Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

Man-machine interface and anomalies

Tech Log The very best in practical technical discussion on the web

Man-machine interface and anomalies

Old 12th Apr 2012, 17:58
  #21 (permalink)  
 
Join Date: Jul 2009
Location: DFW
Age: 61
Posts: 221
Likes: 0
Received 0 Likes on 0 Posts
RR, I find myself without adequate time to do your topic justice, but I will initially say this; in my mind two differing interfaces need discussion. the "man - machine" interface and the "human intelligence - computer intelligence" interface.

The man machine interface discussion is easy, at least for me. I want the machine to give me the information in the least confusing manner, in the most usable manner. And I want the machine to allow my control inputs to be made from the perspective of and in the manner that the average human achieves control in abnormal situations.

For example, I have a talk radio station playing on my computer as I type this and I realize that even though I am interested in listening to the topic being discussed. I have NO idea what was said during the time I focused on typing this reply.
IOW, humans only have limited attention, or focus, and it is fairly well accepted that humans have limited in focus when in stressful situations (tunnel vision) ; so why does an aircraft designer build a interface in which multiple inputs are offered? Why does the operator build a process that requires attention to peripheral issues? The Airbus cockpit, and other modern airliners as well I assume, virtually explode with data in an abnormal situation. Bells go off, lights flash, clicks sound, horns sound, ECAM messages appear, more ECAM messages appear, etc. Operators train pilot enslavement to the machine, AB procedures would convince a pilot that the proper technique in clearing an ECAM message is more important than actually flying the airplane.

But that is only a tangential issue. Back to the topic.

The intelligence interface becomes the problem, especially in the AF447 example, and that is the what I think you really desire to discuss.

The Airbus has a sort of AI, or at least a memory bank of programmed data/intelligence. The problem I see is that the "machine" was designed to be flown by reference to the programmed intelligence with oversight of/management by the human intelligence. The problem with the AB approach is simply that the machine interfaces with the human through the AI interface. When the AI is unable to fly the machine, the human is forced to interface with the machine using the same interface that just failed. When one ponders this, one wonders why the designers did not think it necessary to provide a "last resort" interface that allows the human to interface directly with the machine.
TTex600 is offline  
Old 12th Apr 2012, 18:13
  #22 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes on 0 Posts
Direct Law, then?

Sequential degradation is a myth, and 'graceful degradation' is a tragic and broken promise.

Tex, you have framed the problem in solid Gold.
Lyman is offline  
Old 12th Apr 2012, 20:33
  #23 (permalink)  
 
Join Date: Dec 2002
Location: UK
Posts: 2,487
Likes: 0
Received 25 Likes on 12 Posts
TT, “The intelligence interface becomes the problem” (#22).
I have some agreement with this view, but not as you describe – you cannot equate human and machine intelligence.
Whatever limited ‘intelligence’ aircraft might have is the result of human ‘programming’; so “the bells, lights, clicks, horns, and ECAM messages” will have some prioritisation relating to the safe operation of the aircraft; this is specified by the design team. However, as the design team cannot foresee all possible circumstances there remains a requirement for human ‘intelligence’ to seek out specific information, and apply understanding, but this depends on the context of the situation and background systems knowledge. Thus the ‘intelligence’ interface is a partnership which requires a balance in activities, this is influenced by the overall system context. It is not an easy balance as technology never holds responsibility (exercised via human intelligence) – ‘technology is dutiful but dumb’.

The human has the advantage of being able to construct context – understand the situation (which can be flawed); the machine is very limited, e.g. how to differentiate between a null (failed) value, an erroneous value, and a valid signal.
This is not a failure of the machine intelligence, but a limitation of design and engineering. It is not the division between being able to fly the aircraft or not; – the lack of auto flight control and envelope protection are consequential issues.

Serious problems arise at the limits of the machine ‘intelligence’ where the interface becomes ambiguous (‘capability’ actually as machines cannot think). In these situations the human is required to change mode of operation (thinking) which requires deeper understanding of the situation and its implications.
Many accidents involve the failure of the human to recognise the need to change ‘mode’, or if recognised, then the abilities in the changed behaviour are limiting – all of which is influenced by context.

Without intelligence, technology cannot describe its graceful degradation; whereas a human can indicate the approach to limiting performance, technology just quits.
Why should we expect anything else? Many posts state what they need from technology – this is usually context dependent and biased (due to human cognition), but consider, what if the reasons behind the human wish lists are flawed; we have inappropriate expectations or mental models of the big system involving humans and technology.

An alternative view: “The Design of Future Things.”
safetypee is offline  
Old 12th Apr 2012, 20:52
  #24 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes on 0 Posts
An interface as we discuss must be hierarchical. Value, limit, and ability (consequence) must be innate, self evident. And without any prejudicial load.
Expectations can be the enemy of the well designed system.

At its most basic, Command. One must be taught more importantly than power, is the ability to prioritize importance, not power. Skill must be measured in a form of humility that honors each, the machine, and the man.

Not really a training issue per se, but attitudinal. The 200 hour pilot is not contra indicated (preferred?) if the format is taught without bias, of any kind. In a sense, traditional pilot training is in many ways a negative v/v fluid exchange of command. Is it patience, then, until all the old ones go?

Last edited by Lyman; 12th Apr 2012 at 21:15.
Lyman is offline  
Old 13th Apr 2012, 16:12
  #25 (permalink)  
 
Join Date: Jan 2012
Location: Not far from the edge of the Milky Way Galaxy in the Orion Arm.
Posts: 510
Likes: 0
Received 0 Likes on 0 Posts
Grrr Man Machine Interface

Rule No.1 Always keep the aircraft in the loop.
Rule No 2 Man is boss over the machine.
Rule No 3 If machine disagrees with Rule No 2 Interface is replaced by Inyerface and man MUST take over.
>
>
bleepxxxxxwhirrxxxxxxxthats what he thinksxxxxxsniggerrrrrrrrrxrxrxrxrxxxxx
Natstrackalpha is offline  
Old 13th Apr 2012, 16:51
  #26 (permalink)  
 
Join Date: Jan 2012
Location: Not far from the edge of the Milky Way Galaxy in the Orion Arm.
Posts: 510
Likes: 0
Received 0 Likes on 0 Posts
Smile

The 200 hour pilot is not contra indicated (preferred?) if the format is taught without bias, of any kind. In a sense, traditional pilot training is in many ways a negative v/v fluid exchange of command. Is it patience, then, until all the old ones go?
I like you lyman, you are interesting. If I may conceptualise your comment and hope that I have got it correctly - would I be right in assuming that you were saying that if the 200hr guys and gals approach this interfacery with an uncluttered mind, free of bias or greater depths of absolutes and worldly knowledge, then they (s/he+machine) can get along - `till ode fogies `av slipped mortal coil, never mind surly bonds uv Earth.

Maybe, but - See Rule No.1 and Number 2. above lest, the ground rise up and smite them. ok?

Automation is a tool, you would no way give up to an automated diety anymore than you would trust your tin opener to plan, prepare and cook a banquet for 15 people.

If, if, if the `lil `ol aeroplane did what it wanted and A L W A Y S ended up landing beautifully at destination all on its little lonesome, come hell or high water - then yes-

but until this has a) happened and b) been proven to happen 100% of the time then the best we young `uns (who just look old) can do, is steer the wee bairns in the right direction and hope that the desire to live is instilled into them in such a way that even the cosy rosy flightdeck won`t lull them into a false sense of security. Remember phrases like Garbage in Garbage out referring to computors like the FMS - is a man factor. Remember, things go wrong but that certain machines - mentioning no names, like the A320 Airbus, will not only tell you what is going on but also make optional suggestions on what to do next, backed up by a seriously structured training program like the one you get on your type rating course which in itself is backed up by all your ATPL training and all your CPL training and all your PPL training and all the flying experience to boot, including loads of experience gained flying aeroplanes the pilot (she or he) has a most composite time working through all the problems as they have studied the aeroplane which will do what it is told to do, or, made to do. Pilots of said machine should always be one step ahead of events so that automagicary - if it decides not to work, can be overriden by the pilot by taking it out of Managed mode and putting it into Selected mode thereby, choosing to press buttons to tell the aeroplane exactly what to do (as it has misbehaved, it must be told off) and failing that lop everything off and hand fly the thing yourself - you can do it in other aircraft and you can damn well do it in this and other aircraft still - ad nauseam. Just as they know when it is best to cross the road by looking very carefully for traffic, like lorries and cars, so it is the case in the swish jetty whirry machine with computers - a toy..
Computors are there to help you (a little bit) to satiate human`s need to be lazy b-----ds. Would you trust a stranger who was directing you while you walk fast towards a brick wall while you were blindfolded? So, you cannot therefore, expect to go hurtling along at 8 miles a minute by a bit of kit, without monitoring the whole shabang.

If it doesn`t work properly then downgrade its authority, just as they do with humans. YOU have the power to do this (as pilot) as you are the boss, the human-boss, the president/chairperson - the pilot, the brains and mind of the aeroplane. The aeroplane does not care who is flying it nor what is flying it, it cares about nothing but responds symbiotically to very sensible piloting.

Sure, you can fly Auto-Cento per Cento all the time, no worries, but should something go beyond non-normal ops then you will have to work for a living. There is less interface between the controls than conventional aircraft as fly-by-wire is all the rage and glass cockpits help by stuffing all the info into a nice ergonomic square for you so - you don`t even have to move your eyes, much. Also, feel has all but gone too - as pilots inputs go to the flying surfaces, through computors and via wires with little or no feedback, synthetic? No, not even that, there is simply nothing there at all, in feedback, except for a slight forward pressure induced just before touchdown - thats a tiny bit of feedback for a 4 hour flight..

It is YOUR machine that you are playing with, not a toy that is in command of YOU.
And, all the time it is still a great and beautiful aeroplane, which you FLY, so don`t let a tantrum of a bit of plastic mess that up for you - take over, quit playing Xbox commanders - and fly., it. Downgrading from automagicary means, the machine becomes an - aeroplane, again. And, Aeroplanes are flown by pilots who have learned piloting skills from other (instructor) pilots and from "learning to fly"

Last edited by Natstrackalpha; 15th Apr 2012 at 11:28. Reason: ts Sunday, just wondering whether to go down to the pub.
Natstrackalpha is offline  
Old 13th Apr 2012, 17:34
  #27 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes on 0 Posts
Attitudinal. Case in point? In a forensic follow up, everything needs to be spilled onto the conference table. Everything. All the traces, data, and artifacts. Withholding anything is withholding everything. No secrets. Without transparency, the old ones hold too closely the POWER. I foresee the loss (good riddance) of the old mercantile dynamic of "Trade secrets".

As with the tribes of old, and long before common knowledge, the Shaman had spells, icons, totems, and chants. All to cow the tribesmen into a fear that perpetuated the power of the Shamans' BOSS, the Chief, and his allies abroad.

The interface can be pitched as complex, as with anything. It will be soon understood, and well, by the common man. I am optomistic we will soon see the loss through attrition of the old suits, the liars, the con men, the profiteers.

It's a MACHINE, eh?

Command, not Power. The first requires discretion and wisdom, the second, merely secrets held by the unworthy. The old ones fear the loss of proprietary, and to be found out to be shamans.

sayin.
Lyman is offline  
Old 13th Apr 2012, 18:54
  #28 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes on 0 Posts
One thing at a time, if you are a human, and if well trained, perhaps three. With a PM, you are in 6 ville, and if that is not enough, well, the a/c actually wants to fly, that is why they look alike.

The safest pilots are two hundred hours in, and newly certificated, but you knew that. The most unsafe? Around a thousand hours. Instructive of the human condition, reality, and the traditional construct.

Knowledge is useless generally, without experience. The best pilot is one who can remain confident, in front, and on top, and for the longest time; at least until things settle into something one can recognize.

If one has a longer wait, refer to: a/c want to fly, it's in the planform. Level, powered, and within limits.

Grace under pressure is more important than rote. Pilots have retracted gear prior to rotate, in honor of rote. "Light a Pipe" (with apologies, PJ2?). The effective a/c is built to cruise, a long long way, in comfort and economy. The current paradigm should suffice, save for some bonehead plays, terrorists, and Weather, mostly avoidable.

All things considered, this thread is exploratory, I jump onto a twin tomorrow for Hawaii, and will sleep as a babe, and why not? The interface is called into question here, and collaterally in honor of the dead out of Rio. I don't think it was the interface what done it. In fact, I am certain. Mechanical, and PE, in some exotic blend, and we will never know, not this time, again. Gotta make not only a buck, but as many as logic and hypnosis will provide.

Thanks Nat
Lyman is offline  
Old 13th Apr 2012, 19:14
  #29 (permalink)  
 
Join Date: Jan 2012
Location: Not far from the edge of the Milky Way Galaxy in the Orion Arm.
Posts: 510
Likes: 0
Received 0 Likes on 0 Posts
Well, hmm, either you are a machine. Or communicating in a much more different way than most.

I trust that your (twin) journeys will be forever safe and comfortable.

Ha! Rio. Hmmm, well I DID put the point across and got only one attack. felt like I was trying to move a mountain. Lost also my instructor to policy. I always remember what I see. As - ahh, yes, rest easy - it is a safety bred industry - safety this, safety that . . the higher we get authority, the greater WHACK! we have with safety and the more we can influence things to get safer and safer and safer. It always works, some people think you are telling them that their fly is undone - put them on the one out of Rio they will soon forget their embarrasment. Happily, most , no, all pilots, don`t trust Jack Diddly, which is why they check the b--cks out of everything they use and everything they do, confirming this, and that. keeping each other in a kind of working communicative loop, called, the loop.

Statistically - flying is safer than riding a bicycle.

Also, Captains need balls. To be able to enforce their authority and safety policies, they trust nothing and no-one, lest they (their flights) are "safe" which is why thousands, but, thousands of flights every day are flown without hiccup.

Also, Security don`t like terrorists one little bit.

Ground Crew are also safety orientated. Hope Brasil was not one of your family.
Don`t remind me- I saw it on TV. No way will circumstances get me in that situation.

If it helps, we study a lot of . . incidents to find out what went wrong, so as we can make sure it won`t happen to us. Is is a lifelong . . .quest for ever greater safety Ask any of the pilots on here . . they will tell you the same thing. Happy landings.
======================================

Added the next day - No, Rio was due to them wandering off into the rainforests and there were survivors.

I was referring to this one -

Here is a clip from youtube - courtesy of youtube.com etc., this is the one I meant, of which I think zero survived, extremely not good. (If I remember rightly), the pilots glided away from all the built up areas and landed it in a wood.

Situation, if I remember it, was . . .coming from Brazil, they encountered unforecast headwinds depleting their fuel, then their destination was socked in with bad weather, thus slowing down the number of aircraft landing there for they had to stack up all the aircraft, in like a vertical waiting queue, if you like, everyone is at a different level, (altitude), as each aircraft lands, the next one up in the queue is told to commence an approach and the one above him or her is descended to the next level down in the stack - one can jump the queue in an emergency - Avianca here - as you can hear have told ATC that to hang around is hardly an option and the rest . . well, its on the tape. I dont think they were stacked up I think they were just trying to get what they could out of the approach. There are times when we should be able to beat our way out of a paper bag. sometimes, it actually does pay to declare an emergency (whereby the controller will give you any cleared runway you want) its all fine and dandy being cool calm and collected, serene and reflective, this has to be trainined, or we would all be screaming because they got the wrong colour ice cream in the inflight catering meal, but somtimes it pays, to return to "proverbial has hit the fan = actions stations! And cooly and calmly and swiftly race with the devil to put rubber to tarmac, prontissimo!!

Poignant, for want of a better word. there are other tapes on the dialogue, giving a fuller lead up to this disaster.

Seems to be drifiting away from Interface and spreading across to lack of go juice.


N.B. It was Columbia not Brasil

THREAD - this is more about my fuel thread the other day, inadvertant thread spread.

Avianca Flight 52 - Wikipedia, the free encyclopedia

Last edited by Natstrackalpha; 14th Apr 2012 at 15:24.
Natstrackalpha is offline  
Old 14th Apr 2012, 15:19
  #30 (permalink)  
 
Join Date: Jul 2009
Location: DFW
Age: 61
Posts: 221
Likes: 0
Received 0 Likes on 0 Posts
TO: no one in particular.

TTex600 is just a average dumbarse pilot. I've been doing it for well over 25 years and I've managed to survive, but sometimes only by the skin of my teeth. I represent the vast VAST majority of the men/women flying your tail around the friendly skies.

Either build yourself an heavier than air high speed transportation device that is entirely independent of my human failings and buy your tickets on it, or build yourself a heavier than air high speed transportation device that requires (because I do have the ability to recognize degradations etc) my flawed personage and build it to help me cover my failings.

Recent accident stats show that fuel CFIT etc kill aircraft. Cockpit confusion kills aircraft. IIRC, BritishMidlands lost a 738 some years ago when the crew shut down the good engine, AF447 was lost when the crew couldn't decipher the instruments after suffering UAS,

Recent trends in automation do nothing to reduce confusion when the excrement hits the fan. Give me a bird that gets out of my way and let's me fly it when it can't fly itself, that's all I ask.
TTex600 is offline  
Old 14th Apr 2012, 21:25
  #31 (permalink)  
 
Join Date: Dec 2010
Location: Middle America
Age: 84
Posts: 1,167
Likes: 0
Received 0 Likes on 0 Posts
Natstrackalpha,

Well, hmm, either you are a machine. Or communicating in a much more different way than most.
It's the latter, although at time its seems like the former...
Turbine D is offline  
Old 15th Apr 2012, 08:40
  #32 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 876
Received 0 Likes on 0 Posts
Interaction with the A/C

Hi,

TTex600:

AF447 was lost when the crew couldn't decipher the instruments after suffering UAS


Recent trends in automation do nothing to reduce confusion when the excrement hits the fan.
Airbus SAS designed and introduced their products "in close contact" with pilots, we obviously believe.

Normally all incidents and accidents ideally should provide a strong feedback to the design.

Considering "economics" could we expect improvements in the aftermath of AF447?

Considering what we heard from BEA we could be skeptical.

One possibility here in this thread is to characterize "where is the problem"

There is a trend? i think so. Why? In simple terms is:

A complex System when degrades (for any reason: obsolete pito'ts, harness damage due uncontained eng failure, etc.) presents complex challenges. Crescent complexity should require more R&D investment in the "interface". And crescent "capabilities" in the crew.

The antagonic factors in the problem seems generating a concerning trend. Example:

Require the crew to diagnose consequences of obsolete parts (Pitot's). This problem could be EASILY tackled buy the System. The (lack of) automation here IMHO is because "the Design" doesn't considered important (and/or feasible) help the pilots. Just to inform. It seems the automation was considered more important to act (like auto THS) than to provide means to reduce confusion.

Is there a deep and thorough analysis (from the A/C manufacturers) of possibilities of failures and how pilots would react? Is it possible? The time required to diagnose is compatible with the contingencies you may have?

Pitch and Power "solution", memory items seems tentatives to solve problems without required investment (in the interface).

It seems the "training" is being pressured by lack of investments in the ifce. And in this situation, crew would be vulnerable (to be prone to error).

TTex600:

in my mind two differing interfaces need discussion. the "man - machine" interface and the "human intelligence - computer intelligence" interface.


safetypee:

I have some agreement with this view, but not as you describe – you cannot equate human and machine intelligence.

We will characterize this in our discussion. Indeed interesting point. What concerns me is both issues work together and this mix is complex. Pilots could be required to solve (unpredictable issues) very fast. Is this possible? Or will create a "serious "CRM" issue between the man an the machine, via this complex interface?

When man interacts with a complex interface it seems you have something like a CRM involving not just the crew, but a CRM issue between PF, PM and the machine. Auto THS may be used as an example: A System going to a high FL stall should autotrim to the limit? (13deg NU). And how the should interact with auto features? And in presence of failures?

To be continued
to comment on posts of: Bear (machine ), Nat and safetypee. I am preparing to the denser posts more requiring. .

It's the latter, although at time its seems like the former...
Well, hmm, either you are a machine. Or communicating in a much more different way than most.
I need to be prepared to "on the fly" reconfiguring to discuss with "flying anthropologists".

PS

safetypee,

You put very interesting points and i am studying and thinking. I started to look the issues you commented more recently. And in your thread on monitoring & intervention you are ahead many miles. I need to study to be capable to present good arguments. . You will feel the heat! The theme is fascinating. An R&D rich field. We will fly (with air superiority) over it. TTex600 put the basic. Good to start.

Rule No.1 Always keep the aircraft in the loop.
Rule No 2 Man is boss over the machine.
Rule No 3 If machine disagrees with Rule No 2 Interface is replaced by Inyerface and man MUST take over.
I liked rule #1. We need to be proactive. Is it always possible? The interface must be friendly to allow. The trend seems not allowing. You can't override (in Airbus). System can decide when degrade. And caught you in surprise. This concerns me.

Misguided technology?
The trend has it's own power. Concerning.

safetypee, Thank you good links

Last edited by RR_NDB; 15th Apr 2012 at 15:24. Reason: Add links to quoted phrases
RR_NDB is offline  
Old 15th Apr 2012, 12:03
  #33 (permalink)  
 
Join Date: Jan 2012
Location: Not far from the edge of the Milky Way Galaxy in the Orion Arm.
Posts: 510
Likes: 0
Received 0 Likes on 0 Posts
Man, your English is weirder than a CAA exam question.

Look - it is not a case of Human intelligence V Computor intelligence - because computors do not have intelligence - they do not have a mind, nor do they have a lifetime of emotional and practical intelligence programmed into them like us cool Human Dudes and Dudesses do. We, Humans, designed the thing - so it is unlikely that we would design it if it was going to do us any harm. In training we realise that there is greater future in pointing the gun away from you when pulling the trigger. So, being pilots we have the human ability to switch it off - and - if we get p----d off with it then we will do it. Like the other pilots have stated - there is more facility to switch it off and fly the aircraft than there is for it to switch us off and fly us. This is very important - for if it were not true - then tomorrow, when you turn up at the airport - all the aeroplanes would have gone, disappeared. They would have taken-off into the night and they would be sunning themselves on some sexy beach in a deep blue ocean - surrounded by other blonde aircraft and holding a cool Pina-Colada in each hand. So, all the passengers would need a refund as there would be no aircraft left to fly them. Although, they, the aircraft, would be doomed - as, when they tried to take off from the lovely sexy beach, to fly somewhere to get some fuel, they would not be able to move. Not because they are fat on Pina Colada, or exhausted due to too much blonde, aircraft company, but because there is no hard tarmac on a beach and like the dinosaurs, they would sink into obscurity and be replaced by aeroplanes that are designed, manufactured and flown by hu-mans and hu-womans who have been filtered through a gate of training and, if they pop out the exit of the training the wrong shape then they don`t make it to the flightdeck. To make sure, they, the pilots in training, are taking shape, they are tested, not only on their flying ability but also on their ability to reason. this testing is provided to them, covert, to test their psychological make up - not only in their medicals which they have to pass every 6 months (or more if they are ill, with lassa fever or something) but also by trick questions, especially in their early years of flying. They have to think out of the box, back in the box and be trained in such a way that in the rare event of an appalling flight, it will seem like easy peasy lemon squeezy - and indeed be so, as usually they, the pilots, have been to hell and back in order to pass through the training.
They are trained to fly aeroplanes, having a computerised Xbox type thing is a glittery luxury, a `fun` thing to have. To operate such jolly kit takes, added training at great expense, thousands and thousands of dollars/pounds/yen (ache, ouch!)

All of the qualified pilots on here - could just as easily fly a fully automated flight from A to B or indeed, fly it manually, as that is what they were trained to do and that is what they have been doing all of their flying lives, `till FMGS came along - and now they get to play with something on those long flights, its fun, for everything that the pilot has just worked out with a pencil, why, there it is in computor LEDs, it is also entertaining to see how accurate the box of tricks can be, and as mentioned before, if it does not behave, then switch it off and fly the aircraft - just as we are trained to do. You know, you should start a thread on Artificial Intelligence.

Last edited by Natstrackalpha; 15th Apr 2012 at 12:17.
Natstrackalpha is offline  
Old 15th Apr 2012, 16:29
  #34 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 876
Received 0 Likes on 0 Posts
Where is the problem?

Hi,

Natstrackalpha words:

As i told in my post the comment was from TTex600:
in my mind two differing interfaces need discussion. the "man - machine" interface and the "human intelligence - computer intelligence" interface.


I agree with safetypee:
I have some agreement with this view,
(my bold)

The "dangerous confusion" may come from an automated interface (with no AI built in). Simple protections can present "inputs" to the crew difficult to be understood in certain situations. This can occur with just combinational logic. E.g. a simple interlock. When we introduced microprocessors (i designed with the first one: Intel 4004) you put the so called "Finite State Machines" in the "loop". They memorize and it's outputs DEPENDS on new inputs and the MEMORIZED data. look, i am not talking of AI. Just saying crew may face situations where an almost immediate understanding could be vital and could be forced by training to clear several memorized states, just to allow the System to return to normal thinking (operation).

Problem as i see is: The interface must be with K.I.S.S. principle built in to allow a safe operation (in abnormal situations). Even dumb interfaces can present challenges to a crew. And training is concentrated on predictable issues. We are capable before new situations to deliver an interface that really ALWAYS help the crew? I think is just not possible to assure. What's the solution? Very simple: a well prepared pilot (not just by Pavlov training, memory items, P&P approach, etc.) can do miracles. Why. Because his Natural intelligence is "open" to highly creative solutions. We may list examples of this and examples of incidents and accidents where confusion was installed from even dumb interfaces. like the LOC of Thiells 727.

You know, you should start a thread on Artificial Intelligence.
After understanding the issues presented by microprocessor based interfaces with just algorithms built in we could discuss something on AI in an "off topic" way. I don't see reasons to complicate the thread with this component. The problems we observe in the current interfaces IMHO recommends attention. Because seems complex enough to, in some cases, cause "CRM issues" between the pilots and the machine. In a sense tha both (partners) don't work together to effectively use all available resources to save the A/C.

AF447 may be is a dramatic example:

1) The plane had no failures* (the sensors were operating as per design)
2) Crew including Capt. had all resources available (engines, system, etc.)
3) The had no chance to "decipher" the interface outputs and were forced to rely on other inputs
4) They aggravated the initial issue and were caught in a "unsolvable surprise"
5) They never understood what really happened

Despite the interface being MUCH MORE advanced than the 727. Confusion. And the result was the same. Crew were put in a "terminal state" requiring a super pilot. Even test pilots could not leave this "state"

* Just a temporary "cold" in 3 sensors

In the aftermath of AF447 case what could be made? It seems to me this is a (dramatic) wake up call to the importance of the interface (and it's characteristics) to the Safety of the "advanced planes". I don't operate "the thing". I know what an interface MUST present:

ALWAYS present to the crew the possibility to NEVER be caught in a surprise impossible to be solved, specially when the plane has no failures. Design (technology limitations, whatever) never could put the crew in the situation PF+PM, then Captain entered.

This (confused crew with no time to even understand) could happen again? I think so.

Why? IMO because the interface could be improved. And probably as i understand better without adding extra complexity, like AI features

Last edited by Jetdriver; 12th May 2012 at 01:13.
RR_NDB is offline  
Old 15th Apr 2012, 17:31
  #35 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 876
Received 0 Likes on 0 Posts
Anomalies

Hi,

TWO anomalies were present in F-GZCP:

1) Sensor anomaly (not capable to operate properly). A well know issue.

2) System anomaly (System was reconfigured by lack of redundancy in the design) Something the System "realized" immediately and never told properly (assertively and with the required clarity) to the crew. Instead the interface informed other issues to the crew. Actually we could never know what PF really received (inputs). RHS was not recorded.

Question:

There is a better way to deal with this kind of problems without relying on pilots capable to deal with external (induced by WX) and internal (inadequate processing of sensor anomalies)?

It's fair to require the pilots (not seasoned pilots) to manage and solve simultaneous anomalies in a rush? (through scan while simultaneously being required to aviate the plane sometimes in difficult conditions)

Or it's just a training issue? You teach the pilots on possible ways to "solve" consequences of anomalies.

IMO an interface must be designed to HELP and not present challenges or surprises. The surprise of iced sensors is enough. To degrade the plane by the first surprise IMO is a way to create chances to confusion. And this is dangerous. Could force PF to "new ideas" that could present "new problems".

In Thiells 727 case a basic crew error triggered anomalies in the sensors and the confusion was due "redundant anomalies" in the interface (alt and climb indicators misleading crew).

In AF447 crew didn't trigger both anomalies, and were required to understand very fast reason of SEVERAL interface outputs, adding to the initial surprises.

Indeed, A/C operated as designed. (no permanent failures) Just a "glitch" (a temporary cold) generated enough outputs making 3 crew not able to decipher interface outputs.

It's not necessary to review or improve the interface? Or it's just enough to train on similar situations (after the more obsolete sensors were substituted)? And the current sensor with know limitations? May trigger again anomalies in the System? The Redundancy issue (lack of) was solved?

Last edited by RR_NDB; 15th Apr 2012 at 17:43. Reason: Text impvmt
RR_NDB is offline  
Old 15th Apr 2012, 18:12
  #36 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 876
Received 0 Likes on 0 Posts
Secondary anomalies

A problem (of interfaces):

Even simple interfaces can generate secondary anomalies. A complex (not K.I.S.S.) interface could always be "deciphered" when you need most?

The responsibility of the interface is huge: V/v FO (PF) of ANA 738 rolling with the Captain locked behind cockpit door.

Or should be simply considered crew error? Automatism (the crew interacts through the interface) can be dangerous. And can act "ahead" of the crew.

Is it possible to stay ahead of the "automated interfaces". Silent THS move, non observable (directly) SS, System reconfig (due sensor and System lack of redundancy) allow you to be ahead of the plane? And to enter the correction loop with a clear and fast understanding to execute your pilot capabilities? (aviate, etc.)

I see a conflict here.That played a role in a big number of incidents and some high profile accidents.

IMO training is not the only solution.
RR_NDB is offline  
Old 15th Apr 2012, 18:16
  #37 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 876
Received 0 Likes on 0 Posts
System, interface (or both) issue?
RR_NDB is offline  
Old 15th Apr 2012, 19:16
  #38 (permalink)  
 
Join Date: Jan 2012
Location: Not far from the edge of the Milky Way Galaxy in the Orion Arm.
Posts: 510
Likes: 0
Received 0 Likes on 0 Posts
Are you sure this is not thread AF--- no 8?

This is to lyal, safeteezee, TTEX600, RRNDB, Lyman and anyone I have forgotten or have spelt their name wrong. (you know, people are beginning to talk about us . ?)
===========================================

Guys. sorry. I was generalising. I thought you were discussing a mere concept, I apologise. Also, I was not being offensive when I asked if you were a machine, but I see you took no offence, excellent.

Taking a more practical and serious view, as you mentioned the AF, I think you mean the 330 / Atlantic?

I can comment on this and it is easy for me as I saw it on a documentry, on the T.V. but - for that reason only. However, much can we "apply" from such horrendous happenings - I also think I can understand and address what you are saying.

The situation in its worst moments (before the final one, ) had obviously provided the crew with a surreal situation.
Also, at high altitude at night having their little minds reversed as little Alice through the looking glass herself!!

At this point the situation is really a problem. I am not going into the detail as everyone knows the conflicting factors. The resultant factor (of this discussion) is what we have learned, which is: At a set power setting, and a set attitude we/they could have maintained level flight (which is the minimum a pilot could ask for) from that point on we would consider the nav implications as we have just taken it out of A/P which had obviously and most noisily tried to kill us. TTEX600 would have already have disengaged the A/P, me too in fact, and gone for Standby Artificial Horizon, which should have been playing ball ok, the standby Altimeter would have been tits up as would anything else from the static vent pitot tube . . system. This is not an Airbus problem, this is a very severe icing problem
Without dragging it all out again, but, as you are interested. Instead of the FMS, well, bon d`accord, it was the ADIRS actually, instead of the brains of the thing saying to itself, hey, we are all iced up therefore none of the information is reliable, it, the system(s), believed all the nonsense it was receiving from the . . inputs from the static Pitot and the ADIRS which complicates things further and everything is [also] derived from Groundspeed, as well as airspeed [inputs] - the FMGS Flight Management and Guidance System - therefore, simply conducts the flight in Association with its inputs, into its system - unfortunately all the inputs were giving wrong information due to the fact that all the sensore (they are not sensors per se` am just trying to Keep It Simple - `cos am Stupid.)

So extreme icing has block correct readings into the system.
The system not only displays incorrect data now but, the FMGS is telling it (the aeroplane) to stuff the nose down as we are losing airspeed. Coupled with some not so yummy overspeed indications all straight through the computor like a laxitive) and, and, stall warner, all more or less at the same time - this is where TTEX600 comes in . . dis-engages the A/P to stop the aircraft from commiting Hari Kari - puts the attitude where he wants it (8deg nose pitch up I think it was) and 80% N2? Sounds plausible for a setting, and hopefully, switched off all theGoddamn noise on the flightdeck, which, was obviously a cocophony, or a cresendo, a mess basically - blame Mother Nature and her huge ice bearing storm clouds and the surrounding air which was cold enough and moist enough to ice up the aircraft beyond its de-icing capabilites - however, the aircraft was flyable. If the same thing happened tomorrow, obviously, we would handle it and think "that was easy" as we have seen it, and learned how to cope with it and how to "fly the aircraft" in this situation.

You know aviation is safe, but it is young, it is barely 104 years old. We learn new stuff all the time. Only during the second world war did they discover what jetstreams (near the tropopause) were and how they could be used by pilots.
We are discovering ever new things about the jet engines we fly - how icing can occur in air which is clear supposedly will only produce ice crystals, how reduction in power at altitude can produce ice induction, no accelleration, vibration and scare us fartless, thinking we have an engine problem, where there is nothing of the sort and there is a cure to this and other "anomolies" most of which are not dangerous, provided we - we fly, the aircraft. Because, we started flying with a simple aeroplane which had little in the way of instrumentation and nothing in the way of FMS, computors, a myriad of pitot tubes and static vents or ADIRS systems which has to balance everything with its gyros, GPS inputs and relative position and its groundspeed - so to fly, is easy, or not depending on which way you look at it.
It is very simple to fly, if not easy . . ? It is not as complex as trying to figure out why the multimillion dollar piece of French pastry is just tripping out of its brain and simply going beserk.

It is just a noise, turn it off and - fly.

EVERYTHING IS EASY IN HINDSIGHT

IF THE MACHINE GETS SICK, SWITCH IT OFF, IT WILL THANK YOU FOR THIS, BY BRINGING YOU INTO LAND - after you have got rid of all the ice. you might even have the ADIRS if the IRS is still in Nav but your ETAs will be up the Swannee, but not to worry, we can also navigate right?

So then its to DR navigation, as we have all been trained to do from our baby flying days known as PPL training.

On that night, rather than the crew figuring out what the machine wanted, should simply have disconnected it Autopilot off - Autothrust Off (TLA to the do-nut, I know)

There it is.

If it does not work, at anytime, then it does NOT work - it cannot cure itself, it cannot figure or evaluate anything. It can only calculate from many different areas, SPEED, PRESSURE, TEMPERATURE, VELOCITY, ACCELERATION, WEIGHT, BALANCE DISTANCE, TIME, THRUST and even KINETIC ENERGY, via many computors, wow! but that is all! If the inputs go wrong, from any sensor to the computor, then - ((say it with me guys, all together now)) "Garbage in Garbage out"
It, the machine, can only do so much, and that is it - it won`t lick its own wounds, it WILL NOT make itself better, nor dial 911 or 112 or 999 or contact PC world, nor will it rectify or stand back from the situation nor can it realise that anything is very wrong, it just shines startling lights and blaring noises so as to wake up the pilot from any idea that this night was going to be an easy ride. AF, your flight in question, was most unfortunate and heartbreaking, but if it is any consolation is HAS dramatically rammed it to our attention this case, this scenario, so that, if it should happen, ever again, the flight will be fine and the people on the aircraft will be fine and the families will all land safely, and not be harmed, not be hurt or afraid - because of what happened over the Atlantic - all these pasengers will be saved, again and again because of that one flight over the Atlantic and all those poor people who lost their lives.

Last edited by john_tullamarine; 15th Apr 2012 at 21:50.
Natstrackalpha is offline  
Old 15th Apr 2012, 20:17
  #39 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 876
Received 0 Likes on 0 Posts
Basic requirements of the Interface

Hi,

The Interface should be:

1) Simplest possible (K.I.S.S. design rule)

2) Allow the crew ALWAYS be capable to control the plane, when required

3) Allow the crew to be ALWAYS "ahead" of the plane

4) Always Help the crew

5)Capable to orient crew how to better enter the loop to act better (fast, precise, etc)

6) Keep crew ALWAYS ready for ANYTHING

7) ALWAYS inform on "what's doing (protecting, processing in background, etc.)

The interface additionally must be:

1) Always reliable for inputs from crew and outputs (inputs to the crew)

2) Never mislead the crew

3) Provide properly most important indications to allow a safe aviate

4) Allow immediate interventions as required by circumstances

5) Allow "recover of what happened" (for design feedback, diagnostic, etc.)

Mac

If i missed something, please include it.

Questions:

1) This is what we should expect from the Interface?

2) The Interface used in A320, 30, 40 and 80 have this characteristics?

3) If not, Why?

Last edited by RR_NDB; 15th Apr 2012 at 20:21. Reason: Incl.: properly
RR_NDB is offline  
Old 15th Apr 2012, 21:53
  #40 (permalink)  
Moderator
 
Join Date: Apr 2001
Location: various places .....
Posts: 7,241
Received 132 Likes on 82 Posts
Chaps,

If I might ask that we restrict the overtly gratuitous use of those sorts of expressions not generally admitted to genteel discourse ...
john_tullamarine is online now  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.