Latest research on Automation Dependency - Regulator please note..

[PJ2]

Maurice Chavez;

As for the wonderful accident rate compared over the years, compare traffic from 1950 and actual now....

QED.

From the 2009 Boeing Summary:

"Accident Rates:In general, this expression is a measure of accidents per million departures. Departures (or flight cycles) are used as the basis for calculating rates, since there is a stronger statistical correlation between accidents and departures than there is between accidents and flight hours, or between accidents and the number of airplanes in service, or between accidents and passenger miles or freight miles. Airplane departures data are continually updated and revised as new information and estimating processes become available. These form the baseline for the measure of accident rates and, as a consequence, rates may vary between editions of this publication."

alf5071h, thanks for the link to the "paradoxes" paper - well worth reading.

PJ2

[p51guy]

A few posts back a poster used the term mode confusion. Is this an acceptable term for not being able to control your aircraft as you desire? Why not get unconfused and disconnect the autopilot until you can get it going where you want again. Is it acceptable now days that confused means you have to press buttons to get unconfused? I hope not. I know I am of the old school but what happens if the crosswind exceeds the autoland limits on an unprotected ILS approach and the computer pilot has to figure out how to land it manually if he hasn't in months? I've had localizers deviate full scale on an unprotected approach into MIA at 100 ft and if it had been coupled the 250 hr guy with the push button talent would have been in serious trouble. Stick and rudder is laughed at a lot on this thread with automation taking over but it will bail you out of a lot of situations in real life situations.

[A37575]

Stick and rudder is laughed at a lot on this thread with automation taking over but it will bail you out of a lot of situations in real life situations.

Except this captain couldn't fly stick and rudder to save his life - literally:

A Kenya Airways Boeing 737-800 took off on a dark night and entered a slow right roll that continued for nearly a minute without the flight crew or the autopilot engaged. The captain - the PF -was preoccupied with the weather and had lost situational awareness.

The F/O who was left out of the loop of the captain's "planning" was not effectively monitoring what was going on and he did not notice the autopilot had not been engaged as intended. Confusion and spatial disorientation prevailed when a bank angle warning sounded. The captain responded with erratic flight control inputs that aggravated the situation and precipitated a spiral dive. The pilots were wrestling with the controls when the 737 disntergrated in a mangrove swamp killing all 114 people aboard.

[Tmbstory]

The loss of Manual Flying Skills

The continuous loss of these flying skills in the present Aicrew of the Aviation Industry will cause the accident rate to increase with unfortunately, more fatalities.

There has to be a way that Automation and Manual Flying Skills can be kept at an acceptable level.

Tmb

[Phantom Driver]

The continuous loss of these flying skills in the present Aicrew of the Aviation Industry will cause the accident rate to increase with unfortunately, more fatalities

.

Really? Better take a look at the accident rate before automation was introduced. Not pretty. One guy on this forum wants to disconnect and fly manually after a long haul flight because he "wants to do something". Another contributor can't understand how the PM can get overloaded in his monitoring duties. Excuse me?!

The Atlantic Barons in their Stratocruisers, Constellations, 707's,DC8's were indeed a breed apart as far as pure flying skills were concerned, but you have to admit-they lost quite a few.

The accident rate in those days would be totally unacceptable now. Todays mass market aviation environment caters for the lowest common denominator in flying skills,(of which there is now plenty, due to massive growth in the industry), and so Engineers, not Pilots, rule. Thats the way of the future, and so be it.

Do you really think you are "flying" these (new) generation aircraft? 777 is FBW, but it's been around a long time now, and yet human factor problems still crop up (Air France/ Air India A/T issues in Lagos/Delhi respectively,). As aviators, our energy would be better spent campaigning for more user friendly systems that are easier to understand. Jeez, how many ammendments to FCOM have you seen come out from both Boeing and Airbus with regard to whats going on with the various modes after T/O or G/A?! This after years in service.

Sure you're going to have to disconnect and fly manually while both of you try to figure out what is going wrong, but in todays crowded RVSM /RNPairspace environment, you are going to have other distractions and conflicts vying for your attention. You didn't have this additional "pucker factor" in the old days.

(p.s. How many more times are we going to have to go over all this manual v automation stuff again? I've heard of history repeating itself, but this is getting a bit out of hand..

[stepwilk]

What the FAA researcher whose report occasioned this thread is saying is not that automation is bad but that current training is not preparing pilots to operate and interact with it properly. It's not a call for a return to stick-and-rudder skills (not that I'm saying there's anything wrong with that) but for better training to prepare pilots for new-generation automation.

[BOAC]

Step - indeed one of my points on the other thread on 'Safety, CRM etc' - we have been let down over the years by poor training, particularly with the AB, and suffered there with over-hyped sales talk. I think the training system here is close to catching up, but there still needs to be a drastic shift in attitude - modern a/c offer such tremendous benefits in automation and efficiency as long as they are both understood and used correctly. Over-emphasis on 'automatics' OR 'stick and rudder' are both fallacious - we need to instil in this up-and-coming generation the ability to know when and when not - but of course to retain the basic skills more than once every six months..

[Maurice Chavez]

Originally Posted by Phantom Driver

One guy on this forum wants to disconnect and fly manually after a long haul flight because he "wants to do something". Another contributor can't understand how the PM can get overloaded in his monitoring duties. Excuse me?!

The Atlantic Barons in their Stratocruisers, Constellations, 707's,DC8's were indeed a breed apart as far as pure flying skills were concerned, but you have to admit-they lost quite a few.

Dear Sir,
First of all you seem to know better then the rest of the industry, do yourself a favor read this flight safety foundation article. Seconds, the guy that questioned the the workload of the PM, whilst hand flying was me. Again I'm still waiting for somebody to point out this overload...Please explain that whole philosophy......
Third, I'm an ex Diezel 8 driver, and yes we lost quite a few, but not as much as the "new generation" aircraft, or maybe the "new generation pilot" should I say..We did use automation as much as possible, but when "the thing" wasn't doing what it was supposed to do, we took over.....

Read the article posted above and try to be honest with yourself......Hope you learn something.

[Phantom Driver]

Ozy-

Good for him. He wants to preserve essential flying skills.

Agreed, but at the end of a 14 hour slog, even if half was spent kipping in the bunk? There is a time and place for everything.And lets be clear here, we are talking about extensive manual flying on departure and on the approach. I would hope we are all flying manually on finals when visibility permits, especially in adverse x-wind conditions. No excuse for not doing so, and hopefully no company advocates disconnecting the autopilot on short finals, or worse still mandatory autolands (unfounded rumour that this was the case initially with the A380)

I suspect this particular contributor believed that if a pilot can't cope with PNF duties he isn't likely to cope with PF.

Maybe, but I would suggest that even the best of us, when PM, will find things getting a little hectic on a night departure in a busy ATC environment with an SID requiring extensive overwater manoeuvring coupled with low level initial altitudes, followed by non standard vectors/level/speed changes.

In the old days, the PF flew instruments, and the PM (PNF!) watched. No longer so in todays machines.

PM is often working like the proverbial "one armed paper hangar". The main point is -who is monitoring what? PF is busy flying instruments, PM is busy with (maybe) ATC/MCDU/FMC/checklist. OK on nice sunny days, but night IMC? Inevitable breakdown in X checking, as shown in many accidents.

Once again-"A time and place for everything".

Stepwilk-

What the FAA researcher whose report occasioned this thread is saying is not that automation is bad but that current training is not preparing pilots to operate and interact with it properly. It's not a call for a return to stick-and-rudder skills (not that I'm saying there's anything wrong with that) but for better training to prepare pilots for new-generation automation.

Amen. The bottom line? Basic requirement, (when things go pear shaped), is the old "Level the wings and find the horizon", the foundation on which our instrument flying is based. If we can't do that, then there is something definitely wrong. I don't think that is the case; the Aviation system is basically sound, no matter what some may think. The average guy/gal in the cockpit is competent. International standards demand it. Accidents will happen, no matter what; thank God they are at historical lows, despite increased volumes.

But forget about manually flown raw data approaches to Cat 2 minimums. Maybe the Ryanair/EasyJet folks doing 2/3 sectors a day can hack it, but I speak from the perspective of a long haul heavy jet operator who maybe gets one landing a week (if lucky).


Maurice C- Thanks for the link. I will try and read it and hopefully learn something.As the saying goes, We only stop learning when we hang up the headset for the last time!

[ChristiaanJ]

Originally Posted by Phantom Driver

As aviators, our energy would be better spent campaigning for more user friendly systems that are easier to understand.

I thought this hit a BIG nail on a BIG head.
I'm astonished it didn't solicit any comments.

CJ

[p51guy]

Our airline couldn't fly cat 2 approaches hand flown. We had to on check rides but not in the airplane. I hope we never get to the point where cat 2 approaches can't be done by pilots in the sim. We have no pilots under 40 years old so don't have to worry about that unless we start hiring again. Underqualified pilots need to do what the rest of us did and pay their dues and get qualified. Why can't they do what we did?

[bearfoil]

PD, ChristiaanJ

"Campaigning for the logical solution" used to be not ironic. It is of course the least used approach, as it generally results in marginalization, or blank and dim looks. What a state we are in. There are simply too many people noticing the problem. Training? Software? Airframe? Command? In today's environment it takes newsworthy events to frame the dysfunction, and that is insufficient for co-operative address.

"Compromise" in most pursuits is all too welcome. Oakape 08:51 has the problem well-defined. Given Phantom's recent post, it is clear the problem is defined, and at least a preliminary approach, if not attitude, is available.

Safety vs. Money. How did we get in this mess? More to the point, why are the wrong people making the call?

bear

[md80fanatic]

If I understand correctly, when the a/p disconnects it is usually due to flight conditions it cannot manage, for whatever reason. That being said, I was wondering if there could be a worse time to pass control to a pilot not exactly proficient with stick and rudder skills?

[p51guy]

It was unexpected to me on that approach to ATL at 300 ft when the MD80 autopilot said you got it. We did fine but it wasn't what we expected. It was turblulent and unexpected. Don't trust automation. It helps keep the workload down but it can not be trusted to get you on the ground.

[alf5071h]

Whilst I don’t believe that the problem is overuse of automation, nor the lack of stick and rudder skills, the apparent lack of ‘appropriate’ manual flight practice may indicate which mental skills have degraded.
Thus, what are these skills – airmanship, and why aren’t they being practiced during automatic flight. Again, this points towards training, but also the equipment.
Does modern technology require airmanship skills to operate it, or alternatively enable application and practice of these skills – are the new systems ‘easy’ to operate mentally.

Re “… more user friendly systems that are easier to understand”
This is not a single issue (a big nail); there is whole matrix of options here. We can have complex systems, but easy to understand and use, and simple systems but difficult to comprehend and use, etc, etc. The problem revolves around what is nice, what do we need (essential), what is tolerable, and for all of these, in what context both situation and use.
Then of course you might have the same range of options for people - the complex, yet understandable, and the nice but not so good pilots, etc.
Most of the above can be modelled in the SHELL diagram, but the problem is not entirely in the boxes it is in the jagged edges of the interfaces.

As a thought, what if the perceived automation/technology dependency is actual an aversion to manual flight? What would have caused that?

[protectthehornet]

aversion to manual flight

alf made a good point here. I looked back on my humble career and rememberd the first time I flew a sabreliner at FL450...and the autopilot was broken (out of service). No passengers...ferry flight.

It was a handful. I was very new in the plane...my first jet. I was copilot. It took all I had to keep her in the ball park on altitude and heading. but I got the hang of it...might have taken 20 or 30 minutes to really feel things out.

Now, flying my first year at the big airline. Autopilot out at night on a DC9. Lower altitude of course. Copilot there too. I asked the captain for FL260 as a final...relatively short flight. I knew by this point the stability would be easier at the lower altitudes. I could feel the flight attendants move around the cabin. things worked out just fine.

the next day, we had the same plane and it was the captain's turn. he went up to the planned altitude and he had a handful of plane...3 hours, no autopilot. I offered the suggestion to fly lower...he told me I was a rookie. ATC told him he was off his altitude.

He had never flown the plane at max altitude by hand. And these were the old days.

Simulators can't or won't produce the ''feel'' of MANUAL FLYING...and airlines are too cheap to let you go out and wring the plane and yourself out in real life.

If you can't fly ''manual'', then you shouldn't fly automation in that regime...because if you lost the automation, you couldn't handle it.

flight training...that's what needs to be changed.

[DA50driver]

Training

Training needs to stretch your comfort levels. When I had students that thought 45 degree banks were difficult to do while maintaining altitude I made them do 60 degrees and all of a sudden 45 was a piece of cake.

A lot of people who have entered the business have no training outside pass the check ride and oral. I have had "fully qualified" line SIC's that don't want to do a visual, because they have never done one and are uncomfortable with it.

I have to admit that the only days I look forward to flying now is when the weather is crappy with low vis and blowing snow on a slippery runway. ILS on auto pilot all the way down is so boring I would rather stick needles in my eyeballs.

I am a pilot, not an accountant's button pushing little be-atch. (I fly an airplane that uses a mouse instead of button pushing by the way, so as we progress we can't call them button pushers anymore).

My mother thinks I am a piano player in a whorehouse, I don't want her to be too ashamed of me.

[Lonewolf_50]

Really? Better take a look at the accident rate before automation was introduced. Not pretty.

Respectfully, in those days before early auto pilots(which preceded automation), and before the current generation of automation, the ability to forecast weather was comparably nil to now.

Code:

Aside: is a trim wheel "autopilot?"  I don't think so, I think it is a secondary flight control.  It seems that autopilot requires us to have on hand something electronic, using logic and comparison statements in electric form to make a control input to do X part of a pilot's task.

The art and science of human factors and crew coordination was primative. The materials and design art was hardly advanced.
The quality of navaids was lesser.
To attribute to automation the improvement in safety record is, to my view, both reductionist and inaccurate.

Another contributor can't understand how the PM can get overloaded in his monitoring duties. Excuse me?!

Are you referring to the guy monitoring the automation, or the PNF?

As I see it, autopilot functions are intended as an aid to the pilot, a workload reduction, not a replacement for his function. Automation I suggest is something different from autopilot, since it has grown in so many ways. As I see it, it has been sold to both the military and to aviation companies and operators as, in part, a way to replace the function of the pilot.
If I am being inexact or fuzzy in this thinking, I sincerely believe that this has been the outcome, desired from the outset or not.

The Atlantic Barons in their Stratocruisers, Constellations, 707's,DC8's were indeed a breed apart as far as pure flying skills were concerned, but you have to admit-they lost quite a few.

See above concerns in re state of the art across the whole gamut of issues involving air transport.

As aviators, our energy would be better spent campaigning for more user friendly systems that are easier to understand.

100% agreed. Design ought not to add workload with systems intended to reduce workload. Seems contradictory to me.

I mention this anecdote again, as I did some months ago. A colleague of mine was an instructor at the Navy Test Pilot School (Rotary Wing), who had under instruction a couple of Seahawk pilots who were lacking in stick and rudder skills despite having each over 1000 hours in the Seahawk. The competition/screening to get into TPS is fairly stout.
He had to let them go. Helicopters are seriously stick and rudder creatures ... or are they?

With AFCS and automation, and missions that rely on them, the opportunity to exercise and grow stick & rudder skills in these basically good pilots had atrophied. (Add some inane SOPs that precluded certain training maneuvers, and you have a systemic refusal to keep S & R alive across a range of skills).

That many veterans in the air transport industry see an almost identical problem with the erosion of flying skills now (just under 20 years after my colleague's experience) strikes me as a sign that over automation is endemic in all facets of the aviation industry, from the design and requirements end.

[Phantom Driver]

No, it isn't inevitable. A competent crew can cope - if they can't, then what are they doing on the flight deck of an aircraft

Of course they cope. Every day. please note the distinction; "Night/IMC". I repeat-there is a time and place for everything. In the sim, I have to demonstrate my ability to fly manual raw data approaches to minimums. Can't hack it? "You failed, Captain!"

In any major airline today, if the examiner is willing to sign that licence (and thereby putting his neck on the line as well), then you at least meet the minimum standard.

But to subject 400 customers down the back to a hand flown approach to minimums in IMC just to prove that you still have the Right Stuff doesn't make sense to me. But then, how often is this required? Very rarely, I would suggest. An autoland is usually mandated anyway by most companys in such conditions.

And I beg to differ from P51 Guy:

Don't trust automation. It helps keep the workload down but it can not be trusted to get you on the ground.

Todays Cat 3B systems, with the built in redundancies & safety back ups, do an excellent job of getting the aircraft on the ground. Again, we are required to demonstrate the ability to recover from a multitude of bad situations on the approach.

However, 99% of the time, I would hope that every one of us is flying the whole final approach manually when conditions permit. (I'll save the raw data stuff for the sim and use the flight director on the line, simply because these days the FDAP "Big Brother" is watching every move. One small step outside very narrow parameters & it's a call to visit the Chief Pilot's office for tea (without biccys). At least that's the way it is on most major operations.

Quite rightly so--it's the least our customers deserve; a smooth, troublefree flight with the minimum of drama or excitement.

Hornet-

Simulators can't or won't produce the ''feel'' of MANUAL FLYING...and airlines are too cheap to let you go out and wring the plane and yourself out in real life.

Well, if the Ace of the Base is up front and having a good day, then he might do as good a job as the autopilot where nailing the needles is concerned. But with the average Joe attempting same after a long haul, i.e hand flying down to minimums (night/IMC)--well, I suggest there would be a few white knuckles down the back. Does anyone actually try to do this on a regular basis? I would hope not.

Das;

I have to admit that the only days I look forward to flying now is when the weather is crappy with low vis and blowing snow on a slippery runway. ILS on auto pilot all the way down is so boring I would rather stick needles in my eyeballs

.

Can you guarantee to do it right EVERY time? One or two go arounds and the Chief Pilot will be giving you a call. As professonal pilots, our job is to deliver a safe comfortable product. If I want to have fun I'll go fly a Pitts Special. Solo.

Alf.

As a thought, what if the perceived automation/technology dependency is actual an aversion to manual flight? What would have caused that?

Now you have hit another big nail on the head. Ozy makes the observation;

I've seen a few pilots who are too dependent on the automatics, and that is very worrying.

. Very true. I recall an accident report some years back; DC10 overrun landing at (Boston?). The A/T had a known fault of holding excess speed (Vref +20) all the way down to touchdown on a short, slick runway.

The report stated that the Captain had been so used to flying with A/T engaged that he was reluctant to disconnect, even when he saw it not performing to standard. So the problem with automation has been around a long time.

Humans will err. We will continue to reinvent the wheel. The question is-Why? I leave that one to the psychologists and CRM experts.

Re the ongoing debate-My personal opinion?--On sim recurrency training (not check day), lets spend more time "Flying"; more manual, more raw data etc, rather than the beloved "Loft exercises". We try to cover so much in the sessions that half the time you come out at the end wondering "what exactly did I learn there?".

Once well understood (and thats where my plea for simplicity comes in),save the automatics work for the line, And there, the bottom line, (as Mr Airbus and Boeing have always told us) is " Never forget there is something called a disconnect button".No matter how rusty we are, we should still be able to revert to basics, and I'm pretty sure that is the case.

[DozyWannabe]

Originally Posted by Lonewolf_50

As I see it, it has been sold to both the military and to aviation companies and operators as, in part, a way to replace the function of the pilot.

Really? The current standard in commercial aviation is 2 flight deck crew. That number was introduced with the DC-9 and the Jurassic 737 - neither of which could be described as an automation-rich product.

I've said this before and I'll say it again - the "automation to replace pilots" meme has only ever come from the press, and given the disdain shown by a majority of people on here for the press's opinion on other matters, why do they take this one so seriously?