FCU Altimeter Setting

[Capt Scribble]

Our Company has decided to change the Airbus SOP so that ALT selections on the FCU will be made in 100ft increments (for "safety Considerations"). I have scratched my head for some weeks to understand the dangers of setting the ALT in 1000s as we have always done, but have drawn a blank. Does any other Airbus operator use this method or perhaps I am missing the obvious.

[VinRouge]

Could be more hazardous as the panel will be taking the pf out the loop for much longer?

[FLEXPWR]

Some airlines have this for specific flight phases, like below transition level or below MSA.

Making use of it on a continuous basis is plain stupid. Another management powertrip idea, like a dog pi$$ing on a lamp post to mark the territory... Companies are full of these, trying to re-invent the wheel -or teaching birds how to fly- in order to justify their positions...

By putting more rules, we ask pilots to follow blindly and this decreases airmanship and initiative. Any professional pilot will use good judgment to find out when using 100's or 1000's is more relevant to the task at hand.

[Uplinker]

Our company's SOP is to use the 100's setting below transition altitude and the 1000's setting when above it for flight levels.

Doesn't seem to be much of a problem - only takes a second to switch it over.

But Capt scribble, are you saying that ALL your selections must use 100's ? If so, that seems a bit weird. Ten times the number of turns to dial in a big climb ? Pointless and stupid, it won't make selections more 'safe, it will add extra errors because you will have to spin that knob so much that it will probably actually reduce precision.

[lifeafteraviation]

Ha, I guess you don't have to fly to countries that use meters...

Usually such rules or procedures are born from somebody's major screw up....did something happen recently within your company that prompted this as a the solution?

[Capt Scribble]

Uplinker, Yes, we used to use 100s with altitude but now its all the time. As this change is allegedly for safety, I just wondered whether anyone else world wide found setting in 1000s a great risk. By the time I get to the Alt, I'm likely to have forgotten what it was.

[streetcar driver]

If you are cleared for FL320 and you turn the rotary selector just 1 click too far you would have entered FL330 iso FL320 when 1,000ft increments are selected. During phases of high workload, such an error could remain undetected.

With 100ft increments, the selected level would be FL321, which does not look right at all. Therefore, the 100 ft setting increases your chances to notice some fingertrouble.

Of course we are talking of propablities of 10E-x.

[vilas]

There is no standard procedure. But generally below TL or FL100 it is changed to 100s because you can get intermidiate altitudes like 3700ft etc. Also it mitigates the error, mistake in 1000s can easily be below ground level. Some always leave it in 100s but use as required basis but come back to 100s. Using it in 100s only is not logical nor practical.

[VinRouge]

Why not get the PNF to confirm the selection? That way, not only is e selection checked, but also a common understanding of the cleared level that was read back by the PNF. Useful in parts of the world where comms are difficult.

'Just one click' doesn't wash, as a crew should pick up on the fact they have been cleared a nonstandard level.

Not to forget mode s provides yet another level of protection, albeit not universal yet. The distraction of using hundreds I don't like.

[chimbu warrior]

I have scratched my head for some weeks to understand the dangers of setting the ALT in 1000s as we have always done, but have drawn a blank.

This is nothing new; various airlines have tried to re-invent the wheel and discovered (usually the hard way) that square wheels don't work.

What about an emergency descent? If you reach up to the altitude select knob (with it set to 100's at cruise level) and give it an anti-clockwise twist, then pull, descent will be initiated, but here is my guess at what will happen next -

you will pull heading and speed, each in turn
you will read the FMA, and perhaps be comforted by the "THR IDLE, OPEN DES. HDG" that you see
you will then deploy the speedbrake
at about this point you will go into ALT* with full speedbrake at high altitude
you will utter #$%^&*()


The suggestion to use 100's below the transition level and 1000's above would seem to have considered the above scenario, where perhaps your company has not.

If you are cleared for FL320 and you turn the rotary selector just 1 click too far you would have entered FL330 iso FL320 when 1,000ft increments are selected. During phases of high workload, such an error could remain undetected.

Errr....whatever happened to reading the FMA?

Why not get the PNF to confirm the selection?

One would sort of hope that they do this already.......

[Chris Scott]

Using the hundreds setting to change the selected altitude all the way up to the cruise alt and back again will wear the mechanism out prematurely.

We had proponents of this idea as far back as 1988/9 on the A320. The argument in favour was that, at the end of a given selection when you push or pull the knob, you can accidentally click it one notch clockwise or anti-clockwise without realising it. When the setting is on hundreds, that mistake only changes the selection by a hundred feet instead of a thousand.

Some of the guys that came on to the A320 fleet from the B737 and B757 seemed to be particularly susceptible to making this mistake. The reason was that they were used to having the altitude window on the FCU as the only indicator of the selected altitude, and were reluctant to change the habits of a lifetime.

The solution to the problem, of course, is to use the selected altitude indication on the PFD, not the one on the FCU. That way, you will pick up the mistake immediately. In fact, once you have found the right knob, the whole operation should be performed using the PFD indications. (And the same philosophy applies equally to the VS/FPA knob, not to mention the SPD/MACH and HDG/TRK knobs.)

The bottom line is: at the moment you push or pull the altitude selector knob, you MUST be looking at your PFD, not the FCU.

[woodja51]

What that man said... ! The PFD is where you should look to confirm the selection .. Not the FCU .. For all modes selected ...that is the philosophy of the FMA wja

[Chris Scott]

Thanks, woodja51,

Some of our trainers used to cover up the FCU windows (on the simulator) to stop guys using them...

[TyroPicard]

Some trainers still do.....

[PEI_3721]

An interesting discussion. I am not familiar with the particular equipment, but would it be correct to assume that there is both an ALT SEL readout on the FCU near the selector, and a readout on the PFD ?

If so, this follows general design principles (SAE S7 recommendations circa 80/90s), where annunciations should be provided both at the point of selection (next to the knob) and at ‘the point of operation’ (PFD for inst flying / monitoring).

The current use of these systems appears to contradict the design assumptions; apparently due to the risk of mis selection. Thus if true, procedures have been introduced to overcome a design weakness, yet we have to and can accommodate many weaknesses because no system is perfect, but this should not be at the expense of introducing another error: – reaching for the selector knob without looking and choosing an incorrect control; or looking at the selector then back to the PFD with associated increased workload.
There may be no answer to such dilemmas, but it would be interesting to understand some of the rationale and balancing safety cases.

[Chris Scott]

Well, thank the Lord for that. It would be rather sad (although not in the least surprising), if the lessons we learned 25 years ago had been lost by our successors...

Cap'n Scribble, Zirrr,

Can I respectfully ask you to amend your thread title to something a bit less confusing? "FCU Altitude Selections", perhaps? At first I thought you were on about the altimeter sub-scale (hPa/inHg) selector.

[vilas]

Chris Scott
What you said is perfect. That is the way to fly Airbus FBW.

[Chris Scott]

Hi PEI_3727,

Your reference to SAE recommendations is interesting, but I'm afraid your being "unfamiliar with the particular equipment" shows clearly when you go on to remark that:
"[...] but this should not be at the expense of introducing another error: – reaching for the selector knob without looking and choosing an incorrect control; or looking at the selector then back to the PFD with associated increased workload."

So let me explain the situation from basic principles (A320/330/340/380 pilots can go off and make a cup of tea, and/or concentrate on the Six Nations rugby match).

When the AP is engaged, the PF makes his/her own altitude selections using the selector knob on the FCU. (S)He must monitor the flight-path on the PFD and ND (nav display), and - in climb or descent - needs to minimise the amount of time looking at other parts of the cockpit. The PNF monitors the selections made by the PF, and - when they have been completed - calls the associated FMA changes (i.e., any required by company SOPs) as seen on his/her PFD.

So: back to the selection itself, in this case an altitude or flight-level change. The PF switches his gaze from the PFD up and across to the FCU, and reaches for what he thinks is the altitude selector knob. Touching the knob but not yet moving it, he averts his gaze back to the PFD. The values of the 4 parameters that can be adjusted by the 4 FCU knobs - SPD/MACH, HDG/TRK, ALT, and VS/FPA - are all visible on the PFD in one form or another.

The PF turns the altitude knob, simultaneously observing the changing value of selected altitude or FL. If he is turning the wrong knob, this will be apparent immediately. (The easiest pair of parameters to mis-identify on the FCU itself are the adjacent SPD and HDG knobs, whose associated read-out windows both show three-digit values. On the PFD, the parameters cannot be confused.)

Finally, the PF pushes or pulls the knob, depending on what type of climb or descent he wants (or leaves it alone if he wants to use VS or FPA). Now for the crucial bit. The PF continues to watch the selected altitude until his hand is well clear of the selector knob, to ensure that he has not inadvertently turned it. The PNF can now announce the FMA visible on his PFD, e.,g.: "Open-descent flight-level two-four-zero."

When the PF is hand-flying, the selection roles are reversed, except that, if the FD is being used, the PF first has to tell the PNF what (s)he wants; e.g., (s)he might want a change from managed-descent to open-descent, or even VS or FPA. That increases the workload for both pilots.

To sum-up, PEI_3727, there is no harm in choosing the wrong knob, provided you don't turn it until you are watching its effects on the PFD. And the less time your eyes are away from the PFD, the less likely you are to deviate from the intended flight parameters.

[Check Airman]

At my company, the PM makes all altitude selections, never the PF. Both pilots will then call out the new altitude.


Note...I don't fly Airbii

[vilas]

Check Airman
PF handles FCU unless he is flying manually or in some extra ordinary case. If your company wants PM to handle FCU even on AP then your company procedure is against the principle of Airbus flying. Since you are not AB pilot you may have misunderstood.

[Rick777]

At my last airline the PF never touched the altitude selector. On AP of course he did everything else. We had AB and Boeings and it was the same. Boeing's procedures are for PF to handle all selections also. We just did it differently. Whether to use 100 or 1000 was up to the pilot.

[Uplinker]

Just to make an important point:

The FCU altitude readout is the INPUT to the computer. The PFD altitude readouts are what the computer has received. The three should of course always be the same, but if there was a fault somewhere, they might not be.

Therefore, as has been said the only place for Airbus pilots to confirm the selected level/altitude is on their PFDs.

Our SOP's are for PF to make all selections unless they are hand flying, in which case PNF makes selections on request for them. If a new Alt/FL is selected, whoever made the selection calls what they see on their PFD, including the colour of the digits*. The other pilot looks at their own PFD and also calls what they see. The two should of course be the same, but this is an extremely important check to make, because it can catch incorrect selections, computer faults and therefore Alt busts.

*Because on Airbus, constrained intermediate altitudes are coloured magenta, unconstrained ones are Cyan (though we say "blue").

[Chris Scott]

Hello Uplinker,

You make a very good point about the greater integrity of the indications on the PFDs.

Your fleet's SOP for FCU selections sounds similar to our A320 fleet's, although things may have changed in the 12 years since I retired (after doing the first 14).

The main difference was/is that the pilot who made the selection did NOT make the FMA call - the non-selecting pilot did it, unprompted. The logic for this is threefold, and designed to address human fallibility in high (and even very low) workload situations:
(1) the pilot making the selection may "see" and call what (s)he thinks (s)he has selected, or even call before fully completing the selection;
(2) to force the other pilot to pay full attention to monitoring the mode change, rather than (perhaps) merely parroting what the selecting pilot has called;
(3) to halve the number and duration of calls.

Re (2): occasionally the non-selector will miss a mode-change and/or fail to call; due to other tasks or distractions, or ATC calls. The selecting pilot is then tempted to make the call, but I advised people where practicable to choose a suitable moment and ask something like: "Check your FMAs?"

Including all mode changes involving manipulation of the ALT knob, our FMA calls were strictly limited to those that are safety-critical. When we started A320 ops in 1988, several of us, including the chief training captain, had previously been on the BCAL A310, which has an AFS nearly as complex as the A320. On that fleet, we had called every mode change, but it was realised that to do so on the A320 - with considerably more of them - would be counter-productive.

[PEI_3721]

Chris, many thanks for the thoughtful explanation.
This is much as I had imagined, although the operational description highlights the biases of my dated experience. With the (then) relatively new 2 crew flightdeck, including early EFIS, AFCS, and FMS, the systems did not have such an extensive suit of controls and displays.
Subsequently, although the number of controls has been limited (occasionally and hazardously dual purpose) the automated systems capability has undergone significant expansion enabling a range of operations which require greater clarity in depiction / annunciation. This has led to complexity, which generally has been managed with training and procedure.

My concern is twofold:
First; that reliance on training and procedure to manage increasing complexity is approaching a limit – particularly in combination with situational/task workload and organisational/systemic demand, e.g. complex GA procedures, path and communication demands with extensive dependence on cross monitoring – see BEA AGASA study.

Second; that extending or changing the methods of managing with procedures (however well intended) might encounter situations which have not have been considered, neither during design or procedure change. These may generate some surprising incidents, often cited by modern SMS views as emergent hazards, as they are a property of the total system in which we operate (Man, Machine, and Environment).

Thus although a procedure has been well considered for ‘normal’ operation, the demands on human performance, due to the developing and complex operational environment, and interwoven safety responsibilities (regulator, operator, airport, ATM), can change the ‘normal’ to that requiring exceptional human capability (again see BEA AGASA study).

Thanks again.

[Capn Bloggs]

After that explanation from Chris, It's Boeing for me. PF does it all, the PM checks it all. Surely the operation isn't so bad/incompetent that the PM has to call the result of a selection by the PF? Surely the PF isn't going to call something incorrectly from the FMA, followed by the PM also not checking that what has just been said is actually correct?

As for:
occasionally the non-selector will miss a mode-change and/or fail to call; due to other tasks or distractions, or ATC calls. The selecting pilot is then tempted to make the call, but I advised people where practicable to choose a suitable moment and ask something like: "Check your FMAs?"
The obvious solution is that, if the PF doesn't get the answer he needs from the PM eg "Track 270 selected", then he simply repeats the original statement and waits for the PM to respond correctly. "Check your FMAs?" What for? The PM obviously missed your first call so now he has to play the mind game trying to work out what has changed on his FMA?

Aviate, Navigate, Communicate comes to mind...

The three should of course always be the same, but if there was a fault somewhere, they might not be.

...The two should of course be the same, but this is an extremely important check to make, because it can catch incorrect selections, computer faults...
That's getting a bit paranoid, isn't it, creating an SOP just in case the hard-wiring might be faulty? I fully agree with the PM checking (calling what he actually sees if you insist) what he sees on his PFD but using a possible system fault as a reason for a SOP is going a bit over the top IMO.

[Check Airman]

PF handles FCU unless he is flying manually or in some extra ordinary case. If your company wants PM to handle FCU even on AP then your company procedure is against the principle of Airbus flying. Since you are not AB pilot you may have misunderstood. At least 1 other company which operates types from 3 different manufacturers (including Airbus) employs the same technique. Perhaps it's a US thing. I like it for 2 reasons:

1. Altitude busts are always a hot topic here. The PM is kept in the loop. This is particularly true with level changes at the pilot's discretion, where the PF may opt to level off at an intermediate altitude.

2. Consider a clearance I was given last week "you're 6 from the marker, fly heading 120, maintain 3000 until established, cleared ILS runway 9. 170 or better to the marker."

As I'm making mode selections for speed and lateral path, the PM is taking care of the altitude. One less thing for the PF to do. Not much extraneous talking either. The only required callout in this case would be "3000" by both pilots. As a technique, one pilot may state "cleared for the approach", but that's about it.

[Capn Bloggs]

If instructions from ATC are so complex (5 instructions in this case) they require both crew members to "do their own thing" and then have a conference after it's all set to confirm that what they just did individually is actually what is required, there is something wrong.

What's wrong with write it all down, then execute in a methodical, controlled way with ordered setting and cross-checking by the crew, together?

[Kak Klaxon]

It's a UK / USA thing. I have both licences and if I fly with an American pilot our hands will often meet at the Alt selector as a lifetime of SOP is hard to break.
Same as V bars and cross pointer flight directors.......just different but the same job is being achieved.

[Uplinker]

Hi Chris, that is an interesting thought, and I see where you are coming from. I've just double checked our SOP's and it says that all boxed or Alt changes are called by PF and confirmed by PNF.

Capn Bloggs, you might characterise it as 'paranoid' but I wouldn't. It probably came in when Airbus were new and not fully 'trusted', if I can put it that way - however, I personally think it is a very sensible check. Too many people don't understand this input/output distinction with computers, and you do still see pilots looking only at the FCU alt readout. Remember garbage in - garbage out? Now that computers are in the loop as opposed to just 'switches and wires', things may not be what you expect. I have never seen it go wrong in an Airbus, but if a wiring fault or a computer fault occurred one day which dropped a digit and the computer saw your selected 17,000 feet as 7,000 feet or 1,700 feet, you might be in trouble - and well on the way to an alt bust if you hadn't checked and noticed the discrepancy.

[Chris Scott]

Hello Capn Bloggs,
It probably looks complex because I tried to cover all the scenarios. Re Boeing, there's nothing in the system I describe that cannot be applied in any 2-pilot EFIS cockpit with an FCU, provided it also has a full suite of FMAs and selected parameters displayed on the PFDs. (Wouldn't work on a B757/767, for example.) I'll try and answer your points one at a time. I see that most of them seem to relate to operation with AP, where the PF makes his/her own selections?

"Surely the PF isn't going to call something incorrectly from the FMA..."
Wanna bet?

"...followed by the PM also not checking that what has just been said is actually correct?"
If you are constantly spoonfeeding him, he might do just that - through sheer boredom, or even the assumption that the boss (or hit-shot youngster) always knows what's best and gets it right. To pay full attention to cross-monitoring, he or she needs to be motivated. Puttting the onus on him/her to read his FMAs unprompted gets him back into the loop. Some call it CRM.

"The obvious solution is that, if the PF doesn't get the answer he needs from the PM eg "Track 270 selected", then he simply repeats the original statement and waits for the PM to respond correctly."
What "original statement"? I infer your scenario is that the PF has made a selection, and has already announced the FMA change. In "my" system, the PF would never have made that "statement". It's the responsibility of the PM (the non-selector) to make the FMA call, to avoid being suckered into the same mistake already made by the PF. The pilot who has made the bad selection is likely to call what he thought he had set. He and the PM are more likely to notice the error if the PM reads his PFD and makes the call unprompted. And what's the point of BOTH pilots making the call? Just adds to the noise.

"...'Check your FMAs?' What for? The PM obviously missed your first call so now he has to play the mind game trying to work out what has changed on his FMA?"
No, he simply has to read his pitch-control FMA, plus the selected altitude. We did not call A/THR or roll-control FMAs; except on take-off, go-around and localiser intercept.

"It's Boeing for me. PF does it all, the PM checks it all."
Surely that only applies when the PF is using an AP? I presume that, also on your a/c, when the PF is handflying with FD, the PM becomes the selector, and normally has to wait for the PF to call for what (s)he wants?
In "my" system, the PF calls for what he wants, monitors the PM's selections, and then calls the result. The PM says nothing, unless he requires clarification.

"Aviate, Navigate, Communicate comes to mind..."
Precisely. Keep verbage to a minimum. Avoid duplication, or mere rhetoric. Ensure that all commands, responses, and standard calls are as concise, unambiguous, and effective as possible.

[Capt Scribble]

Thanks for all the considered replies. Ultimately, it doesnt matter how you twiddle the ALT selector, the FMA is there to confirm its been set as you want. It seems that our Co have just made the process that much more fiddly under the flag of "safety", when there is no such justification. KISS... Onwards and upwards!