Automation affects pilot performance

MD11Pilot

MD11Pilot

Line Up and Wait
Joined
Oct 6, 2014
Messages
567
Location
Broken Arrow OK
Display Name
Display name:
MD11Pilot
Interesting article. My longest flight on a regular basis is 10.5 hours and other than the first and last hour, we talk to no one. All done by Controller Pilot Data Link sending automatic position reports and other than the usual thunderstorm deviations over the equator we just sit, fully alert and prepared for any contingency :yes:
The Feds have reversed their opinions on automation and require us to demonstrate more hand flying in the sim. In reality for me, I tend to hand fly up to altitude and from top of descent down but sometimes am so tired that I just click the A/P on short final. Just a little something for thinking about.

http://www.pddnet.com/news/2014/12/cockpit-automation-can-cause-pilots-lose-critical-thinking-skills
 
Fatigue and monotony also affect pilot performance.

I see nothing about hand-flying for hours on end with map and pencil and E6B and calling FSS every so often that is better than the way I fly today - and all that came from automation. I'll take autopilot and ipad any day.


I get to sit back and think about "what ifs" and "what would I dos"


.
 
Last edited:
We are allowed to have the autopilot on as much or as little as we want. Usually descending below 10,000 I will click off the autopilot and autothrottles and hand fly it to the ground. Or hold off on putting it on until I reach 10,000 ft. in the climb.
 
Interesting. So I have a qeustion for you airline guys,
are you required to switch on/off the auto pilot at a cerrtain point or time?
 
I would tend to agree with the article, or at least the summary presented,based on what I've seen in the simulator...

Just about every flight instructor can spout the Law of Recency and the Law of Exercise, but very few pilots seem to comprehend the implications of LACK of Recency or Exercise...that which is not practiced is lost.

Most pilots have no difficulty hand flying the two approaches required for ATP/Type Rating, but have them fly a full VOR approach from a procedure turn (with or without the autopilot) without a magenta line on the MFD, and the majority seem to find themselves 12-15 miles from the fix during the procedure turn when <10 miles is specified. Not due to lack of ability to fly the airplane, but lack of ability to remember that the 10-mile criteria exists until it's too late due to lack of Recency.
 
Last edited:
farmerbrake said:
are you required to switch on/off the auto pilot at a certain point or time?
Click to expand...

The autopilot is required to come off at 200 feet per the airplane's flight manual. The auto-throttles can stay engaged from the time I push the power up on take-off until I engage the thrust reversers(where they automatically disengage)
 
I didn't say I would give it up. I am just starting to fly small airplanes again and use my IPad and Foreflight. I sit there and go "How in the world can you get lost now?" I just try to stay more aware of complacency. Non Precision approaches with the flight management computers etc. with options of VNAV, Prof (a Douglas thing), autothrottles etc. seems like it should be easy but then throw in autothrottle inop. Behind on systems knowledge, fatigue and you can have some very dicey situations. It is a matter of only doing them once a year in the sim and then when you are faced with them at the last minute...

A friend and I were riding our motorcycles, his a BMW 1200 GS and Mine a Honda Goldwing. He said "I don't need reverse like your big machine" and I said "I don't either BUT I don't need air conditioning in my house or car either...but they sure are nice!

petrolero said:
Fatigue and monotony also affect pilot performance.

I see nothing about hand-flying for hours on end with map and pencil and E6B and calling FSS every so often that is better than the way I fly today - and all that came from automation. I'll take autopilot and ipad any day.


I get to sit back and think about "what ifs" and "what would I dos"


.
Click to expand...
 
My experience giving instrument refresher training in light planes is that pilots flying systems like the G1000 with the digital autopilot or the equivalent Avidyne system often tend to be weak on basic aircraft handling skills. We d up spending more time on autopilot-inop than we do on anything else, even to the point of taking off the hood and going back to basic Private Pilot aircraft handling tasks.

One solution is to do what MD11 said and hand-fly it to altitude (or at least maybe 3000 AGL) and then take over by hand again on the descent. However, it's also important to maintain proficiency on "full system" operation, which on systems like the Cirrus Perspective can be more difficult than it seems (VPTH and all that, not just the basic functions) because when you get in complex high-density airspace and bad weather, you really do want to use all that functionality to take the physical flying load off yourself and free your brain for more executive functions if at all possible.

Someone flying professionally or otherwise getting 300-800 hours a year in all sorts of conditions gets enough opportunity to exercise all those capabilities. However, the average owner-pilot flying 150 hours a year (or less) may not, and that's where regular refresher training with a lot more than just three approaches and an IPC endorsement comes in. The pros get that mandates by 121/125/135 or corporate insurance rules, but the people who need it most often don't, and that's a problem.
 
Automation can degrade stick and rudder skills. But that's not enough to indict automation as some people do (not the OP, not even necessarily the FAA). In other words... so what?

In some phases of flight, stick and rudder skill is critical and can't be replaced or even much enhanced by automation. That's why we all hand fly takeoffs and landings and patterns. There is too much minute adjustment by feel and split-second human decision-making that is needed in those phases - especially if the engine quits or winds are squirrelly.

But on climbout, in cruise and in high descent/perhaps even approach my concern is not with hand-flying the airplane precisely on the magenta line or watching the ASI/VSI/HSI. I need to think about threats (engine-out what-ifs, look for collision hazards).

I don't need to be able to execute a flawless chandelle or steep turn in order to avoid hitting another airplane or a bird but I do need to spot the plane or bird.

I don't need to be able to fly a perfect steep turn to fly an acceptable holding pattern. I do need to have the approach briefed and know what I'd do if X, Y or Z instruments failed at that moment, run my checklists, etc. None of this requires me to be the most awesomest hand-flying pilot but (and I'll bold this and still people will probably get me wrong) I do need to be able to achieve acceptable results by hand flying if that becomes necessary in any phase of flight, any maneuver.

So I guess it comes down to how much performance degradation matters in a given phase of flight and what the pilot is doing with time he's awarded by using automation.
 
petrolero said:
Automation can degrade stick and rudder skills. But that's not enough to indict automation as some people do (not the OP, not even necessarily the FAA). In other words... so what?

In some phases of flight, stick and rudder skill is critical and can't be replaced or even much enhanced by automation. That's why we all hand fly takeoffs and landings and patterns. There is too much minute adjustment by feel and split-second human decision-making that is needed in those phases - especially if the engine quits or winds are squirrelly.

But on climbout, in cruise and in high descent/perhaps even approach my concern is not with hand-flying the airplane precisely on the magenta line or watching the ASI/VSI/HSI. I need to think about threats (engine-out what-ifs, look for collision hazards).

I don't need to be able to execute a flawless chandelle or steep turn in order to avoid hitting another airplane or a bird but I do need to spot the plane or bird.

I don't need to be able to fly a perfect steep turn to fly an acceptable holding pattern. I do need to have the approach briefed and know what I'd do if X, Y or Z instruments failed at that moment, run my checklists, etc. None of this requires me to be the most awesomest hand-flying pilot but (and I'll bold this and still people will probably get me wrong) I do need to be able to achieve acceptable results by hand flying if that becomes necessary in any phase of flight, any maneuver.

So I guess it comes down to how much performance degradation matters in a given phase of flight and what the pilot is doing with time he's awarded by using automation.
Click to expand...

You didn't read the article, did you?

It's about what happens when the automation fails, particularly related to situational awareness. If you watch a magenta line all the time, can you identify the aircraft's position with respect to navaids? That's a critical question for remaining in protected space under IFR.

It's also a NASA study, not the FAA (and yes, I do know one of the coauthors). I remember the call for subjects, but I didn't participate because I'm not an instrument rated commercial pilot like they wanted.
 
MAKG1 said:
You didn't read the article, did you?

It's about what happens when the automation fails, particularly related to situational awareness. If you watch a magenta line all the time, can you identify the aircraft's position with respect to navaids? That's a critical question for remaining in protected space under IFR.
Click to expand...
Another area I hit on both initial and refresher IR training is turning off the GPS. Generally how I do this is by dumping the FPL and/or turning the big knob to the AUX chapter. Then the trainee has to enter a ground navaid approach and fly it from the IAF to the MA hold with an entry and enough turns to nail the wind correction. You find out pretty quick how good their non-GPS SA skills are that way. And the little blue plane on the iPad gets turned off, too.
 
MAKG1 said:
You didn't read the article, did you?

It's about what happens when the automation fails, particularly related to situational awareness. If you watch a magenta line all the time, can you identify the aircraft's position with respect to navaids? That's a critical question for remaining in protected space under IFR.

It's also a NASA study, not the FAA (and yes, I do know one of the coauthors). I remember the call for subjects, but I didn't participate because I'm not an instrument rated commercial pilot like they wanted.
Click to expand...

That comment is not related to the article - it's just a general observation.

But this is part of my point:

Furthermore, pilots who relied more heavily on the computers to handle these tasks and who allowed their thoughts to drift during flight were more likely to suffer the effects of rusty cognitive skills.
Click to expand...

What are you doing with the time automation saves you? If the answer is sitting on your hands or daydreaming then you're not going to be prepared if some problem is dumped in your lap. But then... is that really the fault of the technology or the pilot using it?
 
Last edited:
Check out the foldout on this month's Flying; a pic of a G1 cockpit folds out to a G500/600 cockpit. The new ones are clean mean automated machines but the G1 is definitely more macho.

"the study found that pilots often struggled with maintaining awareness of the airplane’s position"​

I'm ashamed to say I got caught on my last flight when center unexpectedly asked me where I was. My first GPS drunken thought was "167 miles from my destination" :dunno:

petrolero said:
In some phases of flight, stick and rudder skill is critical and can't be replaced or even much enhanced by automation. That's why we all hand fly takeoffs and landings and patterns. There is too much minute adjustment by feel and split-second human decision-making that is needed in those phases - especially if the engine quits or winds are squirrelly.
Click to expand...
I suspect not for much longer as robotics marches on.
 
Can AP in a commercial airliner land a plane if you didn't kill it at 200 feet?
 
SixPapaCharlie said:
Can AP in a commercial airliner land a plane if you didn't kill it at 200 feet?
Click to expand...

As warthog said, those certified can do it. If it didn't have autoland, it would probably just fly it right into the ground.
 
It can land the aircraft on centerline, track the runway and bring the aircraft to complete stop on centerline. It cannot extend gear or flaps though.......yet.
 
farmerbrake said:
Interesting. So I have a qeustion for you airline guys,
are you required to switch on/off the auto pilot at a cerrtain point or time?
Click to expand...

Only speaking for my last company (I'm between jobs).
In addition to the autopilot limitations. We were required to couple the autopilot in approaches, depending on the reported weather. For example, I believe 500/1 on an ILS it needed to be coupled. The rules varied for non-precision approaches, certain notams affecting the approach, and MEL'd equipment.
 
So certain times then the AP is flying, but it seems most is personal preference on the way to/from cruise right?
 
RotorAndWing said:
Depends on a lot of factors, such as how busy the airspace/ATC is, complex arrival/departure procedure, weather, etc.
Click to expand...

Just like any tool. I love my oxy-acetylene torch, but do not typically use it for oil changes.
 
As a side note, while automation has many great advantages, it actually makes it harder to stay proficient if you don't fly a lot -- you now have twice as much on which to stay proficient (with and without each of the various systems like a/p and GPS).
 
You must log in or register to reply here.
Back
Top