Breaking news: Ethiopian Airlines flight has crashed on way to Nairobi

[TeeVee]

i think that everyone would agree that today's ac are immensely complex beasts. i imagine that very few if any ac technicians can diagnose a problem and come up with the solution in a matter of minutes. i feel pretty safe in saying that ac are far more complex than your average car/motorcycle that use minimal electronics. i know from personal experience that diagnosing and fixing problems in these relatively simple vehicles sometimes confounds dealership mechanics that have full factory support at their fingertips. i had a 2006 bmw 528 that had an electronic gremlin that not even bmw's factory rep could figure out (ending up in bmw buying the car back from me).

what evan is proposing (and i'm not for or against it) seems to be that pilots (whose primary job is flying), have a deep understanding of systems and inter-operativity, to the point that when something goes wrong during a critical phase of flight, they barely blink to work it out. and yeah, i get the part of "one fly the other work on the issue." that appears to have failed once or more times.

who knows...

[ATLcrew]

Yeah, I'm not suggesting that pilots must have actual electrical engineering knowledge of the avionics, but rather a general engineering knowledge of how the systems come together and interact and what their specific functions are in the grand scheme of things. I'm particularly concerned that they understand the ramifications of a given failure condition and the failure 'cascade' that might result.

So, in your example, let's say you are dispatched with a MEL'd ELAC1, and then you lose the green hydraulic system in flight. Do you lose the left aileron? Or let's say same dispatch scenario and you then have some reason to reset ELAC2 in flight. Will doing that suddenly put you in alternate law without ailerons?

That's the level of engineering knowledge I'm talking about.

Ah yes, the glorious 5-page 27-93-01-2! The answer is no, we won't lose the left aileron. If we only lost the green pump, the PTU will take over, and the green system will remain pressurized. If we lost the fluid, the aileron will still be powered by the blue system, HOWEVER, we will be in alternate law. The answer to your second question is yes, it will go into alternate law momentarily while I'm resetting ELAC 2, but will come right out of it, if the reset is successful. Even if it doesn't, that's not any big deal, the airplane rolls beautifully on spoilers alone, even if we're down a spoiler per wing. Frankly, the rather extensive pre-flight procedure for that MEL is more of pain than not having ailerons. Probably why I've never seen that MEL used, it's easier to just replace the ELAC, takes all of 40 minutes, if that.

Does every pilot actually NEED to know that? Maybe, maybe not. The ECAM and the SD will tell you what's working and what's not, you don't necessarily need to know why, as long as you know what to do about it.

BTW, Evan, I'm surprised you didn't ask me if we'd lose BOTH ailerons in that Green system scenario. Now, THAT would have shown some real engineering brilliance on your part.

[Evan]

I actually agree to an extent. Personally, I have long advocated for more thorough systems training, particularly since it's sort of "my thing". I like systems, I've been known to read systems manuals for fun, including those of aircraft I've never flown, but that's just me. That said, there are pitfalls that come with too much information, I'll just touch on a few.

Firstly, I don't see the benefit of knowing things I cannot address from the flight deck. E. g. it is of little use to know the exact signal pathways through SEC or ELACs since there is nothing I can do to affect those pathways. I can reset an ELAC, I can turn it off, but there's nothing I can do with it internally. Do I need to know what controls each computer affects? Absolutely. But the "engineering" of the computer itself? I suppose it's cool to know for the "gee-wiz" factor, but little else. I need to know what happens if an MMR fails. I don't need to know how it's put together or even what it looks like (and I don't).

Yeah, I'm not suggesting that pilots must have actual electrical engineering knowledge of the avionics, but rather a general engineering knowledge of how the systems come together and interact and what their specific functions are in the grand scheme of things. I'm particularly concerned that they understand the ramifications of a given failure condition and the failure 'cascade' that might result.

So, in your example, let's say you are dispatched with a MEL'd ELAC1, and then you lose the green hydraulic system in flight. Do you lose the left aileron? Or let's say same dispatch scenario and you then have some reason to reset ELAC2 in flight. Will doing that suddenly put you in alternate law without ailerons?

That's the level of engineering knowledge I'm talking about.

[3WE]

***my philosophy has always been to reduce the airplane to a Cessna, make it do what I need it to do and then restore the automation (if appropriate). I don't necessarily need to know exactly how the automation failed. All I care is is it doing what I want it to***

What the phigoid does phulosophy have to do with anything?

Follow the phugging procedures!

[ATLcrew]

I guess the answer to that question is: how much intellectual ability can we expect from pilots?

If I had it my way, becoming an airline pilot (and thus the protector of hundreds of live every day) would require a strenuous discipline both in traditional piloting knowledge and engineering knowledge. The reason for this lies in the new reality of the job. There is no longer a flight engineer. The decision to remove the flight engineer involved the assumption that the pilots could take over those responsibilies through the digital flight deck provisions. Those provisions are meaningless if the pilots do not also have the flight engineering knowledge that has been removed with the flight engineer.


Where do you draw the line? In terms of systems and architecture, pilots should know everything that is contained in the FCOM (MCAS should have been contained in the FCOM). Therefore, the FCOM (and FCTM's) has to be written to include everything pilots must know for both normal and abnormal operations. Knowing the FCOM from cover to cover cannot be considered "too much information". Before being given a type-rating, pilots should be extensively tested on FCOM knowledge. In addition to checkrides, pilots should be periodically quizzed on systems as described in the FCOM and the ramifications they have in abnormal situations. Moreover, pilots should feel a moral responsibility to be constantly augmenting their training with a general study of the aircraft and abnormal ops and upset recovery piloting techniques. I can't say it enough, piloting should come with a dreadful sense of responsibility.

I actually agree to an extent. Personally, I have long advocated for more thorough systems training, particularly since it's sort of "my thing". I like systems, I've been known to read systems manuals for fun, including those of aircraft I've never flown, but that's just me. That said, there are pitfalls that come with too much information, I'll just touch on a few.

Firstly, I don't see the benefit of knowing things I cannot address from the flight deck. E. g. it is of little use to know the exact signal pathways through SEC or ELACs since there is nothing I can do to affect those pathways. I can reset an ELAC, I can turn it off, but there's nothing I can do with it internally. Do I need to know what controls each computer affects? Absolutely. But the "engineering" of the computer itself? I suppose it's cool to know for the "gee-wiz" factor, but little else. I need to know what happens if an MMR fails. I don't need to know how it's put together or even what it looks like (and I don't).

Secondly, too much technical knowledge can create the added risk of going down the wrong path to address a failure that may be better addressed by the relevant system just being isolated (de-powered, disconnected, whatever). In an emergency situation, we likely won't have time to do "engineer-level" troubleshooting. If something is misbehaving, put it out of business and work with what's left. With regards to specifically automation failures, my philosophy has always been to reduce the airplane to a Cessna, make it do what I need it to do, and then restore the automation (if appropriate). I don't necessarily need to know exactly how the automation failed. All I care is is it doing what I want it to or is it not?

Thirdly (and this is particularly important, I think), too much knowledge can create even more expectation bias, that is to say, one may have seen a failure that was caused by a thingamabob with which he/she is familiar through amazing "engineering knowledge", and the next time he/she sees this failure, he/she may try to address it from the standpoint of the aforementioned thingamabob, while in fact the new failure was caused by a whatchamacallit. So, there's that.


On a side note, I wouldn't deify F/Es too terribly much, they weren't necessarily system gods either, especially since the majority couldn't wait to get out of that seat and move forward.

On another side note, better-written and better-designed FCOMs would be a big help. Both Bombardier and Airbus Frenglish are painful to deal with.

[Gabriel]

Only somehow related (no MCAS involved, and the -500 is not an NG let alone a MAX), but interesting:

Incident: UTAir B735 at Surgut on May 25th 2019, trim runaway

http://avherald.com/h?article=4c87a767&opt=0

[BoeingBobby]

Spoken like a person who puts his life in the hands of professionals. Feel free to add your own wisdom.

I will have to discuss it with LH, I'll get back to you.

[Evan]

Spoken like the professional that you are right?

Spoken like a person who puts his life in the hands of professionals. Feel free to add your own wisdom.

[BoeingBobby]

I guess the answer to that question is: how much intellectual ability can we expect from pilots?

If I had it my way, becoming an airline pilot (and thus the protector of hundreds of live every day) would require a strenuous discipline both in traditional piloting knowledge and engineering knowledge. The reason for this lies in the new reality of the job. There is no longer a flight engineer. The decision to remove the flight engineer involved the assumption that the pilots could take over those responsibilies through the digital flight deck provisions. Those provisions are meaningless if the pilots do not also have the flight engineering knowledge that has been removed with the flight engineer.

Case in point: In the aforementioned MD-81 incident, I feel pretty certain that a dedicated and competent flight engineer would have been aware of the ATR issue and would not have blown those engines.

Where do you draw the line? In terms of systems and architecture, pilots should know everything that is contained in the FCOM (MCAS should have been contained in the FCOM). Therefore, the FCOM (and FCTM's) has to be written to include everything pilots must know for both normal and abnormal operations. Knowing the FCOM from cover to cover cannot be considered "too much information". Before being given a type-rating, pilots should be extensively tested on FCOM knowledge. In addition to checkrides, pilots should be periodically quizzed on systems as described in the FCOM and the ramifications they have in abnormal situations. Moreover, pilots should feel a moral responsibility to be constantly augmenting their training with a general study of the aircraft and abnormal ops and upset recovery piloting techniques. I can't say it enough, piloting should come with a dreadful sense of responsibility.

That said, with an apparent pilot shortage around the world, and with intellectually-driven people preferring other paths in life, this could prove to be a problem. But the solution lies in vetting pilots as best we can for intellectual ability and training them well beyond the minimal 'need to know' standard of the puppy-mills and third-world flight schools.

Spoken like the professional that you are right?

[Evan]

i don't totally disagree with you evan, but you also have to keep in mind the "information overload" factor. in the last month or so, i heard some aviation "expert" rambling on about what if anything boeing did wrong. he agreed that boeing should've disclosed mcas more clearly and develop a training program for it. he also defended boeing by saying that while aircraft have become incredibly complex, there really is no need for pilots to be taught and expected to recall the function of every single system down to the bits and bytes. for example, does an airbus pilot really need to understand the details and decision process the flight computers go through in prohibiting a her/him from going ballistic? i doubt it. but the pilot does need to be told, "hey, no matter how hard you yank that sidestick back, you will never achieve more than x degrees of nose up attitude, and the FCC will decide what that degree will be given the then current situation."

the real question is, at what point does overload begin?

I guess the answer to that question is: how much intellectual ability can we expect from pilots?

If I had it my way, becoming an airline pilot (and thus the protector of hundreds of live every day) would require a strenuous discipline both in traditional piloting knowledge and engineering knowledge. The reason for this lies in the new reality of the job. There is no longer a flight engineer. The decision to remove the flight engineer involved the assumption that the pilots could take over those responsibilies through the digital flight deck provisions. Those provisions are meaningless if the pilots do not also have the flight engineering knowledge that has been removed with the flight engineer.

Case in point: In the aforementioned MD-81 incident, I feel pretty certain that a dedicated and competent flight engineer would have been aware of the ATR issue and would not have blown those engines.

Where do you draw the line? In terms of systems and architecture, pilots should know everything that is contained in the FCOM (MCAS should have been contained in the FCOM). Therefore, the FCOM (and FCTM's) has to be written to include everything pilots must know for both normal and abnormal operations. Knowing the FCOM from cover to cover cannot be considered "too much information". Before being given a type-rating, pilots should be extensively tested on FCOM knowledge. In addition to checkrides, pilots should be periodically quizzed on systems as described in the FCOM and the ramifications they have in abnormal situations. Moreover, pilots should feel a moral responsibility to be constantly augmenting their training with a general study of the aircraft and abnormal ops and upset recovery piloting techniques. I can't say it enough, piloting should come with a dreadful sense of responsibility.

That said, with an apparent pilot shortage around the world, and with intellectually-driven people preferring other paths in life, this could prove to be a problem. But the solution lies in vetting pilots as best we can for intellectual ability and training them well beyond the minimal 'need to know' standard of the puppy-mills and third-world flight schools.

[TeeVee]

i don't totally disagree with you evan, but you also have to keep in mind the "information overload" factor. in the last month or so, i heard some aviation "expert" rambling on about what if anything boeing did wrong. he agreed that boeing should've disclosed mcas more clearly and develop a training program for it. he also defended boeing by saying that while aircraft have become incredibly complex, there really is no need for pilots to be taught and expected to recall the function of every single system down to the bits and bytes. for example, does an airbus pilot really need to understand the details and decision process the flight computers go through in prohibiting a her/him from going ballistic? i doubt it. but the pilot does need to be told, "hey, no matter how hard you yank that sidestick back, you will never achieve more than x degrees of nose up attitude, and the FCC will decide what that degree will be given the then current situation."

the real question is, at what point does overload begin?

[Evan]

Was watching the Weather Channel yesterday which regularly have plane crash shows on which are pretty good. Anyway I was somewhat thunderstruck to watch a show about Scandinavian Airlines Flight 751 as there seems to have been a very comparable issue as with the MAX and MCAS.

While not a direct cause of the accident, the fact that the crew was insufficiently knowledgeable in how to handle engine surging due to an unknown (to them) feature of the aircraft seems like deja-vu-all-over-again:

"Another contribution to the accident was insufficient training of the crew: they were not trained in restoring engine operation after they repeatedly surged. There was no simulator or other training on the engine surging problem. Secondly, they were not informed about a pre-installed automatic thrust system (Automatic Thrust Restoration or ATR). The reason for this lapse of information was that there was no knowledge of ATR within SAS. However, ATR was described in manuals by the aircraft manufacturer which every operator is obliged to know. Even though the system was developed for use in procedures not applied by SAS, a sufficiently careful study of the manuals should have led to SAS noting the system and training its pilots in its function."

Full Wiki on the accident is at:

https://en.wikipedia.org/wiki/Scandinavian_Airlines_Flight_751

I guess this bolsters my confidence that the MAX issues can be resolved since obviously the MD-81 and its off-spring survived quite successfully after this accident (although admittedly there were no casualties in this one which is amazing in and of itself).

Chris

One thing it has in common is that the incident occurred at very low altitude where there is not very much time and pilots cannot be expected to always think and act reliably. You could argue that, regardless of the specifics of the MD-81, the initial goal in a compressor stall is to reduce the EPR, and CRM dictates that while one pilot is concentrating on flying, the other is working the engine problems (as Sullenberger and Skiles did). Therefore the PNF should be closely monitoring the engine instruments, so if the N1 comes back up despite the thrust reduction (due to the unknown ATR system) or the EPR is not reducing as expected, the throttles need to be brought back further to avoid destruction. More training on engine recovery procedures might have prevented it, but I think that assumption is going to vary by the specific pilot's ability to maintain a clear head in a critical situation at such low altitude. Essentially, ATR was a stealth factor and you don't wan't stealth factors, but the goal of any compressor stall is to obtain a result in the engine instruments so the focus must be placed there and thrust must continue to be modified, for whatever reason, until the goals are achieved. Otherwise, you break the engines.

The main dissimilarity here is that the system WAS described in the FCOM. I really can't fathom why a person entrusted with a passenger airliner wouldn't study the FCOM and professionally familiarize themself with ALL the aircraft systems, but it seems there are a good number of pilots who just want to skip the homework and jump in and fly the damn thing with their almightly universal airmanship. IMHO, there is no excuse for a MD-81 pilot to not be aware of an MD-81 system behavior that is described in the FCOM, even if that pilot is never given training on it. IMHO, there is an inherent responsiblity for a pilot to take it upon themselves to study the manual as if they were monks and it were was the bible and know the aircraft they are commanding. IMHO this is not asking too much.

But in the case of the 737-MAX, the stealth factor wasn't to be found in the scriptures.

[Chris K]

Was watching the Weather Channel yesterday which regularly have plane crash shows on which are pretty good. Anyway I was somewhat thunderstruck to watch a show about Scandinavian Airlines Flight 751 as there seems to have been a very comparable issue as with the MAX and MCAS.

While not a direct cause of the accident, the fact that the crew was insufficiently knowledgeable in how to handle engine surging due to an unknown (to them) feature of the aircraft seems like deja-vu-all-over-again:

"Another contribution to the accident was insufficient training of the crew: they were not trained in restoring engine operation after they repeatedly surged. There was no simulator or other training on the engine surging problem. Secondly, they were not informed about a pre-installed automatic thrust system (Automatic Thrust Restoration or ATR). The reason for this lapse of information was that there was no knowledge of ATR within SAS. However, ATR was described in manuals by the aircraft manufacturer which every operator is obliged to know. Even though the system was developed for use in procedures not applied by SAS, a sufficiently careful study of the manuals should have led to SAS noting the system and training its pilots in its function."

Full Wiki on the accident is at:

https://en.wikipedia.org/wiki/Scandinavian_Airlines_Flight_751

I guess this bolsters my confidence that the MAX issues can be resolved since obviously the MD-81 and its off-spring survived quite successfully after this accident (although admittedly there were no casualties in this one which is amazing in and of itself).

Chris

[Bluefalcon]

And the bad hits just keep coming:

https://www.live5news.com/2019/05/23/another-former-boeing-sc-flight-line-worker-claims-he-was-fired-reporting-safety-concerns/

NORTH CHARLESTON, S.C. (WCSC) - For the second time in two weeks, Boeing has been named in a wrongful termination lawsuit filed by a former employee at the company’s plant in North Charleston....

Hollensteiner protested the practice of the alleged parties but participated in fear of losing his job, according to the lawsuit. The lawsuit details one incident in which Hollensteiner claims a senior manager told him, “Kurt, I need you to get these stamps back into compliance" which he claims the manager knew was a violation of FAA compliance.

That same manager would later fire Hollensteiner for falsifying company records, according to the lawsuit. Hollensteiner claims he was fired for his safety complaints regarding tool accountability and involvement in the union.

He is seeking actual, compensatory and punitive damages.

Hollensteiner becomes the third former flight line worker at the North Charleston plant to sue Boeing in 2019, joining Rich Mester and Liam Wallis.

[3WE]

The swiss cheese model applied to a single point of failure:

I totally agree that there's some painfully lacking redundancy...

HOWEVER

this is NOT a single point of failure. It has numerous layers of swiss cheese.

It needs:
-AOA failure
-Pilots totally forget the way most pilots adjust trim during 99% of their hand flying
-Pilots not noticing (or not responding to) big-ass, noisy trim wheel zooming along in the nose-down direction.
-Pilots, after 15 min suddenly deciding to LET the plane trim into full nose down (Lion Air)
-Pilots too slow to hit Bobby's cut off switches.
-To happen at the worst possible time
-Needs to happen to pilots who lack confidence that their power and attitude are healthy
-Needs to run long enough to get the trim wheel mechanically friction locked
-Needs to trigger an airspeed uncertainty

As much as I don't like the design (especially when Gabriel trashes it), I think there may be a valid argument that the plane probably would not have crashed if Boeing Bobby or maybe even one of us over-pontificating-PPL-amateurs was flying...

On the one hand, you can argue that it was a crash waiting to happen; however, I think you need a semi-rare dual mental brain lock to crash (NO CREW ADMONISHMENT INTENDED- I give myself 70:30 odds if Gabriel was flying- I'll let him give my value)