Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Transportation News Technology

Facing Sharp Questions, Boeing CEO Refuses To Admit Flaws in 737 MAX Design (seattletimes.com) 413

A reader shares a report: In a tense and steely news conference, his first since two deadly crashes of 737 MAX airplanes, Boeing Chairman and CEO Dennis Muilenburg faced sharp questioning but refused to admit flaws in the design of the airplane's systems. "We have gone back and confirmed again, as we do the safety analysis, the engineering analysis, that we followed exactly the steps in our design and certification processes that consistently produce safe airplanes," he said. "It was designed per our standards. It was certified per our standards."

In the case of the MAX, those processes certified as safe a new flight-control system that was triggered on both the Lion Air and Ethiopian crash flights by a single faulty sensor and then engaged repeatedly to push the nose of each airplane down. Boeing is currently flight testing a software redesign of this system -- Maneuvering Characteristics Augmentation System (MCAS). Yet Muilenburg would not concede that there was anything wrong with the original MCAS design, saying only that the system is being "improved" with the software redesign. He said airplane accidents are typically due to "a chain of events," and that "it's not correct to attribute that to any single item."

This discussion has been archived. No new comments can be posted.

Facing Sharp Questions, Boeing CEO Refuses To Admit Flaws in 737 MAX Design

Comments Filter:
  • by Hognoxious ( 631665 ) on Tuesday April 30, 2019 @10:01AM (#58516362) Homepage Journal

    It was pilot error!

    • Re: (Score:3, Funny)

      by Tablizer ( 95088 )

      The Steve Jobs of planes: "They were holding the control yoke wrong"

    • by bobbied ( 2522392 ) on Tuesday April 30, 2019 @10:25AM (#58516508)

      It was pilot error!

      Actually.. Boeing has at least admitted that it was pilot training... Boeing didn't properly document the feature and pilots where not properly trained as a result. Remember, the aircraft IS still flyable, a fact that is proven by the events leading up to the first accident, where the day before the aircraft that crashed had the identical malfunction, and the pilot who happened to be sitting in the jump seat figured it out, popped the trim system breaker and they landed safely.

      This tells me that this was, at most, a training issue. The pilots didn't have enough time to diagnose and work around the problem because it wasn't something they where trained to catch right away. If one pilot was able to figure out a solution all on his own, safely flying though this issue is possible.

      So I don't blame the pilots for making a mistake, even though they could have flown the aircraft with the malfunction, they where not told how so they didn't make a mistake, they just ran out of time before they could figure out what to do.

      • If you train the users to alter the right config file or press the right combinations of buttons then they won't delete all of their data. Stupid users.
      • by omnichad ( 1198475 ) on Tuesday April 30, 2019 @10:50AM (#58516670) Homepage

        Sure. As long as you admit that the training issue is that pilots were told they didn't need to retrain if they were certified for the 737 and were told it has exactly the same flight characteristics. This makes Boeing culpable for that "training issue.". The MCAS system was not documented in the flight manual.

        They only won't directly admit fault, because it will be used as evidence in all the lawsuits.

        • That's exactly what I'm saying. The plane is flyable with this fault, but Boeing didn't inform pilots of how to diagnose the issue or how to deal with it. And Boeing has ALREADY admitted this. They did so after the first accident and BEFORE the second.
          • by Tough Love ( 215404 ) on Tuesday April 30, 2019 @04:05PM (#58518352)

            The plane is flyable with this fault...

            Multiple faults, actually, including the stubby landing gear, the poorly positioned engines causing unwanted lift, the lack of sensor redundancy, the murderous control algorithms, and it goes on. Really, the 737 Max is a junk design and should be junked, along with the evil people who foisted it off onto an unsuspecting public.

      • by Luckyo ( 1726890 ) on Tuesday April 30, 2019 @10:51AM (#58516678)

        Issue being that this is a design problem, not a pilot training problem if your description proves true in investigations. This is because two pilots are far too busy trying to keep the aircraft from nosediving into the ground to be able to troubleshoot the problem in time.

        The extra pilot could do it because he had nothing else to do but troubleshoot, and sitting behind them he could very easily see what was going on with trim wheels.

        This sort of a thing could be labelled a "pilot training problem" in 1960s. There's a reason why a lot of planes crashed back then compared to today. Today, this sort of a problem is considered a design error, because the system as a whole is not safe enough for modern civil aviation. The sum of the issues leading to the failure is the following: system that isn't immediately obvious to the pilot + system lacks backups for critical components + system is mission critical and can cause a rapid and unavoidable crash + insufficient pilot training.

        Better pilot training will not correct for other issues, such as lack of critical component backup or system failure's obviousness to the pilot in event of a critical malfunction.

      • If one pilot was able to figure out a solution all on his own, safely flying though this issue is possible.

        Hopefully that was the only issue.

      • by alexo ( 9335 )

        This tells me that this was, at most, a training issue. The pilots didn't have enough time to diagnose and work around the problem because it wasn't something they where trained to catch right away. If one pilot was able to figure out a solution all on his own, safely flying though this issue is possible.

        If the brake hydraulics on a certain car model tend to spontaneously fail, but some drivers still manage to stop safely (by engaging the parking brake or any other means), does it mean that it is *at most* a training issue?

        I have had recall letters sent to me for far less serious issues. The proposed action was always "bring the car to the nearest dealership, and we will replace the part(s)", not "come over, we will train you what to do in this situation".

      • Comment removed based on user account deletion
        • There was no single failure.. I see multiple failures here... Not just that the equipment failed, but that the training failed and the pilots failed to diagnose the issue. Failures happen, the MCAS system could still fail with out a faulty sensor or if there are three of them, anything is possible. The idea is to layer in as many redundancies as you reasonably can, redundancies that don't share common parts. This is so it takes multiple failures of different types, to cause an accident.
      • by Zak3056 ( 69287 )

        Remember, the aircraft IS still flyable, a fact that is proven by the events leading up to the first accident, where the day before the aircraft that crashed had the identical malfunction, and the pilot who happened to be sitting in the jump seat figured it out, popped the trim system breaker and they landed safely.

        IIRC, while what you say above is true, it is only applicable to that specific instance. In the case of the second crash, the pilots apparently DID pop the trim breaker... and weren't able to re

      • by raymorris ( 2726007 ) on Tuesday April 30, 2019 @12:06PM (#58517094) Journal

        Yes, there is course of action the pilot can take when this system failure happens.
        Yes, there is a training issue in that the pilots weren't trained on what to do when this system failure happens.

        The two prior sentences are about what happens after the failure occurs, how to stay alive despite the system failure. They would not be needed if the system didn't fail.

        So there are at least five problems:
        1. Frequently, one of the two AoA sensors fails.
        2. When one AoA sensor fails, MCAS sends the plane into a nosedive, without cross-checking the other sensor or pitot tubes.
        3. The solution is non-intuitive. Other similar failures can be overcome by the pilot pushing harder. The MCAS ignores the pilot pulling the stick back hard.
        4. The pilots weren't trained on the MCAS
        5. The airlines and pilots were told they don't need training on the Max

        Boeing's statement is, imho, exactly the wrong thing to say. This MCAS thing is a string of failure after failure, problems on top of problems. "This is the result of following all of the procedures we always follow for everything" suggests that the processes they use for everything can be expected to cause a cluster fuck of problems on top of problems. The right response, imho, would be "we have looked into the underlying causes five levels deep and implemented new processes to avoid any further problems from such causes".

        • So there are at least five problems: 1. Frequently, one of the two AoA sensors fails.

          AoA sensors are mechanical devices. They can fail. I don't know that "frequently" is a proper term, however.

          3. The solution is non-intuitive.

          I'm sorry, but that's just nonsense. If you have a runaway trim pushing the nose down, the intuitive solution is to TURN OFF THE TRIM SYSTEM. That's what the emergency procedure happens to be, so even if it isn't intuitive, it is part of the pilot operating handbook, is well documented, and is taught to every pilot as part of recurring training.

          Other similar failures can be overcome by the pilot pushing harder.

          That's just asinine, and so obviously wrong as to be la

          • by jeff4747 ( 256583 ) on Tuesday April 30, 2019 @03:10PM (#58518114)

            I'm sorry, but that's just nonsense. If you have a runaway trim pushing the nose down, the intuitive solution is to TURN OFF THE TRIM SYSTEM.

            The problem is pulling the breaker to disable the entire trim system (including MCAS's input to it) leaves you unable to manually change trim because of aerodynamic forces. The backup wheels can't be turned with enough force to un-do the damage caused by MCAS.

            So now the nose is still being pointed down, and you still can't bring it back up. That's entirely on Boeing.

            That's just asinine, and so obviously wrong as to be laughable. You don't "push harder" if the nose is going down. You pull up.

            Asinine is insisting "push harder" literally means push, instead of "actuate the controls in the desired direction with more force".

            And not a single thought of "gee, the trim isn't working right, I better disable it...", which would have saved the lives of everyone on board.

            They disabled the trim system, dumbass.

            They were unable to fix the trim because the manual wheels could not apply enough force to overcome the aerodynamic forces. They then re-enabled the trim system so that they could use the powered trim adjustment as a last-ditch effort. And it worked....until MCAS fucked them over again.

            It takes no extra training to know how to turn off the trim system

            It does, however, take training to know that MCAS 1) exists, and 2) is the thing fucking over your trim, so you're gonna need to leave the powered trim system on only long enough to get to the point where you can actually use manual trim.

            And, at least four months prior to the second crash, the airlines were told about the problem, what creates it, AND HOW TO SOLVE IT.

            If it could actually be solved by the airlines and the pilots, then Boeing wouldn't be spending months re-writing MCAS and retrofitting the optional warning light upgrade for free.

    • by sycodon ( 149926 ) on Tuesday April 30, 2019 @10:34AM (#58516586)

      Any device that has an impact on control surfaces should have triple redundancy...like the Shuttle did.

      A single sensor taking out an aircraft is pretty much negligent design.

  • by 110010001000 ( 697113 ) on Tuesday April 30, 2019 @10:03AM (#58516372) Homepage Journal

    "It was designed per our standards. It was certified per our standards"
     
    I hear this a lot as an engineer. It makes no sense to me. Your "standards" are arbitrary and might be flawed. Why does following a standard mean that nothing went wrong? It is like Agile zealots who assume because they hit all their benchmarks (or whatever they call them) the product is ready.

    • by Immerman ( 2627577 ) on Tuesday April 30, 2019 @10:09AM (#58516406)

      Indeed. In fact, in a failure like this, due to obvious recklessness or incompetence, saying "it was certified per our standards" seems to be a clear admission that your standards are deeply flawed, and *everything else* you've made should be suspected to have similarly severe problems lurking under the hood.

      • by drew_kime ( 303965 ) on Tuesday April 30, 2019 @10:22AM (#58516488) Journal

        Indeed. In fact, in a failure like this, due to obvious recklessness or incompetence, saying "it was certified per our standards" seems to be a clear admission that your standards are deeply flawed, and *everything else* you've made should be suspected to have similarly severe problems lurking under the hood.

        Came here to say exactly this.

      • Yeah, but when you’re at the point in the discussion you can turn it over to your lawyers. They make their living burying such simple facts in an avalanche of legalese.

      • Or .... given the paucity of crashes with Boeing aircraft, perhaps the standards are pretty solid? Then, again, against what can the standards be compared?
        • Well, the control system that caused the problem had a pretty clear and obvious design flaw (using only one of two available relatively low-reliability sensors), in addition to the more subtle failure-loop that required pilots to be specifically aware of the differences in a system designed specifically to avoid retraining pilots.

          So, there's really only two options:
          They're lying about it having passed their usual standards.
          or
          Their usual standards are clearly insufficient, at least where software and user in

    • by Gaxx ( 76064 ) on Tuesday April 30, 2019 @10:10AM (#58516410)

      Yes - what this very much suggests is that the standards are so badly flawed that they can't be trusted and as a result anything else designed to the same standards should be considered suspect.

      Really, taking the stance that planes that crash meet your standards might not be the smartest way forwards.

    • His public response complies with the company's legal standards. It's a sad state of affairs, but understandable. Every "oops" is opening for another lawsuit or prosecution.
      • Response per lawyers standards. Boeing will be sued for zillions. Guess Insurance coverage also depends on root cause. So CEO compelled to tip toe thru the mine field. Regardless an unsafe flying situation resulted in 2 crashes killing a lot.
    • by Tablizer ( 95088 )

      "It was designed per our standards. It was certified per our standards"

      I'm not a troll, I post to my standards, and I have high standards, you low-IQ basement-dwelling idiots!

    • If this system was designed per their standards and nobody did anything wrong, there's something seriously wrong with their standards and I'm wondering what other problems may come up later. I would have been a lot more reassured if they had just said "we messed up and are taking steps so this cannot happen again".

      They are still in lawsuit avoidance mode: whatever you do, don't admit any mistakes were made! As if the size of this cockup wasn't immediately obvious to anyone with the slightest knowledge about

      • They are still in lawsuit avoidance mode: whatever you do, don't admit any mistakes were made!

        I hear this all the time and it's bullshit.

        They're going to get sued seven ways to Sunday whatever the CEO says. Those cases will be decided by evidence + expert testimony, not what the CEO says.

    • It sounds a little weasely, but that is legal language. A full statement could be that it was designed with compliance to relevant(*) safety regulations with a standard of care appropriate for passenger aircraft.

      While MCAS was a shitty design, especially in only using single sensor data and allowing infinite cycling, it was required by regulatory framework.

      The “standard of care” is professional engineering speak for the fact that not everything is designed to the highest standard possible, but t

      • While MCAS was a shitty design, especially in only using single sensor data and allowing infinite cycling, it was required by regulatory framework.

        Without the MCAS, there may have been more crashes. The root cause of MCAS wasn't the regulatory framework: it was a decision to design the 737 MAX8 such that it wasn't stable and was prone to stalling. MCAS followed from that decision and the decision to ensure that existing 737 pilots did not need re-training to fly it.

    • As an engineer, I also suspect the engineers fought against the design as it stands today but were overruled by management. I can’t see many engineers would think that a software change would be adequate to overcome the problems and that pilot override would be so convoluted.
      • As an engineer, I also suspect the engineers fought against the design as it stands today but were overruled by management. I can’t see many engineers would think that a software change would be adequate to overcome the problems and that pilot override would be so convoluted.

        I can, depending on how the design is done. if seperate groups are ppaoraching different parts of the problem and not working together to keep everyone aware of the overall design approach the system could have been designed as it was and all the engineers sign off on it. I've been involved in projects where someone outside of our group would change something and we'd not been told about the change; we caught it because we kept checking the status of inputs into our system and would catch those that just ma

    • The FAA practically let them self-certify their own aircraft.

    • A problem with complex builds that needs more then a small group of people is the fact that no one really really designing for the big picture.

      They say to the Engineer, I need a part to handle tolerances of X and Y that fits in dimensions of AxBxC.

      Now if this part fails because the part handled conditions of X-5% or Y+5% it isn't the Engineers fault for not doing his job, because the part met the specs given. Trying to make a part that exceeds the specs could add undo cost to the project, or there is plan

    • "It was designed per our standards. It was certified per our standards" I hear this a lot as an engineer. It makes no sense to me. Your "standards" are arbitrary and might be flawed. Why does following a standard mean that nothing went wrong? It is like Agile zealots who assume because they hit all their benchmarks (or whatever they call them) the product is ready.

      This is a classic behavioral response that results from a lack of unfavorable outcomes creating the impression that the actions (standrads in this case) are sufficient to prevent an adverese outcome. Repeatedly doing something without an adverse outcome does not imply the action is safe; but peopel tend to erroneously, in some cases, assume it does.

    • by Luckyo ( 1726890 )

      True, but this was a news conference. Overwhelming majority of reporters are specialists in language interpretation and have little to no understanding of how engineering works. Their profession is about observing something, and then packaging it to be sellable to the audience of people who are also utterly ignorant of the topic. To do this, they do not need to be correct in any of their assessments, they merely need to package it so that it looks like they are to the uninformed. They can even be diametric

    • It is worse than that. Managers that state things like that were likely told their standards or industry norms were flawed. However a manager is personally safe if they, even knowingly, follow a flawed standard. If, however, they try and fix things and something goes wrong then they are personally responsible. So while bad decisions were made at Boeing, each individual manager did the thing that was most beneficial for their career.
    • by AmiMoJo ( 196126 )

      ISO 9000.

      It doesn't matter if you fuck up, as long as you documented your fuck up and followed the fucked up procedure.

      Dissipate blame over such a large area that it's impossible to hold anyone accountable. Cover your arse with bits of paper showing you fucked up according to procedure.

      If there are any consequences for this guy they will be for the drop in stock price or fall in quarterly profits, not for getting people killed.

    • by pezpunk ( 205653 )

      sounds like the corporate version of "we were just following orders".

  • Comment removed (Score:4, Informative)

    by account_deleted ( 4530225 ) on Tuesday April 30, 2019 @10:09AM (#58516404)
    Comment removed based on user account deletion
    • Comment removed (Score:4, Informative)

      by account_deleted ( 4530225 ) on Tuesday April 30, 2019 @10:22AM (#58516486)
      Comment removed based on user account deletion
    • And the power of the system will be limited, so that the pilot can always pull back on the control column with enough force to counteract any automatic nose-down movement.

      This one is interesting. This used to be the distinction between Boeing's automated systems and Airbus'. When Airbus switched to a fly-by-wire system, the complaint among pilots was that the computer was flying the plane, and the pilot was merely making requests to the computer. It was implicated in several Airbus crashes (pilot attemp

      • by Zak3056 ( 69287 )

        I agree it's interesting, but for a different reason: apparently, MCAS' entire purpose is to prevent the aircraft from getting into an unrecoverable stall, and the fix is to allow the pilot to apply control inputs that could put the aircraft into an unrecoverable stall. They may as well just shut off MCAS and call it a day.

  • by amorsen ( 7485 ) <benny+slashdot@amorsen.dk> on Tuesday April 30, 2019 @10:11AM (#58516416)

    Sensible companies do not let the Chairman of the Board and the CEO be the same person.

  • by Going_Digital ( 1485615 ) on Tuesday April 30, 2019 @10:15AM (#58516440)
    So engineering a single point of failure is not a flaw, mmm.
  • by bit trollent ( 824666 ) on Tuesday April 30, 2019 @10:15AM (#58516442) Homepage

    I'm amazed that Boeing's CEO is so eager to proclaim that their corporate culture and safety practices don't face significant flaws.

    Every once in a while we all mess up. Companies let short term profits erode their commitment to safety, and in the worst case people die. When Blue Bell ice cream killed less then 10 people in the hospital, they learned a lesson beyond the initial incident and made a change to their safety culture.

    What Boeing is telling us is that the same flawed and unsafe corporate culture that killed 300 people is still making the same mistakes, and will continue trying to sneak defective airplanes past the FAA, whom they have long since turned into a corporate lapdog.

    Boeing has lost all credibility. The FAA has lost all credibility, as foreign countries turn to European and Chinese regulators now that the former gold standard has disgraced itself.

    I don't see how any prospective plane purchaser can ignore the culture of slapdash design and workmanship that let Boeing transition from a safe airplane manufacturer to the leading producer of planes that fly themselves into the ground at high speed while the pilots rush through incomplete documentation while frantically pulling up on the controls.

  • Using basic predicate logic:
    1. "our design and certification processes that consistently produce safe airplanes," i.e. A = "design standards", B = "safe airplanes", claim A(x)->B(x)
    2. "It was designed per our standards. It was certified per our standards." A(737MAX) = True -> B(737MAX) = True
    3. Observation: B(737MAX) = False

    Hence A(x)->B(x) is not valid and they need a new set of standards.

  • by Bugler412 ( 2610815 ) on Tuesday April 30, 2019 @10:26AM (#58516516)
    yet they still designed a safety critical system with a single point of failure (the angle of attack sensor) Is that part of their "design and certification processes that consistently produce safe airplanes"? sigh
  • Over decades, with intense fascination, I've tried to learn as much as possible about the sociology of technology development.

    To me, it seems obvious that the problems with the Boeing 737 MAX-8 were caused by a lack of social connections between engineers and between engineers and managers.

    To me, considering all I have read, it seems that the Boeing CEO was extremely disconnected and dis-interested in the development of that aircraft. The 737 MAX-8 is manufactured in the Boeing Renton Factory [wikipedia.org]. The Boe
  • He has to say this for legal reasons.

    Admission of guilt in face of impending lawsuits (families, shareholders, airlines, etc.) wouldn't look good.

    Of course he's lying, as evidenced by the fact that they are fixing the software and making a previously optional sensor a standard feature.

    It's not like he'll face criminal charges, he's got too much money. His job might be the most that's at stake (and I'm sure the golden parachute is of a very heavy fabric).

  • by schwit1 ( 797399 ) on Tuesday April 30, 2019 @11:03AM (#58516764)

    Boeing lied when they said the 737 Max was a simple evolution of the 737 family.

    Boeing lied when they turned off sensor malfunction alerts on 737 Max — and didn’t tell airlines or FAA. [hotair.com]

    Boeing lied by omission when they designed a flight control system that allowed the MCAS to over-ride pilot inputs.

    To say the crashes were accidents is a lie. The crashes were easily predictable in a short time.

    I want to see the DOJ issuing indictments of FAA and Boeing personnel.

    • Boeing should be forced to refund all the airlines, all existing MAXs should be scrapped. And they should be forced to do a ground up redesign for a stable airplane that does not need MCAS. That is appropriate punishment.

  • "It was designed per our standards. It was certified per our standards."

    Yet two airplanes have crashed without a satisfactory explanation, is that also per Boeing standards? Insisting that you did everything right is not enough when the end result is hundreds of corpses. Now it's up to Boeing to deliver on what went wrong where and how to prevent it in the future, that cannot happen with a starting point that insists on everything being absolutely rosy and perfect.

  • by Ecuador ( 740021 ) on Tuesday April 30, 2019 @11:26AM (#58516900) Homepage

    we followed exactly the steps in our design and certification processes that consistently produce safe airplanes

    The generally accepted definition is that "Safe" is a plane that does not crash & burn without it being the pilot's fault or a maintenance issue.
    The Max is not "safe" under that or any other definition that people would use.
    Hence that design and certification process DOES NOT *consistently* produce safe airplanes, as at least one recent plane was not safe.

    His lawyers just told him whatever you do do not admit fault, so we can have a fighting chance in court...

  • Most combat aircraft are unstable, but, by DESIGN. Instead of spending the money to redesign the 737, boeing moved the larger engines more forward, which changed the COG of the aircraft, hoping the software system would keep the aircraft in trim. 2.5 degrees doesn't sound like much on the entire horizontal stabilizer, but, THAT IS A LOT! Boeing screwed up, and cut corners it appears....and got caught!
  • Obviously avionics goes through way more scrutiny and testing than your average phone app or SaaS platform. But, being a systems person working with developers, all too often I see the mentality, "Ship it, we'll patch in production, it doesn't matter if there are bugs, all software has bugs."

    Going back to waterfall and multi-year design isn't an option, but IMO there has to be something better than almost no planning whatsoever, then tying your developers into name-your-CI/CD-toolchain and putting intense p

  • Forget about replacing your factory workers with robots. When do we sack Boeing management and replace them with AI?

    Management's job is to produce and maintain work procedures and processes for employees to follow. AI can be trained by back propagating errors to modify its rule base or neural net to improve performance and minimize future errors. If Boeing management doesn't recognize the events over the past months as valid training data for policy/process modifications, it needs to be unplugged and repla

  • Lessee (Score:4, Insightful)

    by Ol Olsoc ( 1175323 ) on Tuesday April 30, 2019 @05:00PM (#58518562)
    So you take an airframe that is already low to the ground, put a much bigger engine with a much greater intake, so you move the engine up and forward, changing the center of thrust, which makes the plane really want to pitch up with thrust, and the further you pitch up, the more the nacelle starts generating lift, so you attempt to keep the airframe from a positive stall feedback loop with software.

    Then to top it off, you make a godamned display that might save your pilots and passengers lives........... an option that you have to pay for.

    This is an unstable and unsafe airframe - an airframe that would rather stall and crash than fly. It is surprising that the Boeing CEO is not in jail.

  • by BBF_BBF ( 812493 ) on Wednesday May 01, 2019 @04:03AM (#58520566)
    Famous last words describing a woefully inadequate systems design that fails to work properly in the real world.

Avoid strange women and temporary variables.

Working...