeat of AF 447 in the South Atlantic a few years ago. !!! How sad it all is.
1
Pilots should be required to announce to passengers that there are no outstanding mechanical, technical, or other safety issues for that plan before take-off, based on the record of the last 20 flights.
11
@Amy Berkowitz
That’s absurd. What benefit would this possibly provide?
In automated systems, if sensors don't agree, you can't rely on them. You sound a big alarm about critical sensor failure and disable the system(s) that rely on these inputs. So why did this system continue to rely on false inputs?
Furthermore, why is it so difficult to put controls in manual mode? Maybe Boeing needs to add a dance routine for emergency situations (sarcasm).
Having said that, the airline, playing hot-potato with failed parts, is definitely liable for this crash and gross negligence. It's inconceivable they just moved faulty parts around to save money.
9
This looks similar to a 777 crash on approach to Heathrow a few years ago. The engines throttled down on approach, without being commanded. The problem was a faulty sensor that registered incorrect proximity to the ground. The pilots, and some passengers in the first class section were killed, as I recall, but most of the aft passengers survived.
Turns out that pilot training (or lack thereof) was a big part of the problem. The pilots thought that because the autopilot was turned off, that this turned off all automated control. However, it did not disengage throttle control during the landing approach.
Death by automation...
23
@keith
All your claims are flatly incorrect if the flight you are talking about is BA 38 from Beijing to London Heathrow.
1
@BoingBoing
Seconded. The only B777 ever to crash at London's Heathrow airport was British Airways 38 (Jan 2008).
The accident occurred as a result of abnormal fuel icing - traced to faulty engine heat exchangers, a design error by Rolls Rolls
Autopilot or crew training issues were simply not involved in this accident.
As computerized intelligence becomes more prevalent in automated processes, there should always be a big prominent red "Manual" button on every device to deactivate automation and directly connect humans with the mechanical controls.
19
Boeing made a terrible mistake implementing a safety system that actually made the aircraft less safe than the original version. Not teaching pilots how to handle the aircraft when a problem sensor wants to kill everyone onboard seems insane. I doubt Boeing even made this sensor. Had they considered how a problem sensor would affect the aircraft?
I read in an earlier article how a button on the left side of the “steering wheel” would deactivate the angle of attack sensor. It also said that the sensor doesn’t work if the wings surface area is fully extended. Is this info still accurate?
Were the wheels at foot level, that manually controlled the rear flaps, new on the 8 or have they always been there? I mention this because if they were new and I was a pilot I’d want to know why they were there. And how and when to use them.
No matter how this plays out politically I still believe Boeing made a fatal decision putting the entire airframe and all those lives inside in the so called hands of an angle of attack sensor.
16
I'm an electronics engineer.
Reading the blog post about the FDR information, I understand why this flight was doomed. The 2 angle of attack sensors showed a large discrepancy as the plane was taxiing. The control computers should have generated an alarm and the MCAS should have been prevented from activating, since the input data was faulty.
Back in the old days, pilots would test redundant ignition systems to ensure that they were both working before takeoff. Looks like this common sense attitude has been lost in the computer age. I hope some software engineers at Boeing are having serious reflections about their negligent code killing 189 people.
55
What a strange new world we have created! High-tech planes that are supposedly programmed not to crash, but then seem determined to do so. An internet that was supposed to make all kinds of information readily accessible, but which actually spreads misinformation faster than real information. Social media algorithms that are supposed to help bring like-minded people together that actually seem to be tearing societies apart. Sometimes we humans are too clever for our own good.
20
I guess the artifically-intelligent, self-flying plane wasn't as intelligent as everyone thought it was. Chalk it up to the belief a microprocessor will ever substitute for adequate pilot training, when hundreds of lives are at stake.
And yet, some continue to believe "self-driving" cars will one day become a reality.
9
@BobMeinetz
Actually you can “chalk it up” to a flight crew that didn’t know how to override the system and take control of the aircraft.
And yes, autonomous passenger cars WILL be on the roads. And soon.
Turkish airlines flight 1951 crashed because of an automated auto throttle system on an earlier model of Boeing's 737. A faulty radio altimeter provided bad altitude information to the flight management software. However, the pilots thought that they had turned off all automated systems with the auto pilot. The pilots and 9 passengers were killed.
So, aggressive automated systems are not new to Boeing. Also, these pilots were not fully aware of that they did not have full throttle control either.
15
The number of fatalities in commercial aircraft world wide peaked in the early 1970s at around 2250 persons per year. Since then the number of passengers has increased from about 350 million per year to about 3,500 million per year! That's a 10 X increase in passengers flying today compared to 45 years ago.
The number of annual fatalities recently is about 250 per year. Stated differently - 10X more people are flying today with 1/10 the number of fatalities from the early 1970s.
There are many factors but automation likely plays a significant role in making it 100 X less likely of dying in a plane crash today then back then!
7
>>> ". . . automation likely plays a significant role in making it 100 X less likely of dying in a plane crash . . ."
Yes, very likely true. In this case, however, it appears quite likely that automated behavior, triggered by a failed sensor, by a system of which the humans supposedly in control were unaware, has killed a large number of people.
The larger body of statistics will be of no comfort to the relatives of the dead.
14
@Ed K Scant comfort to those 250 souls and their families. You don't work for an insurance company perchance?
5
@Mike D
Why do you think all information has to be “comforting”?
What a phenomenal tragedy. Boeing is clearly heavily on the defensive but if I were a pilot, I would refuse to fly any of their planes that had this M.C.A.S. system until the investigation was complete. Other planes using this system should be grounded. Only the Trump administration would not be demanding that itself to protect other flight crews and passengers.
People should make sure they are not booking flights on these planes and booking agencies should put up alerts so that passengers are aware they are agreeing to fly on a 737 with the M.C.A.S. system. In my opinion, it's not worth the risk. Boeing is not owning this problem.
41
@lb At this point, the existence, operation, and method of overriding the MCAS system have been so well-publicized that there is no need to avoid the 737 MAX. Any commercial pilot who flies these craft is by now thoroughly familiar with the problem and the solution. It doesn’t matter in the least if Boeing is owning up to the problem. Pilots want to get there just like the rest of us.
9
@Eddie Well yes, I do want to get there. And right now, from what I know, getting there SAFELY is not guaranteed if I book on a 737MAX. I will not allow anyone in my family from flying them either until Boeing has owned up to the problem, and the solution goes beyond pushing button A and jiggling switch B
2
@Lorca Your safety is never guaranteed, not in your car, your home, or in some aircraft. You can of course avoid this model.
Who checks changes in system that effect safety and practically. Who checks the information is correctly disseminated to airlines and pilots . When new drug is developed safety must be proven to the FDA. It seems transportation operates on trial and error - witness driverless cars forced into the public roads. The corrective measures to override seem very difficult to implement in a jet hurtling to the ground. Is it tested.
6
I'm also not a pilot, but a programmer (as some other commenters), and I also cannot comprehend how a piece of critical software like the M.C.A.S does not check whether the two angle of attack sensor readings are the same or not? That seems to be the only input being used by the system to make its - in this case fatal - decision. Because if the readings show a 20 degree difference, then maybe it should not activate itself at all, but show a warning to the pilots. That is a very poor software design choice in my opinion, regardless if it can be turned off easily or not after activation (well it should be easy to turn off also).
132
@Peter - I doubt that it's that simple. I cannot imagine that the stress tests applied to systems in charge of flight controls do not account for contradicting sensor information. The sad truth is that we've only been able to achieve relative safety in aviation by learning from very tragic mistakes.
5
@Peter Peter, we used to have such a system on earlier 707s, 727s, 737s. It was called an ' Instrument Comparator' system and it activated a clicker and a warning light to call the pilots attention to 'compare" the indications on the instrument[s] involved. So much for future 'improvements' !!!!!!!!
1
Commercial air travel is the safest mode of travel.
After this problem is corrected it will be even safer.
150,000 people die every year in India due to traffic accidents. How many die every year due to commercial air accidents?
2
Yes, but how many cars are on the road and how many people versus airline travel. Nobody disputes that it is statistically safer to fly commercial. However, you can ignore yor check engine light in yor car or pull over if the oil presure light comes on. "Glitches" in airplanes must be taken more seriously and what do you want to bet there was no maintenance check flight after the AOA sensor was replaced. No, that would be expensive and time consuming since we're sure we fixed the problem. Load em up!
8
The flight's outcome is deeply saddening and while this recent news may help them know the pilots did all that was humanly possible, the loss will remain too current.
The other bit of information is that this knowledge refutes the suggestion that Lion pilots were the "bottom of the barrel" as written in an article many weeks ago.
It sounds as if even my favorite Stick-and-Rudder pilot might not have been successful at saving the flight.
What is not addressed: Can pilots disengage the system with the flick of a switch? Of course, if they were not aware that the program had been installed, it might not have mattered.
If they only had 11 minutes of fight time, it seems that they should have had the authority to make split-second decisions about overriding flight control programming.
10
@LivingWithInterest Paragraph 17 addresses your question about disengaging MCAS. Not a single step but a series of steps starting with flicking a switch, which the pilots did repeatedly to no avail.
"Bulletins from Boeing and the Federal Aviation Administration of the United States since the crash indicate that pilots could overcome an incorrectly activated M.C.A.S. with a series of steps. First, they would have had to activate switches on the outside of the control columns in front of both the pilot and co-pilot. Those switches are for electrically controlling the trim — the angle of the stabilizers on the plane’s tail. The pilot of Flight 610 appears to have done that repeatedly to bring the nose up, but the M.C.A.S. reactivated each time, as it was designed to do, forcing the nose back down, and the pilot had to repeat the process again and again."
7
I began flying for the military in 1987 and flew various aircraft for the next 30-odd years. During that time I never had an aircraft I couldn't uncouple from an FMS (flight management system) with the touch of a button or two. In fact as an instructor I would regularly turn them off, and make the student conduct all flight operations (and respond to in-flight emergencies) while flying the aircraft hands-on.
But during the end of my time I began to notice trends: "automation complacency" (the computer has the aircraft) "strength of an idea" (the computer must know where we are and what it's doing) and most troubling, a decline in "stick and rudder" skills (aviate, navigate, communicate, in that order).
The pilots who failed to regain situational awareness (knowing where you are in time and space) often led to pilots who couldn't properly diagnose the emergency and take the right steps to correct it, and perfectly good aircraft with minor electronic issues have been flown into the ground as a result.
135
@Mjxs Agreed, but the article clearly states the pilots repeatedly tried to disengage from the MCAS system which somehow continually overrode their manual command/s. What we should be asking is whether the software had a hardwired command glitch producing this disaster or if it was the result of erroneous tech/ground support.
36
@Terri Cheng Yes, but, the pilots were not aware that MCAS even existed in their airplanes, let alone how it worked, or what malfunction indications might be. They weren't trying to disengage the MCAS system; they were just trying to keep the nose up, and this 'unknown' system wouldn't let them. A major difference in hindsight.
4
@Mjxs. AMEN ; but, it has been going on for at least half a century; I speak from personal and futile knowledge.
3
Boeing's reluctance to accept responsibility by suggesting that the pilot's instructions provided by Boeing were adequate for the 737MAX model when they clearly weren't stems in part from the competition with Airbus.
Boeing's 737 MAX model is in a fierce competition with the Airbus 320 neo which Airbus is currently winning . As of 11/2017 Boeing had 4068 orders for the MAX while Airbus had 5254 neos on order.
A deadly crash apparently caused by a new flight control feature on the 737 MAX comes at the worst time for Boeing in this competition. That's one reason why Boeing is desperate to blame the Lion Air crew so as to shift the blame.
As for computerized flight control, that was introduced in commercial aircraft by Airbus in the '80s and later reluctantly accepted by Boeing. One of the rationales for it was that it enabled the aircraft to avoid dangerous conditions automatically. It's credited with contributing greatly to flight safety.
That's why these days, these aircraft commonly go on AUTO shortly after takeoff and stay on it until shortly before landing when the pilots take over again or don't if they are in very poor landing conditions when, in appropriately equipped airports, land automatically under computer control. If you fly regularly into fog-prone airports, you've probably already experienced an AUTO landing.
21
@GerardM To go so quickly to auto is the choice of the pilot. I know at least one, a former naval aviator, who would never switch to auto until 10k ft regardless of the airplane he was flying. Certainly there are engineering flaws and blame should be apportioned appropriately for this tragic accident, but I'm reminded of the maxim that the best safety device in any aircraft is a well-trained pilot.
7
Why is this system even on a commercial airplane? These are professional pilots and a simple horn with stall warning screaming at them should be enough. Airplane designers overthink the problem and overcompensate with technology that isn't needed.
15
@markd Right. Sometimes it's better to keep older less-computerized systems. Just like cars. The more modern and automated it is, the less control you have. You're not even able to fix it yourself because it's all controlled by the computer.
11
I would have thought, with the proliferation of automation in aviation, that some kind of failsafe (and redundant) system was in place that would prevent a brand new jet from spiraling straight into the ground-as opposed to causing one to do so.
11
@Sswank one would think so, for sure!
I have always believed in technology’s unlimited potential to improve our level of safety and convenience in ways yet imagined. That said, the inevitable explosion of newly automated systems that will come to dominate our lives over the next 20 years requires radical new thinking when it comes to certification and testing. Technologies incorporated into aircraft, autonomous vehicles and personal robots should undergo a far more rigorous certification process than in the past. Given that complexity is increasing exponentially, our ability to test for potential faults must also go up proportionately. Since it is always ahead of the technological curve, it seems the aerospace industry would be a good place to begin.
8
That has to be a pilots worst nightmare.
6
Looks increasingly like criminal negligence on the part of Boeing
Wouldn't be the first time:
https://en.wikipedia.org/wiki/Japan_Airlines_Flight_123
Death toll : 524
12
These poor pilots ( and their hapless passengers) may go down in history as among the first to be killed by machine intelligence. It’s sad that software, which powers so much of the world today, failed these folks so miserably. The possibility that there may have been hardware malfunctions (such as sensor failure) doesn’t absolve the software- it should’ve been designed, coded and tested to be failsafe. And it’s utterly, shockingly, unconscionable bungling on the part of Boeing, the airline and regulators that such an egregious failure was allowed to happen.
51
@V P Kochikar Not even close. Apart from the relatively recent accident in which a pedestrian was struck and killed by a self-driving car, faulty readings and sensors of machinery have been responsible for many, many deaths or at least accidents over the years. Also recently some tourists drove their rental car off a cliff because of fault GPS routes. Factories and power plants have had issues in the past too because of faulty instruments making the wrong choices.
I am inclined to pin this one on Boeing but as a software developer myself I know that bug-free software does not exist, and this is neither the first time nor the last time that humans' partnerships with hardware and software will end up with lives lost.
9
@Ian Agreed there have been many people killed by software bugs in the past, but this is one single incident with 189 people killed !
3
@V P KochikarI I know and it's awful what happened, I'm just saying I had a similar thought to "This will go down in history as the first machine intelligence death" when that woman was killed by the self-driving car but when you think about it humanity has a long history of people losing their lives to machines.
2
So looking forward to this kind of reliable technology putting millions of self-driving cars on the road.
22
@Petersburgh That made me chuckle and feels true, but see the many comments about how much safer air travel is today than it was decades ago. I want that for cars as well.
(I mean, I actually don't want cars at all because they're polluting and inefficient, but while we have cars I'd rather they drove themselves.)
1
@Petersburgh
Can’t even begin to compare the two systems, flight-control systems are a thousand times more complex.
Read more.
Is this simply further proof of the disinvestment with R &D by team USA ? Let's not forget the first commercial jet was built in GB and they basically lost their aircraft industry when the plane design was revealed as fatally flawed. I'm sure the Chinese will be happy to step in and take the lead in aviation. Meanwhile a tax cut is doled out to stockholders further hallowing out American corporate strength. What do we do well ? ...grow bushels of soybeans to be exported to our future masters ? Yeah we are doing so fine.
5
I haven’t heard that the Boeing 737 Max has been grounded pending a resolution of this software issue. Hope worries of financial loss to an aircraft builder does not trump the safety of the thousands of passengers and crews that daily fly the Max.
11
@Gerard Deagle A directive went out telling pilots how to disengage the system quickly. That's all that was required to avoid a repeat of this tragedy.
1
The new generation of super automated airliners are not doing the job well enough. Pilots should be trained to operate an entire flight by hand at least one flight per month. Otherwise we are riding with pilots who do not know what they are doing. Otherwise Automation and convenience will do us in.
11
Sad to find that Boeing has taken a page out of the Airbus school of aeronautical engineering, letting computer programmers and design engineers control the aircraft rather than trust the flight crew. Air France flight AF 447, almost 10 years ago in the south Atlantic had similar issues which resulted in a similar outcome. The industry and regulators designing and certifying new aircraft have decided pilots are too dumb to maintain proper pitch control and have built in "safe guards" where an designer/ computer software coder/engineer put into place systems that do what the pilots is suppose to. When a system has such complexity and is designed to continue to frustrate the operator, as seems to be the case in the Lion Air crash, we can only look to the manufacture and question why they believe the people in the front office (pilots) are not considered capable to maintaining basic stick and rudder functions of their aircraft.
Airbus started this trend with its A320 series and followed on in subsequent lines. Boeing has, until recently giving the pilots the upper hand in controlling Boeing aircraft. We have all learned, tragically, that this is no longer the case.
22
@night mission Other than the fact that AF 447 experienced sensing equipment failure (the icing of the pitot tubes which determine airspeed), the two incidents are totally dissimilar. In AF 447's case, the plane's systems went into Alternate law and the autopilot cut out. From that point on, the pilots had control of the aircraft and it was pilot error that caused the plane to crash into the water. For whatever reason, the co-pilot kept pulling his stick back while the pilot pushed his forward to recover from the stall. The two commands canceled each other out. The plane was in a stall position throughout the descent. The pilot, btw, had called out to take command of the controls; the co-pilot failed to relinquish them.
3
@night mission - why the Airbus-bashing? AF447 crashed because its pilots were not properly trained to fly manually such a heavy and complex aircraft near its operating ceiling, with defective speed data and in the middle of a thunderstorm in the night. It was a near-impossible situation driven also by a good dose of bad luck (all three speed sensors freezing) and the inability of the pilots to correctly identifying the problem.
The case of the 737 Max is much worse - to save perhaps a few dollars, the stall detection system that repeatedly failed only relies on two sensors, one of which clearly failed - and the computers threw the dice, making the random 50-50 choice of "believing" the defective sensor rather than the good one. Had Boeing installed three sensors, the problem would have rectified itself, as the defective sensor would have been isolated and ignored.
Moreover, as I wrote on another post I'm astonished by the total lack of cross-communication between aircraft system and the lack of a "plausibility check" - i.e., if a sensor is giving out data which clashes with physical implausibility against the data coming from other sensors, such as altitude, pitch, line of horizon, and so on, it should self-diagnose as defective. How on earth can we have such a "stupid" system in 2018 that doesn't recognize its own obvious faults, is beyond understanding.
8
@night mission The Air France co-pilot kept pulling back the stick to recover from the plunging stall, which is naively intuitively correct, but wrong. Such untrustworthy pilots would be one reason Boeing added this MCAS pilot override.
The problem with this new automation is (1), garbage in, garbage out, and (2), again trusting the pilots (and airlines, and manufacturers) to have sufficient documentation and training that the override can be turned off in a moment of stress.
1
Automation/A.I., designed by humans, CANNOT anticipate at a moment's notice for instantaneous, chaotic events.
7
@Coffee Bean Unfortunately in the vast majority of cases automation is substantially better at handling instantaneous, chaotic events than human guts.
@Ian
Honestly, would you rather have a pilot with 30k+ actual flying hours respond suddenly to a mid-air crisis situation than some aero-tech nerd who wrote a program on how a plane SHOULD respond to a simulated crisis with inexact variables?
2
So “Lion Air, Indonesia’s largest airline, has a notoriously flawed safety record.” Nice! Blame the Lion Air and try to protect Boeing. There is more corporate malfeasance in this than meets the eye. Boeing made a change in the system but did not update the manual. Boeing didn’t require pilot training. Now they are trying to cover their behind. Sad.
39
@Ravi Srivastava I am from Indonesia and I, too, felt insulted by that sentence. I think Boeing has equal, if not greater, responsibility here -- selling faulty airplanes without any manual/non-automated safety measures!
21
@Ravi Srivastava
Each airline, not the manufacturer, is responsible for training their pilots/flight crews on new models of (existing) aircraft and new flight systems incorporated in those aircraft contained in that airline's fleet. Don't blame Boeing for Lion Air's failure to spend the money to qualify their own pilots on their own fleet.
Lion Air, like many airlines, cuts maintainence and training to the bone to cut costs, and safety is a side note. Weak pilots unions are unable to pressure airlines to spend money on training. Let's face it, even the Indonesian government seemingly hasn't been able to legislate Lion Air into a better safety record. But ultimately, the fault rests with the pilot in command. The final report will most certainly read 1. Pilot in command- failure to maintain altitude, 2. Pilot in command - failure to correctly diagnose and manage failure of in-flight safety system.
Seems like the new process is unnecessarily complicated and not at all intuitive.
4
Why in the world would Boeing change the very simple maneuver of pulling back on the yoke to disengage MCAS into a more complex process?
66
The Colgan Air crash in Buffalo resulted because of that very reason—pulling back in a stall when the stick needed to be pushed down. Had they had the same automatic response from the plane, they may all be alive. EXCELLENT pilot training programs are the only safeguard against a computer program (Lion Air) or a lack of one (Comgan Air).
9
This appears to be a tragic example of what happens when engineers take a back seat to marketing charlatans and software programmers who think that the world is a touchscreen, and are thus forced to abandon sound engineering practices when designing an airplane (or, for that matter, any other complicated machinery). There are (at least) two clear signs of this:
- Failure to include critical information in the flight manual about changes in emergency control procedures, as Capt. Tajer discusses. Why? So that the new plane could be sold as a "plug and play" upgrade of existing 737's.
- The fact that the M.C.A.S. did no automatically turn off if it detected a 20 degree discrepancy (!!!) in angle of attack between two sensors. This obviously indicates that at least one of the sensors is malfunctioning, and perhaps both. The software should have contained a procedure to automatically cut out the M.C.A.S and warn the pilots about it as soon as this failure condition was detected. It is incomprehensible why something this simple was not included in the software.
As I said in a recent comment about this tragedy - what was wrong with hydraulic and cable links between properly trained pilots and control surfaces? More expensive airfares? Hurrah for that - no more discount airlines, no more tourist overcrowding worldwide. I honestly wish that we were still flying in slow and noisy Super Constellations and Britannias.
12
These planes should be immediately grounded until their pilots have had adequate time to practice how to override computer software problems on flight simulators.
7
To Glen
Depends on the software
From NPR, on the US Air landing on the Hudson
"They knew that the airplane's flight-control computers had performed remarkably well, seamlessly integrating themselves into Sullenberger's solutions and intervening assertively at the very end to guarantee a survivable touchdown."
https://www.npr.org/templates/story/story.php?storyId=120355655
The airbus had a software setting for water landings that arguably saved the flight
10
@Gerhard ultimately, though, the pilots should have the final say and technology should only be used to support.
3
From my experience (as a Pilot), the Pilot Operating Handbook will have the procedure documented and Boeing had to approve the procedures through the FAA. If the equipment is flawed, directives will be issued to educated the pilots on the consequences of using the MCAS. I doubt the system will be augmented, but the correct actions will be emphasized. It usually takes deaths to alter training, but there is no doubt there is a procedure that would have saved this plane.
8
@Pilot There is a procedure in place that disables the MCAS. It's described in the article underneath the body bags photo. However, it's also said that either the Lion Air crew didn't know about this procedure, or were trying the alternative they "knew" would work- pulling back on the control columns. This worked on previous models of 737s, but not the MAX.
Hopefully, this accident will force a new, easier way to disable the MCAS in such a situation.
2
As a current 737 pilot, I have heard several of my colleagues suggest that training is just as insufficient in the U.S. as it is in Indonesia. We were not informed of modifications to the trim system here either. Although I'd like to think my excellent airmanship and superior system knowledge would have saved the day, I am forced to conclude that I might have crashed the jet myself despite 29 years flying 737s.
Right now, Boeing is in an inescapable position of having designed a system without proper crew training of its possible faults. Regardless of who is determined "at fault", Boeing must repair its credibility by taking ownership and designing some new procedures PRONTO!
433
@Randy The closest I have ever come to flying an airplane is riding in one so please forgive what might sound like an ignorant question. According to the article below, there are two prominent covered switches in the console between the pilots at knee level called stabilizer trim cut-out switches. If you were flying a plane that kept nosing down, would you not have recognized the problem was probably with the MCAS and disabled it using those two switches? That is what the pilots of a previous flight of the plane that crashed did.
https://www.nytimes.com/interactive/2018/11/16/world/asia/lion-air-crash-cockpit.html
6
@MRod
You are right, the trim cutout switches are located on the pedestal and should have been turned off to deactivate the trim motor. That's the way it works in older 737s and I think that's the way it works on MAX as well. Once the motor is cutout, there is a folding hand crank that can be used to trim the airplane manually.
HOWEVER... There is a third system which will freeze the trim wheel when you trim against pressure on the control column. It turns out that this new, unexplained MCAS system deactivates this trim freezing system. If it were me, I'd be wondering why the trim keeps running even though I'm fighting it with the control column. Would I have thought to use the cutout switches? I don't know. This crew was struggling to control the airplane, the captains stick shaker was going off for the entire short flight. The stick shaker is LOUD. It is not deafening but it definitely ends all casual conversation in the cockpit and is designed to wake the dead. The amount of distraction and confusion had to be immense. I'm not going to criticize the pilots yet. They were dealing with an unknown failure mode and were truly fighting for their lives. Two little switches could have saved them but they didn't know or didn't think to turn them off. An accelerating 737 develops very heavy control forces without trim. With nose down trim it appears to have overpowered the pilots. It will be interesting reading when the final report comes out.
18
@Randy
There is no new procedure, it is the same as it ever was. Lift the cover and turn off the automatic trim.
You don't know how to do that?
I am no pilot or aviation expert, but I know a bit about software. It is inconceivable to me that the software could determine the plane was in a stall and needed the nose pushed down when it was traveling over 400 miles per hour. It also didn't detect the difference between the two angle of attack sensors and determine one sensor might be malfunctioning. So from an outsiders perspective, it seems like very poor programming with little regard for using other inputs to determine if indeed the nose needed to be pushed down.
156
@Glenn
But very poor programming is to be expected from a generation of "programmers" whose most notable experience is with social media "apps", rather than with rigorous programming of a rigorously defined set of specifications.
18
@red sox 9 the people writing web-based apps are not the same as those writing formally verified and formally proven software architectures behind safety-critical systems. in the same way your family doctor isn't a brain surgeon, there's levels of certification, training, mentorship, oversight and auditing that go into being permitted to work on these systems that just don't exist in the web software world.
14
@red sox 9 I'm inclined to blame Boeing for this one but I think your comment is disingenuous. You have no idea what training the software developers had and what other factors led to compromised results. Probably the embedded systems weren't written in JavaScript so it's a silly jab to make.
7
I'm a private pilot with over 3000 hours. I have not seen the documentation that accompanied the rollout of the new aircraft, but it appears to me that the new system had a new catastrophic failure mode that was buried in the new paperwork. If this is so, then regardless of the probable lack of good maintenance, Boeing is in deep trouble. When you build an aircraft that will crash itself with the failure of a sensor regardless of the direct control inputs of the pilots, you have started a time bomb ticking. The story suggests that the pilots tried repeatedly to turn off the failing system, and could not recall, or did not know the correct multi-step process to save the aircraft. How many of us have perfect knowledge of all our aircraft systems, and the ability to recall any of them instantly in an panic situation? What other failure modes are hidden in these aircraft? My confidence in our commercial aircraft design has taken a serious hit with this disaster.
460
@miller mcpherson What do you fly? What are your ratings? I find it stunningly difficult to believe that a pilot with your purported level of experience could possibly believe what you've written. Runaway trim is a very basic emergency (and for a professional flight crew, an emergency is not a "panic situation"). If you have electronic trim in your aircraft and you don't know the process for disabling it by heart, you are either incompetent or a fool. Any properly-trained line pilot flying a 737 certainly would. And even if you don't - which you should - that's why God created airplane flight manuals. There are two people in the cockpit for a reason.
7
@miller mcpherson Any automated system that sends an airworthy climbing aircraft into a full throttle steep dive into terrain is a hazard to any aircraft on which it is installed.
12
@Winston SmithNose down trim is not "a full throttle steep dive". Aircraft have been equipped with automatic elevator trim for many decades and they all have had a switch to disable the system in case of need. These pilots failed to remember a basic aspect of flying anything larger than a piper cub: the switch (located on the front of their console) that disables automatic elevator trim.
4
If I compare to The Healthcare Industry, it takes an enormous amount of testing and evidence before a new drug or new indications for an existing drug can be approved. The aviation industry is at the same level as the pharmaceutical industry was in the 50ies. I.E the companies were producing, testing and also handling complaints. I hope this could be a wake-up call for the industry.
7
@Dan Elson please, keep taking care of healthcare industry. Have you heard of the new scandal about prosthetics in Europe, that is American products tried and sold in Europe before FDA's approval?
Re- the NYT's article, it is surprisingly of good quality on this aviation subject.
P.S. I have 15000 hrs on B737 alone (13K PIC)
Peace
The problem is not automation. Flight automation has made revenue flights more affordable as well as improved safety remarkably. The real culprit is first the failsafe systems are not as easily engaged and second, as demand for pilots increases especially in less developed countries, airline pilots are being instructed in flight management not flight training. The consequence may be tragedies like this where Boeing will point out how, in retrospect, the solution was readily available but the pilots thought process could not transition through the possible solutions fast enough to save this flight.
33
@Frederick It wasn't in the training for the Max. Try digging through a stack of flight manuals while hanging on for dear life. Yes, this was an automation problem that thankfully has been exposed before more lives were lost.
2
I wonder how, in an emergency, pilots have time to start reading a complex manual on how to figure out what to do. There needs to be a master disconnect switch so a malfunction in a computer-controlled environment can be turned off. Those of us who have done programming knew that strange situations frequently develop in ordinary programming. The same must be true for airplanes too.
54
@George
For more real-life drama on this type of situation, read The Devil at 37,000 Feet, by William Langewiesche.
5
I wouldn’t be surprised if these fly by wire planes are impossible to fly the old way. You can’t just turn off the electronics and fly the old way. I don’t know this as a matter of fact. But it seems possible considering how expensive a full hydraulic system would be. Especially as a redundant feature.
2
which entity requires all these automatic correction systems?
is it the FAA or Boeing?
why is the pilot not trusted to be able to manually correct a too rapid ascent ? Do they want to save the plane when the oxygen fails or some similar reason?
Why are the planes getting so automated that the pilot does not even fly the plane except in unusual circumstances?
is this better than before?
what if the machines fail?
why was a used sensor employed for the replacement?
why weren't both sensors replaced?
why if the sensor is not robust, are there not 3 or 4 sensors, with only 2 it is difficult to tell which one is wrong.
where is the redundancy in your automation boeing?
why is it impossible for the pilot to permanently turn off the MACS?
etc etc
better re-examine all your automated systems, because likely some of the others also suffer from these design features.
58
@Will Hogan
Will:
Fixing a problem requires the right questions to be asked. You’ve done a great job in formulating this. Your comment regarding why there are not multiple sensors is something that is a very good observation. I recall reading about the Air France crash over the Atlantic. I remember thinking, this entire aircraft’s survival and well being depends upon a single small metal tube mounted on the end of the wing (pitot tube). Here is another similar situation. By ingraining automated flight controlfurther and further into aircraft operation, we are squeezing out human judgement which seems like a necessary ingredient in recovering from an emergency situation.
23
@David There are generally three pitot tubes, and they "vote". However, in the case off 447 I believe they all froze up. Not a good situation.
3
I sincerely hope the surviving families collectively and successfully sue Boeing for an ungodly amount of money.
68
@ben
Instead, I would hope that the families receive reasonable compensation and immense pressure be put on Boeing to detangle what went wrong--and then fix it, so that in the future, other families will be spared such loss.
30
@anianiau Ben’s suggestion of high payouts is exactly what would supply such “immense pressure.”
11
Ben sue Boeing for what? It's not Boeing's fault that the pilots were incompetent. In every large aircraft there is a clearly marked, prominently placed switch to disable the stab trim. There is also a circuit breaker. They could have also disabled it by the pickle switches on the yokes by commanding opposite travel direction.
1
Those poor pilots, valiantly fighting a machine - man vs machine rarely ends well for man.
And we are rushing ahead to deploy driverless cars and trucks, with 100% trust in these same machines.
200
@ajsdelhi The dangerous, incompetent drivers I witnessed in traffic today cannot be replaced by automation soon enough.
13
@ajsdelhi 1,200,000 dead on the roads last year because of how well humans drive. If machines only kill less than 100,000 per year (initially), wouldn't you say that is a huge improvement?
And no, the guilty often survive in car crashes.
18
Michael Crichton's Airframe. Old fiction but offers interesting scenarios.
7
@Moonstone
One of his best - and some parallels with this tragedy.
If you haven't read it, the pilot of a fictitious 'Norton Aircraft' passenger jet ends up fighting the plane - well, the autopilot - for control, resulting in ever more violent 'porpoising' and thus serious injury and several fatalities (including the experienced Captain) amongst those on board.
The battered US built airliner makes a forced landing at a neutral airport where it's noticed that its Far Eastern operators 'disappear' their remaining flight crew suspiciously quickly. The one injured stewardess remaining in the hospital has clearly been frightened into silence.
How could this happen to a modern aircraft and its highly trained, experienced crew? Is bad aircraft design the cause? What is the airline hiding and why?
You'll have to buy the book..
BTW, I realise that a dreadful air accident with great loss of life has occurred and I'm not intending to make light of that.
It seems to me that bankruptcy is the equivalent remedy. The souls on the plane are for all intents and purposes bankrupt. Everything they had or would have was taken away from them involuntarily (and terrifyingly). So Boeing being sued into bankruptcy seems like a fair bargain. Either that or someone going away for manslaughter. Flipping switches, closing circuits, referring to manuals during an in-flight crisis, yeah, that's begging to get a plane load of people killed. So manslaughter seems fair too.
24
@Richard Mclaughlin I would trust the pilot has read the manual fully prior to flying a new model plane. There’s culpability on both sides.
1
@Richard Mclaughlin
"Boeing being sued into bankruptcy seems like a fair bargain"
So how does making many thousands of Boeing employees unemployed help anything? Top management will do fine, but those below them will suffer.
2
@Richard Mclaughlin Then Airbus takes over, and we live (or die) with their mistakes. No. Hold the people responsible. Also, Asian airlines (for whatever reason) always seem to push the limits on maintenance of their planes, ending in disaster.
Without delving to deeply into the technical details of the systems that evidently led to this catastrophe (I have been following developments closely, in professional aviation forums since the hour of the crash), it simply has to be emphasized that Boeing's failure to include full information about the MCAS system in technical and operations manuals is unforgivable.
With no knowledge of the system that was repeatedly overriding manual control, and no obvious and instantaneous way to terminate its operation of the trimmable horizontal stabilizer, there was essentialy no possibility that the crew would be able to recover while flying at only 5,000 feet.
As one insightful airline pilot has observed, they were fighting hopelessly to regain control while their "airplane was quietly trying to kill them." Which it did, along with everyone else on board.
298
@Douglas: This article states that the procedure to deal with a malfunctioning attitude control system is to first turn it off and then manually trim the plane. It doesn't sound like a complicated procedure, but if the pilots aren't adequately trained, they might not know what steps to perform. I wonder if this particular failure mode was simulated during the pilots' training. My guess is no. (I hold a commercial pilot's license and was a former flight instructor, but I don't have experience with this class of aircraft.)
17
@Bill McGrath:
Hi, Bill. One of the problems on the 737 MAX is that there are multiple switches with new labels that refer to trim cutout and they do different and non-intuitive things.
It appears, from what we know so far, that the accident crew was unaware that the MCAS system existed and thus unaware that, while its trim commands could be overridden by inputs on the electrical trim switches, it would reinitiate its own trim commands a few seconds later, if it continued to "believe" that AOA needed to be corrected. In this case, a faulty AOA sensor seems to have made this so. And, of course, there is no direct AOA readout on the display, so . . .
Finally, at 5,000 feet, manually cranking the trim wheels is just too slow -- way too slow -- too allow for recovery.
39
@Douglas Wouldn't you think that at some point during the 11 minutes this was going that one of the two would have thought about pulling the breakers? Or disabling the stabilizer trim? Instead of just fighting it? It really makes no sense. At some point that had to come to their minds. If they didn't know how to disable the trim, that's not Boeing's fault, that's a training issue. And a major one. You don't need to know a lick about the MCAS to disable the electronic trim.
3
Boeing is in BIG trouble!
15
The same symptom (nose down) occurred over and over for 11 minutes and possibly no-one in the cockpit knew the second or
third step needed to override the malfunctioning computer control. I'm not sure how complicated it would be to put the entire craft into manual pilot control, but if it's not a single red panic button, I suggest that get looked into.
164
@GP
You just can't dumb down flying to that level. It might have worked in this case but in other cases such a switch might compound a problem. Air France 447 is an example of why pilots need to be pilots.
If they can't handle the safety systems, how can they handle all that mother nature can throw at them?
GP I having maintain large aircraft for a major airline can tell you exactly how to disable a runaway stab trim on a DC-10, MD-11, DC-9, Super 80, B-737, B-727, B-757, B-767, B-747 and Airbus A300. Airline pilots generally only are type rated on one or two aircraft types. The real problem is the pilots didn't know how to operate the aircraft.
1
@Pauli , @Sfojimbo
I guess my issue is that disabling runaway tab trim or other control malfunctions could require an immediate correct diagnosis of the problem in order to take the necessary corrective steps. If the "what the heck is going on and how do I fix it" time span is too excessive, we see what can result. It is probable a well trained pilot will figure it out quickly enough, but it seems to me, if the planes have become so complicated that a pilot typically can't control it without computer help, computer malfunctions will guarantee these types of crashes to happen again, try as we might to train pilots and employ computer redundancy. I believe the F-22 Raptor jet fighter qualifies as impossible to fly without computer input, but those pilots have the option to eject when all goes haywire.
1
Automated transportation sounds wonderful until we're confronted with the reality that computers and software fail often. We must make sure there are capable, experienced and fairly paid pilots, train operators, bus/truck drivers, etc and that they can take manual control of their vehicle at the snap of a finger.
85
@Scott H Automated cars are being tested for years and in spite of almost successful they have not been approved yet. How come Boeing just gets away with such faulty automated issues with note even a slap on a wrist?!
6
@Sri Maybe it is because automation is working spectacularly in commercial airlines. In 2017, 36 million commercial passenger flights resulted in ZERO fatalaties. You had a better chance of dying by being gored by a bull in 2017 or falling into a hot spring than by being killed in a plane crash.
1
This isn't the first crash caused by false data from angle-of-attack sensors. The article says this Boeing 737 has two of them... Why not FIVE or more to give the necessary redundancy to prevent a crash like this?
24
@CarbonBoy Why not? Weight, thus efficiency, thus ultimately capability and cost. You can't simply put 5 of everything on an airplane.
2
Self driving cars. Sure.
77
@Patrick G You think human drivers are safe? Really? If self-driving cars fail occasionally, that is still far, far safer than humans who drink, text, drive drowsy, and drive at excessive speeds.
36
@Patrick Self-driving cars are lacking a capability essential to safe driving: Situational awareness. I've worked in AI since the 1980's; it'll be many decades, at least, until that capability is realized.
That many drivers, with all the distractions available, may lack that as well is not an argument supporting self-driving cars. Rather, it's an argument for removing distractions.
28
@Patrick G Your assertion is simply not true. In testing, humans must take over from autonomous cars every 0.8 miles. Hardly an improvement.
"Uber’s autonomous test vehicles in Pittsburgh all have backup human operators and, in over 20,000 miles of operation, those operators have had to intervene every 0.8 miles."
https://www.recode.net/2017/3/16/14938116/uber-travis-kalanick-self-driving-internal-metrics-slow-progress
9
Wait! What? In the other interface between man and robot, robot won?! Wait! The humans were already supposed to know what to do but the robot defeated them? Well you can replace a robot. Guess you’re left with blaming a human. That way nobody really get sued! I feel for those poor souls who believed in the robotic hype.
7
Boeing has not behaved well. Their reputation is on the line and they are resorting to the typical PR playbook. Deny until you can deny no longer that there is an issue.
I expect better from them.
86
I'm a bit fed up with engineers designing computer controlled stuff that thinks it is more clever than humans. It happens even with personal computers, but in planes those things are fatal! There has to be an easy and uncomplicated override to those functions - always! Aren't Boeing engineers aware or their responsibility? As a passenger I want pilots in main control, not computers and sensors!
98
@Christian It's been decades of planes being flown by computer. There will be of course, failures, but flying has become incredibly safe due to technology.
10
@Christian: The article describes the procedure to defeat the automatic attitude control: turn it off and trim manually. The problem here sounds more like poor pilot training in that they didn't know what procedures to follow. Engineers know that systems can malfunction and they build in mechanisms to override errant equipment. This only works if the pilots are well-trained and know what to do in an emergency. That apparently wasn't the case here.
6
@Bill McGrath
Good design also includes fail safe options, for instances just like this one. Easy manual override of automatic flying system, with a push of a button. Allowing the pilots to take full control. Is that really so hard?
4
Horrifying. It reminds me way too much of the Hal 9000 computer.
20
@Mons
Good analogy. Unfortunately, this was not a movie but involved real men, women and children.
1
Let's have more "artificial intelligence" like this, "the Boeing 737’s nose was repeatedly forced down, apparently by an automatic system receiving incorrect sensor readings."
Hard to believe WE THE PEOPLE - average people around the world - cannot feel safe in the air since insatiable greed got ahold of every segment of the industry.
Don't fix anything. Don't train pilots. Spend as litte as you can on employees. Pocket as much as you can. To hell with the consequences. A few people die - so what? We're sorry and we might fix the problem if forced to do so.
Is this what OUR world has come to? I, for one, will not stand by and let the insatiably greedy, morally/ethically bankrupt, socially unconscious International Mafia Robber Barons destroy my life any further.
If every single American and average person around the world does one thing to stop it WE THE PEOPLE will win. Let's Do It.
38
@njglea
It reminds me of the prevailing attitude toward guns and the safety of the general public.
5
Commercial flights are remarkably safe. Airplane fatalities have been dropping since the 1990s. I think your statement is a bit exaggerated. I'll probably avoid Indonesian airlines but the vast majority of commercial airlines are safer than even a decade ago.
1
@Chris And, it is precisely the dropping fatalities statistic, that has buttressed and encouraged the trend for lack of training and systems knowledge that led to this accident. nglea has hit the nail on the head; the almighty dollar and greed for it is the basic, deeply underlying cause of this and many other airline accidents, and it has been going on for over half a century to my own personal knowledge
1
The crowded, noisy, distracting work environment in many Boeing offices is not conducive for enabling Boeing’s engineers and software developers to concentrate fully on their often very complex tasks. While it would be difficult to prove that a distracting work environment led to this particular crash, I do believe that it would be much better for all companies, including Boeing, to provide a work environment in which their employees can concentrate when needed and to collaborate when needed.
49
I'm quite taken aback that Boeing has taken a page out of our federal administration's playbook: downplay the issue. They say the proper steps are already in the manual. Well, obviously, it wasn't emphasized sufficiently in that manual and all these people paid the price with their lives. Would it kill Boeing to send out an alert and a training module to all the airlines using the Max? Obviously, it's more profitable not to go to that expense. Good choice, Boeing.
207
@BA Boeing did issue a Notice to All Pilots within 48 hours of the accident detailing how to recover from this situation.Not taking anything away from your argument, just stating the facts.
19
@B737 Pilot A little late, you think?
20
@BA
This "down-playing" may be driven by their lawyers desire that Boeing not admit that this failure mode is not documented in the existing POH (Pilots Operating Handbook) by issuing a revision to the POH spelling out exactly to do when the MCAS fails.
When they are in court, and they will be there eventually, they don't want a victim's lawyer to say,"If it was documented in the POH, then why did you revise the manual after the crash?"
9
A part of the problem here, and in many other cases, is that in the name of safety we have removed a great deal of control fro the humans controlling our devices and given it over to automated systems. This works fine, as long as the systems work. When they do not, however, they often fight against relinquishing control. Further, the humans do not always possess the same skill levels that they used to have by necessity, as the systems do so much for them normally.
I often wonder if automating so many of our safety is actually a benefit or not.
104
@mikecody
I've heard the same thing - because planes are so automated pilots don't know how to fly the plane when the systems fail. I believe this is what happened to Air France plane that crashed in a severe storm off the coast of South America several years ago. When the plane was in trouble, pilots didn't have the skills to respond although Air France denied the accusation. Too much reliance on automation in the cockpit.
29
@DCNancy And yet crashes and fatalities are down dramatically in the US on commercial airlines. How easily we forget what a less-automated world looked like- manual piloting caused many more crashes.
20
@mikecody Well this is one case of machine bites man. But as far as this reader can tell plane crashes are almost always due to "human error?" I continue to look forward to the time when there is no pilot to be the cause of mayhem.
Same for self-driving cars... accidents with the new technology will get headlines but 4,000 AI deaths per year instead of the near 40,000 human error deaths would be an enormous improvement.
3