If it turns out to be the case where the MCAS erroneously caused a nose-down control then it needs to be investigated as to why. In the Lion Air crash it was determined that a faulty sensor led to incorrect data being fed to the flight computer. We don’t know yet if the same thing happened in Ethiopia or not. It seems the pilot did not follow proper procedure and “flip the swtich” to turn off MCAS as he should have - possibly due to lack of training on this particular model, but again, this is speculation.
MCAS was designed to automatically intervene in the rare situation where the pilot is incorrectly introducing aerodynamic stall (such as the case with AF447). In some ways it’s similar to the ABS system in your car - the computer is trying to correct a situation caused by the human. The pilot does have ability to turn this off - but apparently they didn’t. Why, we don’t know.
If it turns out the system wasn’t working as advertised, or their training was misleading or technically incomplete, that is not human error. That’s regardless of whether there was a system they could have turned off to compensate for a badly functioning system.
I believe that this is part of the problem. The 737MAX is being marketed to the airlines as just being about the same as the older 737s, not requiring a long pilot re-certification, despite being a radical redesign of the 737. This may give experienced pilots with a lot of hours on the older 737 a false sense of confidence about knowing how to fly the new plane after just having got a cursory training on the new features.
We are not flying fully analog systems anymore, we fly systems with a lot of digital components, and there is a lot of software in them. Engines are now digitally controlled, as well as many other aircraft systems. Software is also being used to reduce fuel consumption and emissions, and this can also affect the controllability and stability of the aircraft.
You take my quote out of context. I’m referring to the fact that all system failures ultimately come down to their interaction with humans in one way or another. Either in the design phase or their implementation and use.
That’s right, we don’t know. It will be several months before the public is made aware of what really happened. All we can do is speculate and assume based on similarities to past incidents and anecdotal evidence. Grounding all the 737MAX airframes is probably the prudent measure to take although I highly doubt there is anything fundamentally flawed with its design
What can be reasonably assumed though is that this crash, like almost every other, was caused by a series of circumstances that came together in an unforeseen way which lead to a catastrophic outcome. It’s very common for a combination of human error, bad judgement and mechanical failure to all factor into a crash.
Sure, any combination of factors, or single factors alone could be at play.
But systems aren’t infallible.
If you design a car that has blindspots, despite your best intentions, it doesn’t help to tell people they should have done more shoulder checks, or that they just should have used the brake in situations where there was no indication they should have.
Sure, humans are involved…? That seems like a pretty general point to make. A badly designed car does “interact” with a human when it causes a fatal crash. Drivers make mistakes too, but a badly designed system can also kill a good driver that was not at fault.
If it turns out (the humans at) Boeing designed something that didn’t work as safely as intended, I hope they make it better, so less people are as likely to die.
Same difference when the extreme is size instead of speed. An Airbus A380 is almost as long and much wider than a soccer pitch/football field. It has the usable floor space of a mansion, and has a 10,000 mile range.
A co-worker of mine, who has industry insider friends, said as much to me about those planes. He also said lack of pilot training is due to the move towards unmanned airliners. Like, why bother if the plane is supposed to fly itself?
This weeks events are clear signs that Boeing or any other airplane manufacturer isn’t even remotely close to safely pulling human pilots from commercial flight.
Also clear from this week’s events is that we need separate the FAA from both serving the airline industry and enforcing flight safety.
eta: This week the U.S. followed instead of led. And that is sad.
It’s easy to feel that automation is somehow worse than human interaction for flying and driving, but the statistics don’t bear it out.
You can see that the general amount of yearly airplane deaths in the 2000s and 2010s is generally less than that in the previous decades when there was less automation, despite there being vastly more flights now.
These horrific tragedies do mask the general overall safety of air travel.
This automation does cover up a worrying concern, something known as deskilling. Because airline pilots have so much automation now flying is now 99.99% drudgery (except for takeoff and landing), they are less equipped to handle (increasingly rare) problems that automation can’t handle. This is despite spending 100s of hours a year in simulations of all of these problems, pilots are simply less skilled than they used to be. Automation more than makes up for that fact, though.
Not the problem. We all know we are talking to systems digitally. Flying by flight control wires or direct hydraulics is no longer an option - we all get that - the argument here is that a human being should be allowed to fly the plane when things are going off the rails.
This is more like a car had been put on the market with unstable steering, so left on its own it will turn one way of the other but not go straight. Then to fix this the maker hacks the power steering to offset mechanical deviations from normal.
Then a failed sensor causes the power steering to suddenly push hard towards the curb.
Obviously the right solution is to take the car off the market and recall every car which has been sold.
Only at high altitudes. If they’re still climbing after take off, or about to descend for landing, both pilots have to be in the cockpit. And they generally both will be anyway, excepting using the bathroom.
Still, the oscillation doesn’t require a bad sensor. It just requires the pilot to be fighting the MSACs without turning the system off. That’s mostly about training rather than re-design. If it’s another bad sensor, then that’s probably a recall.
There is a post on pprune about this. The procedure for stabiliser runaway is to switch off electronic control of the stabilizer then physically hold the control wheel for trim and prevent it from moving further. It does not say to retract the trim. But as airspeed increases, the pitch force from the stabiser will increase and make the aircraft harder to control.
The procedure is wrong because with full down trim on the stab, the aircraft can be impossible to control.
