The saga of Automation Addiction. Yet another close shave

[Centaurus]

https://www.flightglobal.com/news/articles/azur-757-nearly-stalled-after-mode-switch-during-cli-450601/

Another result of automation addiction to the extent that common sense basic flying skills applicable to a Tiger Moth, a Cessna or a Wide Body are seemingly lost forever. The lesson behind Children of the Magenta Line briefing all those years ago, has long since been disregarded by manufacturers, regulators and airline transport operators. For every reported automation incident, there are probably a thousand that go unannounced.

[Vessbot]

I know we talk about this a lot, but in this case it seems like your post is the shoving of a case that doesn't fit into the rubric of a running complaint. According to the article, exactly what should have happened, happened:

The crew was alerted by visual and aural warnings, as well as the stick-shaker, and the captain immediately acted to counter the approach to stall, disengaging the autopilot and pushing the control column forward, pitching the aircraft 2.2° nose-down.

It entered a descent with a vertical speed of more than 3,800ft/min. The stick-shaker stopped and the aircraft’s thrust was increased to near-maximum in order to recover.

[safetypee]

Careful Centaurus, a bit close to error, blame and train, whereas looking deeper might help understanding.
As I recall the situation was similar to another 757 incident during go around, perhaps dominated by power, pitch, trim, and FD.

In this incident, perhaps the uncalled for mode change masked these effects, the FD dominated the visual attention vs outside world (do we really look out); vision is the dominant sense thus the ‘feel’ of the aircraft may be difficult to integrate with the FD commands.

Re Magenta Line, if there is a lesson to be learnt (I’m not a fan of the quote), then the Captain did ‘click click’, he saw the need for this.
Why then didn’t the handing pilot have the same understanding. Was there strong belief that the last button press was correct, even if mis selected; a very powerful bias in perception. Or low experience of such situations to be recalled from memory for comparison, or a slower understanding (scan pattern), or updating perception (knowing and looking at the important instruments), than that achieved by the Capt.
Does this reflect a low standard of training, or a mismatch between ability and situation; this was a qualified crew.

The industry labels such incidents a Automation Addiction, more often without understanding - labelled, put in the pigeon-hole, human error, more training, forget it.
Who decides that automation was working as envisaged, why imply imperfect human performance with an the negative tag ‘addiction’.
This tagging reflects more on our perception of how work should be done, vs how things are done on the line.

How many occasions of this situation without reporting, what are the underlying statistics, why.
I don’t know for any of the above, but we could entertain alternative views and ‘what if’; what can I learn from this.

[vilas]

But it says there is sufficient evidence to suggest the crew probably failed to monitor and mention the changes of autopilot modes. How can this be called automation addiction. This is lack of sufficient knowledge of automatic system operation and procedural violation of not monitoring the mode changes. Whether on AP or manual the scan to monitor the parameters and mode changes remains same. In manual one makes the changes as well as monitor the parameters with automation you monitor that automation is doing the required changes. Auto thrust is engaged does not relieve the pilot from monitoring the thrust and speed.

[pattern_is_full]

I'm a strong believer in combatting automation dependence. I revisit the "Children of the Magenta" briefing at least twice a year on youtube. I continually recommend it to others.

But this was a totally reasonable point in the flight to be using automation. It was simply a mistake (not even necessarily a lack of understanding) in operating the automation - pushing the wrong button. Something everyone on this forum no doubt does on their phone at least once a month, no matter how well they understand their phone.

Plus some CRM problems - the FO failed to announce the mode change, or the PIC failed to monitor and doublecheck that action and announcement, or other misunderstanding (maybe the FO announced "Speed" instead of "Vertical Speed," or the Captain only heard "...Speed").

Serious mistakes, especially taken together. However, not dependence on, or addiction to, automation.

[compressor stall]

Yes, from the little info we have to go on, I’m not sure I’d be trotting out the magenta line here.

An incorrect mode was selected, and not noticed nor communicated, and an unsafe situation arose. The captain disconnected and recovered.

for me it highlights the importance of the Airbus mantra (I know this was a Boeing) of know your FMA (flight modes) at all times. I’m sure Boeing have similar.

[Judd]

An incorrect mode was selected, and not noticed nor communicated, and an unsafe situation arose.

If the report was accurate, the gyrations that caused the unsafe situation were pretty frightening. The captain eventually got around to disengaging the AP ( a slow reaction some would think) and shoved the nose to beyond two degrees nose down (as well as applying near full power thus guaranteeing the aircraft would descend like a bomb). The report said it entered a descent with a vertical speed of more than 3,800ft/min. Still being manually flown! A scary nose attitude considering the 757 was already at a low altitude after take off.

The FD dominated the visual attention vs outside world (do we really look out); vision is the dominant sense thus the ‘feel’ of the aircraft may be difficult to integrate with the FD commands.
The OP opines this incident is the result of automation dependency; whether blind and inappropriate following of following of a flight director or wrong button selection. Others have different views but it would be unwise to dismiss his opinion as baseless.
As one writer said in another forum: "It seems that pilots today get overloaded when trying to program flight automation in an attempt to make the airplane do what the pilots want it to do." Few would disagree with that.

[hec7or]

Having flown clockwork cockpit BAC 1-11s where all I had to do was fly an attitude and power setting on an aircraft with benign handling characteristics, I found there were far fewer distractions than on the modern flightdeck. No FMA, no FMC, no MCP, no ND to worry about, just an ADI to monitor. You knew the autopilot was a bit ropey and the autothrottle was pretty basic, so flew manually by choice. Actually, it did have an FMA, it was behind the FO's seat! (Fault Mode Annunciator)

I think advanced automation has actually increased the workload.

[safetypee]

It’s easy to overlook assumptions when forming conclusions.

Monitoring mode changes #4, assumes that an erroneous selection is both seen and perceived - understood. After a button press mode change is anticipated; ‘a’ new mode is seen, but not necessarily understood as ‘the’ required mode. Expectation bias, wish-think, workload, unfamiliarity, …

The mantra of the Magenta Line #5, assumes that the pilot understands the need for ‘click click’. Where does this understanding come from, how might it be false, reluctance to disengage a system which you chose to use, or believed to be programmed correctly.

With a range of behaviours including those in CRM (however you wish to define it), monitoring, calls, and intervention, all have dependencies and assumptions.
‘Know you FMA’ is a useful concept #6, but it still depends on how we know, how we learn to know.

Monitoring is assumed to be a continuous process, that the PM does not have any other tasks, or can multi task, no distractions, or perceived higher priority calls on attention. ‘failure to monitor’, no cross check’, ‘poor CRM’, appearing in most incident and accident reports, has the industry asked why, or how do these beliefs actual work on the line.

It seems that pilots today get overloaded #7, yet isn’t ‘seems’ an implicit assumption, that we will know and understand the need for alternative action.

Many assumptions in the safety concepts, CRM, monitoring, and checking, are flawed, they are invalidated as soon as human engages with a situation. Individually we don’t behave as others expect, we are irrational; collectively those who judge continue to hold on to the idea of a rational human, the basis of hindsight. Incidents and accident analysis attempts to fill the gap between these; but is the analysis rational or irrational.

Answers, none; but we might look within the detail of these incidents for possible influencing factors (detail not published), and then consider what might be learnt; what ‘could be’ opposed to seeking ‘what was’, or what we imagined ‘it’ should have been.

[compressor stall]

Judd - we're discussing semantics on incomplete data, but you can also look at the facts presented and your criticism in that he shoved the nose down for a very high ROD. Well, he got out of the impending stall, got the last of his power on (assume that he would have been at MCT already so no danger of pitch up with the thrust coupling) and didn't hit the ground. Better than wallowing in the primary (or secondary) stall into the ground....

[Vessbot]

Better than wallowing in the primary (or secondary) stall into the ground....

...or trying to fix it with a flurry of button pushes and knob twists.

[rogerg]

and the autothrottle was pretty basic
Ours didn't have an auto throttle till airbus put one in to give us cat 3a capability. It only worked on approach tho. BCAL.

[hec7or]

yes, but it was useful in the hold if you were having your breakfast while waiting for an improvement in the RVR!

[flash8]

Having flown clockwork cockpit BAC 1-11s
A bit like those that couldn't transition easily from the 737-200 to the 737CL (with EFIS)... and folk think glass is easier ))

[hec7or]

A bit like those that couldn't transition easily from the 737-200 to the 737CL (with EFIS)... and folk think glass is easier ))

Glass is easier for the writers of SOPs

[RVF750]

Personally I think the same problem is going on in the car world. More automation, more toys and assisting systems and it seems the standard of awareness and driving ability is declining in direct proportion.

Just how many modern cars DON'T have a Bluetooth hands free capability? Yet day in and out folk are seen messaging or distracted by their phones, surely a result of the feeling of safety of all the systems in their cars.

I hand fly through the clean up and like to disengage long before I start configuring on approach when the ability of my colleague allows.............

[misd-agin]

It's terrible pilot skills to any airplane slow from 195 kts to 134 kts before a pilot takes over. What were they doing at 190 kts? 185? 180? 175? etc, etc, etc.