Hacker News Reader

> - training for the 737-8 and -9 was just short of non existent because in practice, these planes don’t fly differently.

Have you read this complaint from an actual pilot?

https://news.ycombinator.com/item?id=19374386

Quoting the HN post:

"My post flight evaluation is that we lacked the knowledge to operate the aircraft in all weather and aircraft states safely.

The instrumentation is completely different - My scan was degraded, slow and labored having had no experience w/ the new ND (Navigation Display) and ADI (Attitude Director Indicator) presentations/format or functions (manipulation between the screens and systems pages were not provided in training materials. If they were, I had no recollection of that material).

We were unable to navigate to systems pages and lacked the knowledge of what systems information was available to us in the different phases of flight. Our weather radar competency was inadequate to safely navigate significant weather on that dark and stormy night. These are just a few issues that were not addressed in our training."

To me - having had a 25 year career in customer-facing systems engineering and support, this just reeks of dismissing the first-hand account of UX.

If the problem is lack of training, (or even "possibly maybe failing sensors") - and your failure mode includes "maybe killing 150+ paying customers in one shot" ... seems like a no-brainer, to me, to fucking ground the planes, and take immediate corrective action.

And again: just my opinion, it just reeks of someone who spends their days in a "spreadsheet-facing" position, making the decision to not ground these planes, not take their users' (pilots) reports seriously, and continue to risk hundreds of customers' lives (and thousands of their own employees livelihoods) on a daily basis.

> - Lion and Ethiopian Air are near the bottom when it comes to maintenance and follow up with Boeing

Another commenter mentioned that Ethiopian Airline is considered Category 1 by the FAA, and is a member of the Star Alliance. Surely that indicates a decent track record?

I'm not saying it _is_ the plane, but as the EASA Emergency Airworthiness Directive points out, while a connection between the two incidents cannot be established, neither can one be ruled out. And for that reason, they are grounding the planes as a prudent temporary measure till we know more.

Quite. That doesn’t pass the sniff test.

If it turns out this is Boeing’s fault, for the reasons suspected, and the FAA have staked their reputation on it - I fear for the future of airline safety.

It seems to be that the crash situations are way too complex to blame a single party. Since a chain of events (automatic systems, pilot actions) seems the be responsible for the crashes.

I imagine that besides what is being done to make sure this doesn't happen again, a lot of time is being spent on making sure that whatever changes, the crashes cannot be blamed on them.

> It seems to be that the crash situations are way too complex to blame a single party.

It seems to me like Boeing is very much intent on blaming a single party, ("3rd world pilots/airlines").

Of course they are. Any rational entity caught up in a situation like this will try to blame someone else. The cost of being liable for anything in today's society is very high so every party fights tooth and nail to make it 100% not their fault. This is especially amplified in a highly public case like this, the politicians will not rest until some entity has been punished, politicians are under a lot of pressure be seen "doing something". The incentive structure of modern society is simply not conducive to accepting responsibility for mistakes and is even less so when politics get involved.

We already know that crash #1 was caused by poor maintenance and communication (pilots not sufficiently aware of the implications of a questionable AoA sensor) of the plane's condition to the pilots combined with changes to how the plane responds to certain pilot inputs and how MCAS works which resulted in a situation the pilots didn't properly handle (Boeing and the airline's failure to train pilots combined with some amount of human error).

I will be very very surprised if the cause of the second crash is as simple as "yup, Boeing built a POS that fell out of the sky" or "yup, the pilots were idiots who flew into the ground".

Thanks to decades of incremental improvement we don't really have "single source" accidents in commercial aviation anymore.

It's worth noting that this particular plane was new. So 'poor maintenance' is not likely.

I don't think anyone serious literally thinks that 'Boeing built a POS that fell out of the sky', but there are indications that they built a plane that required substantially different handling and tried to abstract that fact via the onboard systems, which makes the pilots unable to handle the actual behavior of the plane when the systems were off for whatever reason, which to me does sound like a pretty big 'oversight' on the part of Boeing.

Now, perhaps the pilots could have handled the situation better, but it does seem like the FAA is trying to protect Boeing, (an American company with enormous lobbying power) and Boeing is sort of positioning this as "3rd world pilots" being at fault, the FAA potentially risking American lives here.

That does not look good to me whatever the cause is, (there's unlikely to be a single one and Boeing is unlikely to be completely clean here).

I would consider pilots not being made sufficiently aware of the implications of a faulty AoA sensor a maintenance failure.

Clearly people have fucked up because planes have crashed. We're basically just arguing over whether the responsibility split goes 33/33/33 or 20/20/60 or something else.

> I would consider pilots not being made sufficiently aware of the implications of a faulty AoA sensor a maintenance failure.

From what others have said, I gather there is a problem of this plane behaving differently than the rest of the family, which is being 'masked' by the on-board software so that it 'feels' similar to the other planes in a regular situation and thus requires minimal additional training. This however goes out of the window when the pilot cannot rely on the software and suddenly has to deal with how the plane 'really' behaves, which is substantially differently.

If that's the case, I'd consider the root of the problem to be significantly on Boeing's side.

>From what others have said, I gather there is a problem of this plane behaving differently than the rest of the family, which is being 'masked' by the on-board software so that it 'feels' similar to the other planes in a regular situation and thus requires minimal additional training

The problem that's (presumably) causing crashes isn't how the plane flies. It's perfectly flyable but has a tendency to pitch up more than other planes of the type. As a response, Boeing added functionality to MCAS to compensate.

The first crash likely wouldn't have happened had the pilots turned all the electronic nannies off and flown raw dog. The just didn't realize that was what they had to do for a combination of reasons.

Not knowing that they were flying with a screwy AoA sensor and what the symptoms of that would look like is a maintenance/communication/human systems failure.

The fact that a faulty AoA sensor can have those symptoms and the fact that the pilots didn't know about it well enough to make the connection themselves and disable the system is a failure on Boeing.

Think of it like having a crossover that wanders more than the comparable sedan at highway speed so the OEM silently turns on lane keeping when you use cruise control to compensate. There's a risk that if you're not paying attention it will find a lane that isn't a lane and drive you into a barrier at 70 and on this particular one the relevant sensor was faulty and more prone to doing that.

I fear for the immediate future of the American aircraft manufacturing industry's global reputation.

This is totally fucked up. For an administration whose core-competency which was supposed to be "protecting American Workers" - they're sure trying like hell to get the airlines of the world to turn to Airbus and other foreign manufacturers for trustworthy aircraft.

I mean - holy shit - every engineering effort fucks up. But to refuse to admit it, and cover it up, that's a violation of trust. Who wants to do business with that shit? You can buy product from someone who occasionally makes mistakes, and when called on it, admits it, and does their damnedest to make it right and show you they're trying to fix it. But when someone blames the victim or tries to deny there's a problem, you take your money elsewhere.

> - not everything is back from the Lion Air incident, but internally there was strong confidence that a lack of pilot familiarity with the adjusted mechanics of MCAS contributed to the disaster

Sure, if there's a probability you were involved in developing a potentially flawed system that killed hundreds, the natural reaction is to not accept this and blame someone or something else. And it's not suprising that "internally there was a strong confidence" – nobody there wants to be the one saying "uh, I think we messed up here guys". Boeing is a total echo chamber at this point.

It's interesting you mention the MCAS here since...

> - training for the 737-8 and -9 was just short of non existent because in practice, these planes don’t fly differently.

If this were true, the adjustment shouldn't even matter, since the sole purpose of it is to make the plane behave exactly the same way as the old 737, so that the pilots don't have to go through lengthy training for the new plane.

You were trying to patch up changes to the hardware by adding software, so the hardware would feel the same. You claim you have achieved exactly that. But at the same time, you blame pilots for not reacting properly to those changes you made to the new plane compared to the old one. What?

>Ethiopian Air are near the bottom

This is ignorance, Ethiopian have been flying for 70 years and have an excellent safety record on par with most western airlines.

Could there be more non fatal incidents that only manufacturers and safety regulators are aware of ?

Aviation doesn't work that way. If a moving plane screws up in any way (too close to another plane, wrong taxiway, botched takeoff/landing, etc) it will be noticed. The US also has the widely used ASRS system for confidentially reporting safety issues/lapses of any kind for everybody's benefit, and most other countries have similar systems now, although I can't vouch for Ethiopia specifically.

https://asrs.arc.nasa.gov/

I'm not sure we understood each other. I understand aviation is heavily monitored by the field, but maybe only a few (when fatal or grave) sample of incidents gets to the public eye.

Unless walkingolof is also in the field and by 'track record' he meant even 'up to the slightest faults'.

All aviation accidents are reportable in most civilised countries, and the reports will be public. For an airliner (which these are) they'll invariably make the news. For little prop planes in GA nobody cares unless someone died.

Take an example from Ireland, an island near me. A crew screwed up and typed the wrong temperature into the computer, which calculated take-off thrust based on higher density air than was reality. As a result they had inadequate acceleration and smashed a runway light with their wheels because they took off far too late. Nobody died, nobody even needed a band aid, but that's a reportable accident and made minor headlines... even though it happened in another country miles away.

As your parent indicated, the US further has ASRS, so that not only accidents, but incidents short of an accident (situations in which something went wrong, but fortunately it did not cause an actual accident), are reported. These are anonymised to protect the reporter but you can see (anonymous) reports by aircraft type in a public system.

Aight, that's what I was curious about. Thanks

>A crew screwed up and typed the wrong temperature into the computer, which calculated take-off thrust based on higher density air than was reality

May I ask, why does the plane not have an external temperature sensor able to automatically input this data?

On the model of plane they were flying that isn't possible. There was a firmware upgrade which compares the input temperature to the temperature detected by a probe on the engine feed and if they're too different rejects the input temperature (which subsequently means the plane acts as though you didn't input a temperature at all, disabling auto-throttle) but their airline had not yet applied this optional firmware upgrade.

The larger answer to your question is Raymond Chen's standard explanation on the The Old New Thing. By default no features are implemented so by default that's why your preferred feature isn't implemented.

The accident investigation was more interested in the other thing, given that for whatever reason the plane's acceleration is inadequate, why don't we realise and abort before V1? Humans turn out to be bad at this, so not much use asking the pilots. Airbus have built planes that measure their own acceleration and warn if it seems too low. So maybe that's the future.

Have a look at http://avherald.com/ to see for yourself.

> - training for the 737-8 and -9 was just short of non existent because in practice, these planes don’t fly differently.

Except for, you know, that computer system that throws people into the ground if a probe malfunction, and that the "just short of non existent" training doesn't seem to cover properly.

I don't know how this part works, when a manufacturer says "this plane doesn't require full training but merely a small top up since it's the same as plane X before", I assume the FAA needs to validate this matches reality ?

Not only validate it matches reality when the plane is introduced, but continuously validating it via crew reports. This feedback needs to be a closed loop.

If a plane surprises an adequately trained crew, then it means someone missed something.

  > - not everything is back from the Lion Air incident,
  >  but internally there was strong confidence that a lack
  >  of pilot familiarity with the adjusted mechanics of
  >  MCAS contributed to the disaster
  >  <…>
  > - training for the 737-8 and -9 was just short of non
  >  existent because in practice, these planes don’t fly
  > differently.

Don't you think these two statements are at odds? If there is no difference in practice then why does unfamiliarity kill?

"In theory there is no difference between theory and practice, but in practice it is different".

Is this Yogi Berra or was he mis-attributed?

Why the fuck does the investigation need to be concluded __BEFORE__ they ground the goddamn planes??? That makes no sense it has horrible incentives.

Because its quite possible that grounding the planes would result in more fatalities. People may drive or get shuffled onto older, less safe aircraft.

With such a small fleet in service, for such a short time, with 2 crashes already, I fail to see what a "less safe aircraft" could be among those currently flown nowadays.

> but internally there was strong confidence that a lack of pilot familiarity with the adjusted mechanics of MCAS contributed to the disaster

> training for the 737-8 and -9 was just short of non existent because in practice, these planes don’t fly differently.

These statements contradict.

What are insider's views on the MAX being the new MD11 cargo, which had a similar software solution for stretching-induced aerodynamic instabilities?

https://www.pprune.org/rumours-news/619272-ethiopian-airline...

I understand the point that the aircraft itself is probably safe (up to the fact that we don't know what caused the latest crash), but there appears to be an inconsistency between these opinions (emphasis added:)

- not everything is back from the Lion Air incident, but internally there was strong confidence that a lack of pilot familiarity with the adjusted mechanics of MCAS contributed to the disaster.

- training for the 737-8 and -9 was just short of non existent because in practice, these planes don’t fly differently.

My best guess is that the latter refers only to the situation where things are working properly, and may also ignore the what-its-like-for-a-pilot issues raised in ASRS report 1555013, as mentioned by hnnmzh in another reply to this post.

Plus, at the time of the Lion Air crash, no-one had said anything to pilots about these 'adjusted mechanics'.

Saying these areas are near the bottom feels disingenuous. Boeing should make minimum safety requirements explicitly clear and not sell planes to areas that don’t comply. Not just ethically, but for business reputation.

"practice, these planes don’t fly differently."

Until they crash and kill hundreds of people.

Spare me. The pilots themselves have stated that they were completely unprepared to fly the plane safely in any condition with 0 training or warning.