It isn't the automation, it is the Pilot Monitoring, CRM and Culture

[n57flyguy]

US Airways was forced to do exactly this after they experienced a series of hull losses in a short time frame.

I only know this because I benefited from the outcome as we started the E170 program at the last shop.

It really changed, for the better, the foundation of duties, task management in normal and abnormal situations decreasing task saturation and decreasing the resultant errors, either correcting an abnormal or in a task saturated environment.



The first three are answered by what @Seggy suggested in his post. Quite frankly, the details of which are way beyond the scope of what can be covered in an internet thread. However, if you have any friends at airways, grab a copy of their normal and abnormal policies and procedures, operating philosophy, automation philosophy and checklist instructions.

To answer your final question, the pilots have to buy in, just like SOPs. It's not worth the paper it's written on if no one uses it in the real world. See: "How we do it online versus in the SIM"

I can probably get some of that information, thank you for the leads.

I understand, I was just looking for realistic points as I am not in that environment to know of such threats or probable errors. Because of that, is landing at the wrong airport being considered a probable error? I ask for purpose of discovery, it is noble cause to decree how the FAA should look at this, but one can only be so curious as to wonder what else is going on and when/where did SA leave the room.

 

[Roger Roger]

You have human factor experts come in and with management work to develop a training syllabus that would work for that airline.



You talk about all of this in the training.

Seggy, you really do seem to have a lot to say about this topic, and I know that you were involved in a company that went through a huge positive cultural change, but it's really hard for someone on the outside (or probably even some of the 121 guys reading) to gather anything concrete and useful from this because at least to me all I see is safety program buzzwords. And I don't say this trying to mock or slam you, I'm trying to see if we can get communication going on this particular topic that a knuckle dragging 135 pilot like me can understand.

 

[n57flyguy]

I can give examples, but I don't know if you are worth my time to.

Advise? Should I apologize for an interest in human factors psychology from a professional pilots viewpoint who is involved in the environment?

 

[Seggy]

Of course. But the PF is the one taking the plane to where it's going; hence the Pilot Flying. So those questions do make sense in terms of role, not in terms of responsibility necessarily.

Why is the PF committing his error of commission? Why is the PM committing his error of omission?

Doesn't matter who is in charge necessarily, it matters who is driving the train, and who should be backing that person up at the time.

Yes it does matter who is in charge, the company culture, and the CRM/TEM Training.

What happens if the First Officer (who is flying the airplane) as the crew is trying to find an airport at night. The Captain thinks it is 'over there', while the First Officer thinks it is 'no over there'. The Captain then says 'MY AIRPLANE', takes control, and starts heading to his 'over there'. Was the First Officer who was the PF making an error? Or was the Captain who was the PM making an error?

Or better yet. An airplane is landing at LGA. The Captain, who has hardly been to LGA is the PM, doesn't like the way things are looking and says 'my airplane' at 500 feet AGL. What does the First Officer do? What happens if the First Officer is saying 'go around' but the Captain says 'nah, I got this'. Once again, does the role of the PF really matter? Or is it more important to focus on the company culture and proper CRM which teaches the crew to handle these types of situations?

 

[Seggy]

Seggy, you really do seem to have a lot to say about this topic, and I know that you were involved in a company that went through a huge positive cultural change, but it's really hard for someone on the outside (or probably even some of the 121 guys reading) to gather anything concrete and useful from this because at least to me all I see is safety program buzzwords. And I don't say this trying to mock or slam you, I'm trying to see if we can get communication going on this particular topic that a knuckle dragging 135 pilot like me can understand.

As @Polar742 said, it is WELL beyond the scope of an internet thread for every single CRM/TEM situation and possible examples to be discussed. Furthermore, operators have different procedures and policies that may work at one place, but may not work at another. So I am not sure why you are looking for something concrete or even what you are getting at??? In order to get a good CRM/TEM program in place, you need human factors experts to develop that type of program.

I guess the best specific example I can give is that the CRM/TEM programs I have been through, facilitators are used to generate discussions about the mistakes made by crew members in accidents or incidents. Threats that may have lead up to those mistakes are also discussed to allow the flight crews to understand how certain threats can 'back' you in a situation.

 

[Polar742]

Of course. But the PF is the one taking the plane to where it's going; hence the Pilot Flying. So those questions do make sense in terms of role, not in terms of responsibility necessarily.

Why is the PF committing his error of commission? Why is the PM committing his error of omission?

Doesn't matter who is in charge necessarily, it matters who is driving the train, and who should be backing that person up at the time.

You are a with great aviation experience, which is why like, with your kind permission, to use your post as a great discussion point for the change in philosophy @Seggy is referring to.

Yes, the pilot flying is ultimately manipulating the controls to move the machine. However in a poorly run cockpit, either you have a captain as pf who totally negates or ignores information presented by the pm. Also, the other scenario where the FO is the PF, where the CA turns the FO into either a human autopilot or a verbally activated mcp manipulator.

The first, and single most important, aspect of CRM is having competent qualified pilots. Everyone who has trained pilots know on that first flight, or oe leg, you are single pilot as the other pilot is already task saturated manipulating the aircraft and completing baseline duties.

The cockpit concept you present, ultimately, is a single pilot concept with an individual to do administrative work.

In the most tangible sense, a crew that can accomplish all their required duties, either pilot monitoring, pilot flying, captain or first officer, with out direction from the other pilot is probably where we need be. That allows the pilot monitoring really to be pilot managing. The cockpit should be open place with a healty conflict environment ("Hey man, I think we changed that the new revision." "Oh, I'm not sure, let's grab the book") versus what goes on in some places ("I think that might have changed in last revision" "we're doing this way").

The idea is to divide workload and create what I like think of as a mental reserve to deal with any issues that might come , from reanalysis of fuel burn (simple issue) to the obtuse (internal fire over NYC). However if you are working hard just to accomplish normal ops, task saturated almost, any deviation from what was expected can turn right into an overload situation with poor results - landing at the wrong airport.

I'm typing a long post from my phone, so if it reads like it, I apologize...

 

[MikeD]

Yes it does matter who is in charge, the company culture, and the CRM/TEM Training.

You're missing where I'm going young Padawan; I'm breaking down the overarching umbrella into its individual parts, in order to analyze them individually, and in fact, you're doing the very same thing with your examples below.....

What happens if the First Officer (who is flying the airplane) as the crew is trying to find an airport at night. The Captain thinks it is 'over there', while the First Officer thinks it is 'no over there'. The Captain then says 'MY AIRPLANE', takes control, and starts heading to his 'over there'. Was the First Officer who was the PF making an error? Or was the Captain who was the PM making an error?

First off, do we yet know whether this example actually occurred in any of these wrong airport situations? Because this would be a pretty deep breakdown in CRM. While it's indeed an example, is it one that has occurred (ie, has CRM broken down this far?)

In this hypothetical, they are both making errors in their respective roles. If the FO/PF thinks he is right, he needs to say why and/or give the supporting information for it. The Capt/PM needs to listen and make a quick analysis, or otherwise state why he thinks or believes the FO is wrong. If there isn't time to do that, then the crew themselves allowed the situation to progress farther than it should ever have, which now begs the question of why wasn't this caught sooner by either the PF or the PM?

In this hypothetical though, we still don't know who is correct, of the two?

Or better yet. An airplane is landing at LGA. The Captain, who has hardly been to LGA is the PM, doesn't like the way things are looking and says 'my airplane' at 500 feet AGL. What does the First Officer do? What happens if the First Officer is saying 'go aroud' but the Captain says 'nah, I got this'. Once again, does the role of the PF really matter? Or is it more important to focus on the company culture and proper CRM which teaches the crew to handle these types of situations?

They all need to be focused on as concepts, yes. But you have to analyze them individually, then analyze how they are interacting, or not interacting, together.

But you can't omit one and solely focus on the other. You're wanting to focus on the role of the PM, while saying "who cares who's at the helm?" (does the role of the PF really matter?). Yes the role of the PF matters! Because we need to analyze WHY the PF believes what he believed at the time and took the plane, whether he is right or wrong.

Taken back to a single pilot example is one of the ways of analyzing the actions of the PF, and how the PM did or did not contribute. This is part and parcel from the who's in charge, that's a separate analysis. For the PF (assuming a wrong airport landing), would he have made that same mistake if he were alone? If yes, why? If no, what changed his mind? What convinced him otherwise?

 

[MikeD]

Yes, the pilot flying is ultimately manipulating the controls to move the machine. However in a poorly run cockpit, either you have a captain as pf who totally negates or ignores information presented by the pm. Also, the other scenario where the FO is the PF, where the CA turns the FO into either a human autopilot or a verbally activated mcp manipulator....

Like I wrote to Seggy, now you're discussing the same kind of conceptual breakdown I'm attempting to do. Analyzing the individual parts to try and see what, in fact, is truly broken? Or what combination of things are broken? Then seeing how those individual breaks are affecting the system as a whole when reassembled. I agree with you.

 

[Polar742]

Like I wrote to Seggy, now you're discussing the same kind of conceptual breakdown I'm attempting to do. Analyzing the individual parts to try and see what, in fact, is truly broken? Or what combination of things are broken? Then seeing how those individual breaks are affecting the system as a whole when reassembled. I agree with you.

It's all already been done. And as you've astutely observed, it needs, and has had, all three areas addressed.

The biggest problem for companies is that there is no tangible return, nothing to value add in the traditional sense. Until they can conceptualize that not crashing airplanes, not landing at the incorrect airport, not having all of these incidents has value, they'll refuse until regulated.

The funny thing is that it can be traded our incorporated for groundschool they already do. The expensive part, the science, is already done. It becomes a task of translating the basic concepts into practical lessons and applications.

Most companies see any time a pilot in training as an expense, no matter how much they will publicly say otherwise.

 

[MikeD]

It's all already been done. And as you've astutely observed, it needs, and has had, all three areas addressed..

But it hasn't been answered fully, at least not beyond a very general sense. We still don't know what caused the individual parts, of the individual examples, to fail. And what common thread lies therein, or threads. Each situation of how each crew got themselves where they did, and how each crewmember failed in their respective responsibility in terms of what convinced them at the moment of what they fully believed to be right? That can only come from sources within the investigation of each individual case that likely won't be released publically; and that is the crux of the repair in order to help patch the "system" as a whole. I agree there was a problem in these particular cases; do I necessarily agree that there is a breakdown in the system as a whole? Im not thoroughly convinced of that yet.

 

[ppragman]

Personally, in my idealized world of CRM there are three rules.

One - Whoever gets scared first wins. If this means that you go around, you go around. If something isn't right, you go back, reboot, and start again (i.e. get delay vectors, go-missed, or cram and climb as required). If the FO doesn't agree with the captain, you revert to the safest course of action.

Two - Backup everything that you do. OBS the runway/load the approach/pull the sectional out/double check.

Three - Have a plan before you start. This should be a no-brainer, but brief the approach well in advance, have a plan, and stick to it until you can no-longer stick to it.

Granted, in this idealized world, everyone adheres to sterile cockpit rules, nobody is in a hurry, and nobody is willing to cut corners. A cockpit (and an aviation company in general) should be a team. I haven't always been the best example of this in the past, but you live (maybe) and you learn.

 

[ppragman]

Most companies see any time a pilot in training as an expense, no matter how much they will publicly say otherwise.

This is my biggest pet-peeve in the whole world.

 

[Polar742]

But it hasn't been answered fully, at least not beyond a very general sense. We still don't know what caused the individual parts, of the individual examples, to fail. And what common thread lies therein, or threads. Each situation of how each crew got themselves where they did, and how each crewmember failed in their respective responsibility in terms of what convinced them at the moment of what they fully believed to be right? That can only come from sources within the investigation of each individual case that likely won't be released publically; and that is the crux of the repair in order to help patch the "system" as a whole. I agree there was a problem in these particular cases; do I necessarily agree that there is a breakdown in the system as a whole? Im not thoroughly convinced of that yet.

Orange Anchor and I had a long discussion about this before his passing.

All the questions you asked were asked during the "reset" at airways.

The thing is, looking at it from an investigator's standpoint, you'll never be able to hold a smoking gun. Because you'll say either "yes the crew followed procedures" or "no the crew didn't follow procedures". I'll bet in both of the latest high visibility incidents, the crews followed the procedure, making only a minor error.

You'd really have to go into a single company, evaluate the incident reports and look for a common hole. You have to see what isn't in the report. I'm not trying to get all new-age holistic, but that is the only way to solve this issue. Kind of like saying someone who's never been in a wreck is a great driver. Sure they speed, they eat, they text, don't use their indicators, but haven't wrecked. They obviously don't need to change.

When you review incident reports (asap is great for this) from the mundane to the severe, you start to see trends.

At my company, you very rarely, if ever, see a report that mentions a first officer intervening or suggesting or even providing information. If the FO is mentioned, it's either only to mention if they were flying or not, and if so what buttons they manipulated. So that relates back to @Seggy's post.

Another place to see is post-incident changes on the company's part. Do they band-aid what they think is a solid procedure by adding a extraneous step to their perfect procedure (in a previous shop we ended up checking fuel about 5 times prior to takeoff), or do they, as part of the analysis of the incident, try to see if procedures as written will replicate the unwanted results? Do they ask questions of the procedures or give answers because of the procedures?

If you're not using your incidents to test the effectiveness of your procedures (most companies don't), it is cycling back to putting the pilots at fault and not looking for the hole.

It's an interesting thing, this hole, you can't ever point to it. The approach I'd to find the rim - which is the common gap in a company's culture that can only be found by looking for what's not there. So you can't say Southern jets had 15 altitude busts last week. The industry has a CRM problem.

However, what you can see is that Southern jets had the highest rates of altitude busts in their most automated jets, Beachball International landed at the incorrect airport the most times with their most automated jets. The ugly Eskimo had the most gnes with their most automated jets. Why does each airline seem to have basic errors with their fleet that allows the most SA and ability to monitor.

Sorry, that's the best I can attempt to convert what's in my melon. It's very conceptual that you're looking for what's the common omission versus hard evidence.

 

[🦈💜]

Personally, in my idealized world of CRM there are three rules.

One - Whoever gets scared first wins. If this means that you go around, you go around. If something isn't right, you go back, reboot, and start again (i.e. get delay vectors, go-missed, or cram and climb as required). If the FO doesn't agree with the captain, you revert to the safest course of action.

This. Absolutely this. And reinforce it by never second-guessing that decision. This is how I teach my students, and this is how I would imagine an ideal crew environment would work.

Two - Backup everything that you do. OBS the runway/load the approach/pull the sectional out/double check.

Three - Have a plan before you start. This should be a no-brainer, but brief the approach well in advance, have a plan, and stick to it until you can no-longer stick to it.
Granted, in this idealized world, everyone adheres to sterile cockpit rules, nobody is in a hurry, and nobody is willing to cut corners. A cockpit (and an aviation company in general) should be a team. I haven't always been the best example of this in the past, but you live (maybe) and you learn.

I've never flown as part of a crew, and declare my opinion unqualified to carry any weight ... but for what it's worth, I agree completely.

-Fox

 

[nc1982]

Great write up @Seggy. Culture at a particular company starts with the training department.

 

[MikeD]

Thats the thing, while there is no "smoking gun" per se, there are reasons that guys are missing the things they're missing, or otherwise reasons that they are being led down the paths they go, truly believing those paths to be correct at the time. This isn't trying to simply blame pilots, it is trying to run the analysis of why these common errors are getting through the swiss cheese model and not being stopped by the various factors that should be in place. CRM problem? That's well and good, but what and why is what we both agree should be getting looked at. Just as there are any number of ways to teach/practice CRM, its a matter of not only finding out why the checks and balances therein broke down, but what factors initially led crews down this road to where the mistake(s) weren't caught early on and got past those checks/balances eventually. I agree that the what and why is there in the specific analysis of incidents, and hopefully those with the ability to access those within their operation, are making that analysis objectively, and recommending the appropriate actions, and that companies are accepting those recommendations.

 

[Polar742]

Great write up @Seggy. Culture at a particular company starts with the training department.

It starts with management. First empowering policy makers. Policy makers write comprehensive SOP, preferred technique and a cohesive building-block training program. The training department is then empowered to train in a standardized manner. The checking events are really a reflection of the training program. Then OE is to add to the skills learned in the sim - reinforcing basic flows and profiles adding situations and giving the trainee the ability to apply the tools learned. Line checks are as much to affirm the policies are useful and helpful tools for the line pilot. The feedback from linecheck events, incidents and flight crew reports should be analyzed by the policy makers to adjust their policies and tools (SOP, FOM, MEL) and adjust modules in the training program completing the feedback loop.

This works for initial or recurrent modules with equal effectiveness.

However, without foresight from management, everyone is hamstrung.

 

[Polar742]

Thats the thing, while there is no "smoking gun" per se, there are reasons that guys are missing the things they're missing, or otherwise reasons that they are being led down the paths they go, truly believing those paths to be correct at the time. This isn't trying to simply blame pilots, it is trying to run the analysis of why these common errors are getting through the swiss cheese model and not being stopped by the various factors that should be in place. CRM problem? That's well and good, but what and why is what we both agree should be getting looked at. Just as there are any number of ways to teach/practice CRM, its a matter of not only finding out why the checks and balances therein broke down, but what factors initially led crews down this road to where the mistake(s) weren't caught early on and got past those checks/balances eventually. I agree that the what and why is there in the specific analysis of incidents, and hopefully those with the ability to access those within their operation, are making that analysis objectively, and recommending the appropriate actions, and that companies are accepting those recommendations.

Unfortunately the companies that need to make the adjustment either through arrogance or penny punching will not make the changed unless coerced by regulation or via economic penalties.

Most incidents that can paint a comprehensive picture are usually summarily dismissed because they are mundane. Or another issue is dealt with, and time isn't taken to see circumstantial issues.

Oddly, the airlines that do these programs that are truly safety focused, like ASAP and FOQA, got them because the union got leverage due to incidents and/or accidents.

 

[Joshwa]

It's a difficult one. I work for a company that preaches pretty restrictive SOPs, which are partially a control measure due to the diversity of the pilot group.

The fundamental problem is are the two crew on the same wavelength? Does PM know what PF is planning? Briefing properly is a start - you can't monitor the other guy's plans if you don't know what they are. The obvious thing to do is to ask if you aren't sure, but if the culture discourages that, things aren't going to go well.

How big is TEM pushed? For us, it's the training department's religion. We discuss threats as part of the approach brief - be it terrain, ATC, funny local factors and so on. For a visual, things like "correct runway" and the like would also figure highly. I fly into a lot of places where the "taxiway" is/was a relief runway, and with the state of runway markings and low sun angles, it's an easy mistake to make.

 

[dustoff17]

Proper CRM and threat and error training which emphasizes the role of the pilot monitoring and a commitment to a just safety culture. It actually is very easy to do.

There are too many variants to this alone.
It's obvious you're passionate about this.