Automation vs Seat-of-the-pants-flying talking as devil's advocate - so no abuse plea

[Dream Land]

Create a corps of button-pushing SOP monkeys. Safer & cheaper by Rat 5

Spot on, exactly what was and continues to occur at my previous SE Asian airline, along with making managers / checkers out of the worst of the lot - based only on their political position.

[RAT 5]

Have heard of airlines whose SOP is A/P on at 1000 agl on takeoff, dual ch A/P ILS to allow an A/P in case of G/A, even if not LVP's, and then disconnect A/P below 1000' agl on finals. How do they pass their prof checks which do include some manual flying (N-1)?

[Lord Spandex Masher]

Well, most people will go "mind if I hand fly a bit?".

No.

[bubbers44]

I guess if you can not hand fly autopilot above 1,000 ft must be a requirement. Some make it 400 ft because their pilots are even worse. We handflew most departures and arrivals and only turned the autopilot on when we got bored. But we are old now so are sorry to see aviation depend so much on automation because of the inexperience of some.

Asian pilots have shown that twice in the last month at SFO how automation makes you a programmer, not a real pilot. They can[t do a simple visual approach in clear weather to SFO. We all looked out the window and landed, they can't.

[flarepilot]

I beg to differ...it is not seat of the pants flying vs automation...it is smart flying vs automation.

seat of the pants flying is what is meant when instruments are not consulted...the last six inches of a landing is a good example of modern seat of the pants.

automation makes it easy to sell an expensive jet to a third world aviation country with no GA or exceptional military training.

[latetonite]

A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools.

[LeadSled]

Ever attempted a dead stick landing in a simulator? if you haven't, then the chances are high you would crash if for real.

Sheppy,
Yessir, and so did every pilot in the fleet (B767), it was part of the cyclic training program, introduced after the "Gimli Glider" accident -- which sets the period.
Right from the word go, no pilot failed to make a runway, even if they deliberately planned a little "hot and high", based on the logic that it was better to go of the other end at slow speed, than not quite make the runway at 140Kt.
Needless to say, it was not part of the "pass/fail" program, but we all thought it a very valuable exercise in a company that demanded a high level of manual flying skill, as well as a high level of competence with the autoflight systems.
Sadly, I am told that the same company now discourages hand flying --- the wonders of new age management??
Tootle pip!!

[barit1]

A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools.

The ultimate result of shielding men from the effects of folly, is to fill the world with fools.
Herbert Spencer (1820-1903)

[Teldorserious]

The endless thread debate -

'Learn to fly'
'But we don't want to, besides the chief pilot won't let us'

(Another plane crash)

'Learn to fly'
'But we don't want to, besides the chief pilot won't let us'

(Another plane crash)

Repeat...

[RAT 5]

When the USA Airbus lost it tail in a vortex on departure the NTSB investigated the airlines training techniques and discovered a fault in their teaching, plus a lack of information from the manufacturer on the matter. In other words it was not deemed to be a design fault.
I wonder if there has been a case, or ever will be, of an accident resulting from a recoverable incident due to lack of basic piloting skill. If this could be the case, and the training dept was investigated and the airline handling policy scrutinised, I wonder if the AAIB would conclude the pilots were 'not fit for purpose' and take action? True, they would have legitimate LPC's and prof checks, but the cat would be amongst the pigeons if it was concluded that the pilots should have been able to save the day but were not competent.

[DozyWannabe]

Originally Posted by RAT 5

I wonder if the AAIB would conclude the pilots were 'not fit for purpose' and take action?

The AAIB's remit does not extend to apportioning responsibility. Like most civil service accident investigation bodies, it can enumerate in detail exactly what went wrong and suggest remedial action, but it cannot directly lay fault.

[sheppey]

automation makes it easy to sell an expensive jet to a third world aviation country with no GA or exceptional military training.

]

Astute observation In fact I thought one of the best illustrations of this was a letter to the editor of Aviation Week which said (in discussing automation addiction) "Using autothrottle on final approach is like using cruise control to park your car in the garage

[Clandestino]

Originally Posted by Capn Bloggs

Just the truth, sunshine

Great. We got an explanation why the Turkish at AMS and Asiana at LAX were inevitable...

Originally Posted by Capn Bloggs

Take off your automation/Airbus sunglasses and try to understand why pilots can no longer fly aeroplanes.

... but I don't get it completely; they were yoked, classic, manly, seat of the pants, definitively-not-Airbus aeroplanes and yet they got wrecked.

Originally Posted by mross

Could you maintain hand flying skills by simulator sessions alone?

Obtain and maintain, provided they are practiced in the sim. ZFTers are walking among us for last two decades and are not particularly abundant in grim statistics.

Originally Posted by mross

Is the simulator of today accurate enough?

It is certified as accurate enough to provide initial and recurrent training. No data support hypothesis it shouldn't have been.

Originally Posted by mross

What are the shortfalls of the FFS? Apart from the obvious one that a FFS is not an aircraft!

One of the major shortfall are pilots who get into them with "It's only a sim" mindset. Good example was made by a certain TRI - 7200 hrs on type yet when the birds hit his fan, he was unable to replicate for real the maneuver he was practicing in sim for ages. Unlike host of others that fly as trained and only get cursory mention on Avherald.

[flarepilot]

dear sheppey

the observation (and thank you) is from a major argument I had with the professor occupying the boeing chair at M.I.T. (mass institute of technology) in 1992 or so.

he was explaining to me why skilled pilots would no longer be needed as boeing and airbus had adopted the philosophy that automation would take care of everything.

I asked him what a 300 hour pilot flying a plane the size of a l747 would do when the gadgets quit.

he simply said: they won't quit.

a really good pilot, who keeps his skills high by practice, and a plane with good automation, intuitive to a good pilot, can't be beat. but a crummy pilot with a good automation plane is asking for trouble.

even now we are finding problems with robotic surgery, and I recall when San Francisco built the Bay Area Rapid Transit system. They were going to run the rail trains without drivers...but the computer system didn't quite work right.

oops (oh and the drivers may be on strike soon)

[vilas]

Hi everyone
Tons of words are written against automation, manual flying to the hilt is recommended and lot is said about the value of experienced pilots against 200hrs guys in RH seat. But what is actually taking place in real world? Are we missing the point? I said before civil aviation is a commercial activity. Commercial viability takes precedence above everything else. Regulatory bodies also are on board when they lowered the requirements to be in RH or LH seat and accept the automation. Surely some studies were conducted and the findings were accepted word wide. Now what is the industry response to this clamour against automation? More automation with better safe guards and definitely not manual flying aeroplanes. Man may not be replaced as yet but surely will be reduced to monitoring role. Say what you like but it is writing on the wall. I have seen Navigators exit the cockpit followed by FEs they also had some points in their favour but now nobody is uncomfortable without their presence. Automation has its merits. Without it longhaul, all weather operations were not possible.

[main_dog]

Vilas, I don't think anyone would argue that automation isn't an extremely valuable tool. The problem is that as long as the automation is fallible (and it most certainly is, and will be for the foreseable future) then a pilot is still required at the pointy end to make decisions and perhaps even (gasp shock horror) fly the airplane.

If said pilot has "kept his hand in the game" and occasionally clicks all the automatics off to practice his skills, then the day those skills are required it will be a non-event. If instead he's a child of the magenta who depends on the automation to make up for skills he never acquired or lost, then chances are that it will end in tears.

Mind you we are all at risk of becoming "magenta children", even if we came from GA/military or "steamgauges": the wonders of LNAV/VNAV (or "Managed" if you're on the bus) can lull anyone into a false sense of complacency. The only effective mitigation strategy is a continuous critical review of what the automatics are doing at all times, and once in a while when conditions allow, disengage all automatics, practice your scan and remind yourself that you can still handfly!

MD

[Teldorserious]

Flare nailed it.

The current myopia is that the gear won't break. This is self delusion at it's best.

AF447 went down because a lighting strike flashed the ROMS, knocked out the tubes, no iron gyros and now in turbulence, you got pilots in the dark trying to handfly an aircraft with no attitude reference.

The unthinkable happened - The plane broke and wouldn't fly itself.

[flarepilot]

I am highly distressed at the course modern aviation (of all forms ) has taken.

Automation...great as it never fails (didn't anyone see "2001, a space odyssey"? Hal, HAL, daisy daisy give me your answer do)

We have shrunk the single most important instrument (airspeed) to a sidebar

We have taken the mental situational awareness away...my brain could, with two VORs , a DME, and an NDB/ADF know exactly where we were all the time.

We are asking olympic athletes to use elevators and escalators and get old and fat. (of the aviation brain that is).


I know how I would design a jetliner...the cockpit would have a giant airspeed indicator and a giant horizon and a giant altimeter, there would be a HUD for airspeed, horizon and altimeter too.

The control system would be Douglas strong...cables, no computer interference and the plane itself would be strong enough to handle me flying the wings off it ONCE to a safe landing.

Of course I would fly it like an old lady to avoid having to use the ''fly the wings off'' saving system.

Now we have a plane...nav instruments and wx radar improvements ...sure...but the Basic instruments of flying are there and the standard scan for any landing would be; airspeed, runway, airspeed runway

instrument conditions would certainly include altitude, horizon and nav.

But we have moved away from FLYING in an effort to make the gadgets happy.


Like bubbers said, autopilot is for when you get bored of hand flying. And if you are performing a maneuver or approach using the autopilot, you better be ready to fly it as well as the autopilot or you shouldn't even try the maneuver or approach.

Even 20 years ago I watched (laughingly) as one pilot I was flying with had a devil of a time HAND FLYING at cruise altitude. He couldn't do it within ATP standards. I finally said, why not descend to an altitude that allows the plane to be a bit more stable. What to do if the autopilot fails?

What to do till the doctor arrives?

Back in the day, a pilot had to fly.

[A37575]

the cockpit would have a giant airspeed indicator and a giant horizon and a giant altimeter

Agree. Have you noticed how most wristwatches now have hands (pointers). Before that came digital numbers. Before that too were hands on the face of the clock. It is quicker to gauge the time by looking at the round dial watch. Don't ask me why but certainly it seems that way.
Round dial ASI's took your attention as their rate of change of airspeed in either direction stood out. Drum type ASI's need interpretation of a different type to round dial ASI's.

Take a look at the "modern" artificial horizon in glass cockpits. Usually a tiny triangle as "the little aeroplane" if you have a good imagination, and nothing like the old type of artificial horizons of yesteryear with a big "little aeroplane" that stood out like dog's balls and much easier to fly on instruments.

The glass cockpit AH's which are usually half camouflaged by coloured bakgrounds are designed primarily for flight directors and often surrounded by a plethora of additional flight information. Garmin displays are like reading a colour blindness chart. Somewhere among those colours is a "little aeroplane".

It may be why the average airline pilot brought up on button pushing often has trouble with basic instrument flying

[flarepilot]

It is incumbent upon the industry to remember that flying the plane is the most important thing one can do.

then comes navigation.

(lump in wx radar in the nav area...but you could clean things up by having a voice say: fly heading 220degrees, instead of cluttering up the cockpit with things so big they detract from basic flight instruments)

then comes communication...heck, modern times could have the controllers switch radio frequencies for you and you could forget about that.

But to lessen the importance of the flight instrument...no wonder we have crashes like Asiana.

And yes, the Air Speed Indicator with a pointer and a vref somewhere near the 3 or 4 o'clock position, V2 also is just about right.