WP is usually paywalled, but I think they allow some links from some paths. That one worked for me, but it may have because of the page I started from.
Most of the big-name lefty sites (WP, NYT, LaTimes, SFChron) are paywalled. Few of the more conservative sites are. Curious.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
I presume you mean change of altitude. That would also be an error prone and slow acting indication.
In the presence of conflicting data, knowing something is "wrong" is insufficient. You need to be able to determine which data to ignore and which to trust. In the general case that is a difficult problem to predict and solve /correctly/.
"Sensor fusion" from widely different sensors is a difficult problem to solve correctly.
Relying on a single AoA sensor is going to be "problematic" for Boeing's reputation.
Not curious at all.
They get money from sources that want to stay in the shadows. You are the product being sold to those sources.
We can be sure that a large corporate producer of aircraft has 'serious and deep' problems, that's why they hire folk to solve 'em.
As for 'serious and deep' problems that can pull an aircraft untimely to the ground, though, it is common to not certify (nor sell/deliver) aircraft when THOSE are likely. For this year, the news is, Boeing is correcting those problems at high priority, and certification authorities are doing the best they can as well. If FAA and NASA and NTSB can be trusted, the results should be satisfactory.
If you're thinking 'Boeing is jinxed' and that the CEO must be sacrificed on an altar, that's just wacky.
snipped-for-privacy@highlandsniptechnology.com wrote in news: snipped-for-privacy@4ax.com:
So, were I like you, I would use some lame ploy by saying the opposite of what is true so that you saying I am always wrong goes against you.
No. I will not say that you are intelligent, because I have not been wrong, and you *are* most definitely a self imposed utter retard.
As I understand it, by moving engines forward in 737 MAX 8, it caused a stronger nose up attitude when applying power.
One design goal with MAX was that pilot having ordinary 737 rating could use MAX with minimal conversion training, making it easier for airliners switch to MAX , One way to make the MAX behave the same way as older 737 models, the MCAS was introduced.
Did the MAX conversion training also include disabling a malfunctioning MCAS ?
When working with critical control systems, it is important to verify that you get dependable sensor inputs, otherwise this is just a GIGO (Garbage In Garbage Out) system. Typically this requires multiple redundant (sensor) system to detect that something is wrong and at least a triple redundant systems to vote out an erroneous sensor, In redundant systems, it is important to avoid common single point of failures.
If relying on a single sensor only, use all means to check that the signal quality is inspected. In case the AoA sensor that it works consistently on the ground and at a nose up attitude before takeoff.
Even a programmer used to make critical systems should have noticed the lack of redundancy and alarmed his/her bosses of the bad design.
The 737MAX, unlike the current Starliner, is not a fully automated system without human pilots. There is no need to solve all possible problems, in certain cases it would be appropriate to switch off and issue an alarm. The MCAS system was badly designed not to do this, even when it would have been possible to take auxiliary measurements to detect that something is going wrong.
What is even worse is the reasoning behind that: the MCAS system was only designed in to save money for the clients (and thus make the product easier to sell) by skipping the requirement for pilot training. It is a bolt-on system that performs a noncritical function. When it would have been an essential system, it would also require corresponding training and thus defeat the original purpose.
So the fault is both with Boeing (for designing a system that does not perform self-evaluation of its actions and does not turn off when it is clearly causing trouble) and with the regulators (for setting up the maze that includes typerating and the classification of essential and non-essential systems).
The requirements were determined by the nature of the (MCAS) system. MCAS is a bolt-on feature that was added to make the 737MAX a part of the same typerating family as the 737NG and earlier types, so the 737 pilots could fly it without additional training.
Such a system is in a category where it should use only simple methods not including things like redundant sensor inputs with automatic or manual selection (and manual overrides). If it would be an essential system it would be able to do that, but at the same time it would require the pilots to be informed about it and trained in its operation, thus defeating the original purpose.
The pilots were not even told about the addition of the MCAS system! When MCAS started to rotate the trimwheels, they knew nothing more than the existing systems that do automatic trim on the 737 would be doing that. And they are supposed to know how to turn off those systems. However, they were never alerted that there now is a new system that is doing those trim actions and remind them that it should be turned off when it going awry.
They were thinking about a new plane but it was far ahead in the future and a large customer was considering moving over to Airbus which already had announced such a plane (A320 NEO). So a quick move was required and it was decided to design another 737 model with the newer engines.
The big problem was not this plane, but that they decided they wanted to keep using the same typerating (probably required to keep that customer, as that could be offered as an advantage over the Airbus).
Not really. It is a requirements specification error at a systems level which is *FAR* more serious. The software did what it was intended to do
- the problem was that included flying into the ground at high speed to avoid stalling in the event of of a single defective sensor.
At least some people in Boeing tried to point that weakness out but the marketing imperative to pretend it was just another 737 with no extra training requirement won out over sound engineering decisions.
FAA were asleep on the job and/or in Boeing's pocket.
-- Regards, Martin Brown
:
ing-starliner-launch/
not bad coding
s that the requirements
g that pointed that
the airplane on the
sensors are
e told, when
n't write
only becomes a problem with poorly trained foreign pilots with insufficient aptitude for flying complex systems.
ave
angle
nstead
et,
s the
are
being
teps,
ctly
ot
oat
down
the
at out
rily
gh
g manuals.
is you
Simple
artly. The American pilots who experienced this same phenomenon stated all they had to do was shut the friggin MCAS off and fly manually.
shows
ly 100% on every auxiliary flight control possible, and can barely fly the aircraft manually, kind of like a blind person driving a self-driving car.
ilots
, but is actually wrong about everything:
total bs corrupted by FAA and NTSB bureaucracy.
s and took control of the plane themselves. Obviously none of these flights crashed."
s American pilots, not even close. And failure to recognize this fact....ne ver mind... backwater mindless people like you aren't worth the time it tak es to finish a sentence.
attacks on the foreign pilots as well as anyone who doesn't agree with you is unjustified.
I believe the answer is no, but it exhibits itself as runaway trim and all pilots are trained in how to deal with that. It doesn't matter what is forcing trim to extreme, incorrect positions. Could be a stuck switch, short, etc or MCAS. MCAS went amok
3 times that we know of. In the first case, the Lion Air flight immediately before the flight that crashed, a jump seat pilot correctly identified the runaway trim and told the pilots flying what to do, which is turn off electric trim, then trim manually. They did and the plane flew on to it's destination.The second crash, Ethiopia, the young co-pilot with just a couple hundred hours, correctly indentified the runaway trim and followed the Boeing procedure that was put out after the first crash. He turned off the electric trim, but then said he was unable to move the trim wheels by hand, which is what will happen if the trim is way out and the plane is going too fast.
Yes and it's shocking Boeing didn't realize this.
Maybe. Depends on what the programmer knew about what effect MCAS would have on the airplane if it malfunctioned. They may have believed that it would be minor. Boeing did classify it as serious, but not catastrophic and apparently that was what enabled the use of just one input.
Yes, but still runaway trim can happen in any plane with electric trim and all pilots are trained how to deal with it. Turn off the electric trim with the switches right at their side, next to the spinning trim wheels and indicators. Then trim manually using the trim wheels. Electric trim can run amok from a bad switch, short, etc, in addition to MCAS. Apparently those other things very rarely happen or we should have seen other crashes.
Missing in all this, still, is what was the root cause? Apparently it was a bad AOA, not the cabling, not the circuit that reads it, etc. But then what? One would think we would have heard why two of these failed in brand new planes. I think one theory was that a bird strike may have happened in Ethiopia.
ner-laun
ding
requirements
nted that
ne on the
re
en
rammers did
he
s.
mes as
ch
oftware
tware process as it related to MCAS was quite informal to put it mildly.
at's how they ended up with 2.5o trim capabiity.
d,
Obviously not, with the 787 being a good example. The 797 is under development right now, totally new composite design, mid-size plane.
It was also that their CUSTOMERS would not have to train and certify it's pilots on a different plane. And also, moving pilots from an existing 737 on one flight to a radically different 737 on the next flight isn't ideal. There were benefits to keeping them very similar. The competing issues of keeping it the same, but making it more fuel efficient, different engines, is what made it problematic. I'm not sure at what point, ie how far along in the development they were, when they realized they needed MCAS.
Not just that one customer, but all customers flying the world;s most popular plane.
The issue here is that this is so STUNNINGLY bad, so OBVIOUS, that until someone proves that a whole lot more isn't screwed up at Boeing, call me deeply concerned. This wasn't some obscure failure, some unknown manufacturing defect that caused an engine failure or similar. It was Boeing designers and the FAA thinking it was OK to build a plane that relies on just one sensor that can so easily cause a plane to dive itself into the ground.
I didn't say jinxed, which implies something supernatural. I said my faith in Boeing has been shattered, I prefer to fly on Airbus now and that CEO should have been gone long ago. BTW, did you see the new screw up? They can't build spaceships either.
Strange, but not surprising.
One has to also evaluate how many more possible problems and f
My points, especially those about expecting coders to understand aircraft dynamics, are more general than just the 737MAX.
Nonetheless, the points you make are pretty sound.
To some extent I have sympathy with the regulators. Some people seem to expect that the regulators should know more about the specific aircraft than the manufacturers. That's clearly unrealistic.
OTOH, they appear to have, for whatever reason, allowed too much de-regulation and self-certification. But such a "light touch" /because/ "the market knows best" has been an unchallenged political imperative for the past couple of decades.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.