Hacker News Reader

Comment by tzs | Hacker News Reader

My understanding is that one of the major selling points of this plane was that pilots of previous 737 models would need very little additional training for the 737 MAX. I think it was something like an hour of reading and they were then considered good to go.

It is also my understanding that IF pilots were fully informed of everything MCAS does and its failure modes, and were trained in how to recognize and handle those failures, and maybe also if the plane has the optional AoA features (display from both sensors, AoA disagree light) which could help recognize when MCAS is using faulty data, then the MAX would be fine.

Question: doesn't there come a point where it is cheaper for Boeing to give up on the very little additional training thing, and go ahead and eat the costs of retraining all the pilots to fully understand MCAS and installing the optional AoA features on all 737 MAXs?

View on HN

All the pilots needed to do was identify the problem as MCAS, use manual electric trim buttons to neutral and then turn electric trim off.

But Boeing never trained pilots how to identify the problem, and told them the old procedure was the correct way of dealing with it. The old procedure told pilots to turn electric trim off, never turn it back on and return stabilizes to neutral with manual control, which turns out to be impossible.

That's not one, but two pilot training issues with Boeing's official training materials.

A related problem is that because of the no extra training requirements, Boeing explicitly designed MCAS to be near impossible to disable, because if it was disabled then the plane now had different flight characteristics, which the pilots weren't trained to handle.

And this is also the reason why MCAS only used data from one sensor. If the sensors disagreed and MCAS disabled itself, then that would require pilot training.

The whole system would have been designed differently, and more safely if it wasn't for the "no extra training" requirement.

It was much worse than a lack of training.

Prior to the first crash in Indonesia, MCAS was not mentioned in the formation to the pilots. Boeing has asked the FAA to remove it from the documentation. The FAA did not know that the MCAS was totally different from what Boeing had initially planed, so they agreed.

After the crash, Boeing published new instructions that mentioned MCAS. But they were of no use for the Ethiopian crash. As you wrote, turning off MCAS also turned off stabilisation, and manual control was not possible because of the excessive speed. The pilots did understand that, so they enabled the electric trim again. Unfortunately, this activated the MCAS which send the plane downside 10 seconds later. They had no way out.

I have a personal theory (based on the graphs from the reports from both crashes) that once the stabilisers go far enough out of trim at high enough speeds, that the even electric trim motor isn't powerful enough to move the stabiliser.

In both crashes, after repeated activations of MCAS has pushed the stabiliser almost fully down, we see the pilots do two quick presses of the trim up button. The graph shows the stabiliser barely moving (much less than a press of the same length earlier in the flight).

I suspect the pilots pressed the trim up button, heard a loud noise from the electric motor straining, or a fuse popping, or something? So they stopped pressing trim up.

Even worse, only trim up was failing, MCAS was still able to command more trim down.

Best summary I've read so far, of the whole fiasco.

As the plane is headed down, it increases its speed. At a certain point, due to aerodynamic forces, pilots cannot stabilize the plane with the stabilizer trim wheel in the cockpit. The situation is unrecoverable.

This is explained here: https://www.seattletimes.com/business/boeing-aerospace/boein...

They also need to either add a manual MCAS cutout (separate from the power-trim cutout) and probably add a third AoA sensor too.

The airplane will not be safe as long as the only way to exit a nose-high MCAS runaway situation is to cut out the power trim and move a wheel that physically cannot be moved while the aircraft remains in the nose-high situation. And without a third AoA sensor the chances of a failure remain unacceptably high.

> Question: doesn't there come a point where it is cheaper for Boeing to give up ... and eat the costs of retraining all the pilots

I thought that perhaps it was worse than this. If much training is required, then you are admitting that it's not the same plane anymore. And if it's not the same plane, then why does it get to grandfather in all the unchanged parts from the 1960s? Meaning... perhaps things like doors? I could be wrong about this, but I thought there were a bunch of mechanical parts which you would have to design differently on a new aircraft, to meet current rules.

So, your first two points miss something important.

The two sensor setup would basically mandate Level D simulator training. That was a non-starter to maintain the 737 MAX sales pitch, and would also have likely led to inconvenient questions that would have kept sales depressed because of too much attention being called to the actual implementation details of the plane.

That came out in the original 60 Minutes Expose on the 737 MAX iirc.

Boeing actually benefits from keeping training low, it's actually a barrier to entry for competing planes. Boeing is appealing to airlines that have a lot of Boeing pilots because of the low retraining when upgrading models, this allows Boeing to overcharge those airlines since they can make a margin on the retraining costs the company would otherwise need to put out.

From what I can tell (IANAP) the MAX was specifically designed to maximize this margin by the promise of that hour, possibly this grounding wouldn't have happened if Boeing had stated a slightly higher retraining requirement, and possibly a PHB overrode a suggestion from an engineer for some extra training time to maximize that window. That's all speculation, but it'd be extremely unsurprising.

Reply on news.ycombinator.com