Originally Posted by DoubleWasp
Originally Posted by Astro14
Most crashes are blamed on pilot error.
So, engineers are trying to save us from pilots.
This is the result.
Nice.
Are there currently any examples on record of these types of systems saving a plane where a pilot would have crashed it? Anything well-documented?
Would it actually be a given that any such incidents would become record, or is there a way for the crew to "keep it to themselves"?
I'm just trying to get an idea if these systems have provided a net gain, or if they are basically hundreds of human lives "in the red" as far as being an asset or liability.
Fascinating question. I don't have the answer.
When you look at some crashes, like Colgan Air 3407 (
https://en.wikipedia.org/wiki/Colgan_Air_Flight_3407 ) the warning systems were ignored by the pilots. Now, this captain had a history of poor performance in the simulator, and responded poorly in real life, too, resulting in the death of everyone on board. So, it's my contention that engineers look at this, and by corporate mandate are told to improve the system.
The result?
Well, if a pilot ignores the stall warning system, as the captain of Colgan Air did, then you have to build a system that will respond to that warning without pilot input.
So, you get MCAS, where a stall warning is acted upon by the airplane, which forces the nose down, in case those pilots don't do the right thing.
And that's what happened here with Lion Air - a newly-designed, more forceful system to protect from pilot error. The system response over-rode the pilots, it responded to a stall warning, and the pilots tried to keep the nose up, and because the system was looking at a false warning, it over-rode the pilots and forced the nose down. The pilots were seeing the real picture. Now, they should shut off the trim, but...
So, you ask an excellent question, and I can think of two ways that a system "save" would get back to agencies (or manufacturers); pilots report it, or a data-collection system captures it.
The first is, frankly, unlikely, for a couple of reasons. The pilots would have to know that the system saved them, and the situational awareness (SA) to know that was the case was probably lacking in the first place, or the system wouldn't have been needed. Further, many pilots operate in an environment where any error can cause a loss of their job and/or a fine. So, there really isn't an incentive to report, and in fact, fear of repercussions really suppresses reporting.
The second is something that we've had at United for a long time - data recording. And that is protected information at United, for a lot of good reasons. But the data recording to which we (ALPA/United) agreed can only be used for analysis, not repercussions. This is not the case around the world. In China, for example, data-logging is used to punish pilots. Exceed 30 degrees angle of bank by one degree? $200 fine comes out of your paycheck. 5 degrees? $500. You could lose your whole paycheck in one flight! No wonder some airline pilots are slaves to the autopilot...
The data from the data-logging is captured by the airline. So, perhaps they share, and perhaps they don't. We share. It's called FOQA, and we've identified threats and trends before they become crashes, and the identification of those areas allows us to focus, in training, and in flying, on managing those common threats. There are huge potential safety benefits to this forward looking approach, which identifies threats and errors before they become crashes, when compared with conventional accident analysis which kicks through the wreckage...
In fact, to incentivize reporting that the data logging would've missed, the FAA and UAL entered into an agreement in which the incident on which pilots file a report cannot be further used by the FAA or the airline for certificate (license) action* so there is a growing volume of pilot reporting under this program (known as FSAP, and adopted by other airlines). But the data is protected. I've filed several FSAP reports. Each identifying a safety concern that can be added to the database.
Want a really fascinating read into the intersection of pilot performance and aircraft systems engineering? Read about the A-320 crash at Toulouse**. An Airbus test pilot, who knew the flight control logic intimately, was conducting a demonstration of the fail-safe nature of the Airbus flight control logic with respect to stall during a sales flight. So complete was his faith in the system that he was puzzled as the airplane crashed into the trees, killing everyone on board.
He had not remembered that the system was disabled below 50 feet, or you wouldn't be able to land the airplane, and so, the demo didn't go as planned. In this case, the system was given full authority, the airplane was stall-proof...and the very experienced, fully aware test pilot managed to stall and crash anyway.
*Certificate action is administrative in nature, like a photo-radar ticket. Unlike photo radar, the pilot is presumed guilty of the infraction, and is typically fined $10,000 and out of work, because their license is suspended until he/she can prove their innocence.
**
https://en.wikipedia.org/wiki/Airbus_Industrie_Flight_129 I would read the original report, if you have time.