I cannot believe that in 11 years of flying you have not been in a situation where it would be better to disconnect rather than trying to let the autopilot "cope" with the situation. Do we not fly the same equipment?

I don't claim to be an excellent pilot, nor will I ever disconnect AP, AT on an approach (whether its IMC or VMC) without the approval of the PNF. As many have already mentioned there is a time and a place to do this. I'm no idiot.
Bottom line is though that flying an ILS is a pretty basic flying skill.
I would be terribly ashamed if I would not be able to do this.

Absolutely fantastic CRM. Do you also beg the approval of the PNF before you go to the toilet or to blow your nose. :ugh:

You've perhaps never flown a "classic" airplane: on the 747-200 which I had the absolute honour and pleasure of flying for almost three years, disconnecting the automatics was often absolutely the only way to get out of sticky situations (like the autopilot turning the wrong way -towards the mountains of course- on localizer intercept).

On the equally lovely 747-400 I have been on for the last two years, I have already had to disconnect the A/P at least once as it was clearly doing something patently silly. Come to think of it, even on the ultra-modern A321 I flew for seven years, a couple times it went "wonky" leaving no alternative but manual flight (one classic example was the dual FMGC timeout as I recall).

The bottom line is, automatics are wonderful and quite reliable (especially on newer airplanes), but sooner or later they will fail/misbehave. As long as the guys/girls on the "sharp end" are the type that use automation as a mere tool, are well aware of the attitude and thrust required for each phase of flight and are comfortable flying manually, it will be a non-event.

Automatics are your slave, NOT the other way around. By all means use them, but depend on them at your own peril. :ok:

The following, I think, are facts:

1- the better you hand fly, the better you understand automation
2- the better you understand automation, the better you can master automation
3- the better you are at hand flying, the safer pilot you are
4- the better you are at automated flying, the safer pilot you are

wait, a better order would be 4,3,1 and 2, maybe

Microburst2002: I couldn't agree more. :ok:


Shifting the debate slightly to one side, perhaps some of you more experienced pilots might be kind enough to suggest a practical approach for inexperienced pilots to improve their handling without letting their inexperience endanger the operation?

Putting the query in that way risks an outpouring of scorn from people who see my inexperience as a weapon with which to oust me from my 'shiny jet', but I firmly believe the self-improvement option is the only way ahead in a risk-averse industry. We all know there is a time and a place for 'practising'; so what do you think a 1000hr baby should be doing?

MH152

Microburst2002 - 2,3, and 4 OK, 1, NO.

There are still people out there who will, as a first reaction to any problem, dump the autopilot and start handflying without proper analysis of the problem, the annoucement, "Disconnecting" usually comes a second or so after the buttons are pressed!

More than once I have seen the automatics dumped, (on B747-400), when they were actually doing what they were programmed to do, either the FMC had been incorrectly programmed or the MCP had been mishandled - No, automatics are not always right, but, on aircraft such as the B757, 767 and particularly the B747-400, more often than not they are.

IIRC this turned out to be a major factor in the Kegworth accident. The decision to shut down the (working - incorrectly diagnosed as faulty) starboard engine occurred simultaneously with the decision to disengage A/P and A/THR. This had the effect not only of masking the very real problem in the port engine by reducing the fuel flow to near idle in preparation for the descent, but also markedly increasing the workload on the flight deck, so much so that when the Captain started troubleshooting - "Now, what indications did we actually get?" - he was interrupted by having to receive and reply to radio calls, the opportunity to correct the mistake passed and the rest is history.

Not being a pilot myself, I can't say - but most of the correspondence I've received over the years, as well as a considerable amount I've read on here, tends to indicate that modern automation is about as good as it can get, and while it is no substitute for analytical human minds on the flight deck, any assumption that an unexpected change in vertical or horizontal track must be due to a fault in the automatics (as opposed to inadvertent mis-programming) is likely to be mistaken. If you're not happy, then disconnect the FMS routing, but surely using a basic altitude hold while diagnosing the problem should be the next step. If that doesn't have the desired effect, then go hand-flying, but bear in mind that it will consume a lot of effort - possibly more than you anticipate.

To be clear, I'm not saying that handflying should be a method of last resort. If it's a good day, your airspace is relatively clear and you feel comfortable, nothing should stop you. However, if and when things head south one should give careful consideration to the increase in workload that handflying will put on the PNF, as well as yourself.

But in your post you say "Not being a pilot myself, but"

I honestly don't mean to denigrate your sentiments but I believe you do need to have been a jet transport pilot with both significant hand flying and automatic pilot monitoring experience to understand the subject under discussion. Hand flying these aircraft should be a normal accepted skill whatever the weather conditions. The automatics are there as an aid to navigation in general - they should not be treated as a crutch because of a pilot's lack of competency in pure flying ability. Judging from accident investigation reports pertaining to jet transports it seems the latter is now the norm.

Hopefully a bit more than that! Under the right circumstances, properly programmed, they can 'fly' the aeroplane whilst the humans sort out the problems, with the correct emphasis on the problem as well as flying the machine, more relevant in the days sans Flight Engineer.

As DozyWannabe points out, Kegworth was a classic example of a time when full and proper use of the automatics may have brought about a better outcome.

No argument there, just don't ignore the automatics when they can make life a bit easier, as a public transport pilot you have nothing to prove, there are no machismo prizes on offer.

is not exact, I agree. It should be ammended as follows:

...the better you may understand automation.

which means that if you don't study the books frequently and learn from experience and keep studiying from time to time, you will never fully understand your automation.

If you lack hand flying skills (for instance because you hardly have flown a few hours in a cessna, after which you directly came to the jet airliner and only hand flew it a little bit in the sim) automation becomes a "black box". You insert inputs, and outputs are the result, but you don't understand how things work inside the black box. That is, in my opinion, inacceptable from a professional pilot.

I sometimes get surprised when fellows with over 2000 hours on a jet demonstrate a very poor understanding of the principles involved in flight and how they have a lot of misconceptions about automation modes.

If you come from another jet airliner that you did fly manually frequently, then you will understand easily any automation, as long as the aircraft still has wings and engines.


The following, I think, are requisites por a transport pilot:

1- master your automation
2- don't be afraid of hand flying (no AP/FD, no A/THR)

I think that regular hand flying practise is good for both

Really? A lot of NTSB investigators (including Greg Feith) never flew the line, and I like to think I do my research properly (My well-thumbed copy of HTBJ is never far away). I may have worded things badly, so...

I agree with you completely (except for maybe the last sentence - there are still accidents and incidents where premature disconnection of the automatics increased flight deck workload). I think that airlines limiting hand-flying time is counter-productive, but at the same time I believe there's a time and a place to practice when you've got a planeload of people in the back. Would you disagree that practicing handflying through severe weather is something that largely belongs in the simulator - for the safety of pilots as well as passengers?

Parabellum has done a good job of putting what I wrote more succinctly.

Precisely. The role of F/E has not "disappeared" as such - the troubleshooting role that the F/E once filled has been divided between the two flight crew at the controls (with the aircraft systems management role being handled by the automatics). The troubleshooting role becomes a lot easier to play when you're not trying to manually handle the jet at the same time. Again, I think that parcticing handflying in a troubleshooting situation should definitely be part of recurrent training, but in these modern times that also belongs in the simulator. Any other time, as long as you're confident of your handling skills, go knock yourself out! Just please understand that along with the rest of the human cargo, I want you to be *really* sure you can handle it... :)

I agree and that brings us back to my thread starter: But I also agree with Tee Emm when he says "I honestly don't mean to denigrate your sentiments but I believe you do need to have been a jet transport pilot with both significant hand flying and automatic pilot monitoring experience to understand the subject under discussion. "

No offence intended, so please don't take it as such. I don't know your background (nothing in your profile) but unless you have spent considerable time on an airliner flight deck, an even if you made a good post, you're not in a good position to fully understand the subject.

Respectfully,
Sabenaboy

@sabenaboy

My views on the subject are in agreement with yours, and I probably expressed them best here:

For the record, I'm just a software engineer - but one with a love of aviation since childhood which manifested as a stint as an Air Cadet and an intent to join the RAF until I turned 16 and got long hair, rock music and pacifism. Part of me would love to fly airliners for a living, but the fact is that I simply couldn't afford it in this day and age. As far as this forum goes I try very hard to not stick my oar in unless the info I have is confirmed and/or documented, and I'm always willing to accept if I've overstepped my mark.

As far as the technical aspects of hand-flying airliners go, naturally I'm speaking from a position of relative ignorance (but like to think I'm reasonably well-versed in aeronautics, aviation history and accident investigation for a layman), and as such will defer, but as far as the business aspect of the things you are discussing goes - believe me it's not just happening to airlines. I got some positive feedback from a post I made the last time this subject was discussed, which is here:

http://www.pprune.org/tech-log/43780...ml#post6183737

The knowledge I have regarding high-level design and specification of modern automatics (particularly as they relate to the A320 and her descendants) came from my Software Engineering/Reliability professor, Peter Mellor - who you'll find deeply involved in many of the discussions on the development of airliner FBW going back to the late 1980s. The A320 story became the introduction to our first-year Software Engineering module - both as an example of how complex, multiply-redundant safety-critical computer systems are designed and specified, and as a cautionary tale - how important it is to make sure that things are done right and how dire the consequences can be if they are not. It reappeared in the final year Software Reliability module, which went more into the detail of how redundancy was designed into the systems and implemented.

This discussion doesn't really relate to that a great degree, since for the most part we are talking about modern autopilots (FMS/FMC etc.). I'm willing to stay in the background here, but as an interested observer I hope you don't mind me putting a view forward.

@DozyWannabe
Absolutely not. Thanks for your interesting contribution. :ok:

By all means Dozy.

But you must give the majority of us credit for the fact that we do not hand fly into busy TMA's, we do not hand fly through bad weather/around bad weather, and most probably do not hand fly in IMC. We do not do it because we can't, but for the fact that monitoring becomes a much more important task during these times. Good pilots know how to manage workload as well as many other things, and as such hand fly at appropriate times.

I would add though that pilots do not have a good grasp of punctuation or spelling (for pilots add I)!

I fly because I love it (bar the night flights!), I loved it since my first AEF flight out of BOH on the Chipmunk, why on earth do I want to manage the whole flight, bar the last 400', via the MCP and FMS.

Regards

GBD

Believe me I do - in fact I'd be mortified to think I'd implied otherwise!

Absolutely. That thread I linked to contained some information that I frankly found frightening regarding how some operations were making junior F/Os worried about hand-flying the aircraft under pressure from the business to keep costs down, and I wondered then as I do now what's going to happen when those junior F/Os become Captains in due course.

I do agree with your other statements, however that i do not share. Yes, we should not overload us and the other pilot, however flying in IMC is a major part of our skill set and as thus it needs to be trained. Especially to be able to call upon it when everything is not working out as it should.

I agree with the other statements, but I will hand fly in "benign IMC"
A few harmless (layered, non convective, no embedded CB's forecast or reported) cloud layers and good visibility below a cloud base of let's say 500' AGL would not necessarily stop me from hand flying raw data down the ILS.
With lower cloud base and or low visibility I'll let "James" take care of tracking the ILS.

Even though I've hand flown the A320 simulator without F/D along the ILS, to a full stop with 25' ceiling and 100m vis, doesn't mean I'd ever want to do that in the real a/c! :)

That just about sums it up.
I can only speak for longhaul but why anyone would want to practise raw data flying after a 8-12 hour overnight sector defeats me. :confused::confused:
Such flying can be practised on Caribbean visual approaches if one is desperate. And on my fleet the A/T is always on.

Just thought the last minute in this episode of ACI was worth a mention, I suppose any awareness the public gains into this can only be good.