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 7th May 2012, 00:57
  #61 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 756
Snoop

I finally read these 20 pages " The Problem of Automation: Inappropriate Feedback and Interaction, Not Over-Automation. Donald A. Norman UCSD :
I see that the mandatory (with controllability) observability of dynamic systems is the most important point in all the described exemples.
In Norman's paper it concerns the Human Machine Interface often forgiven, as we are looking separately human and machin. The concept of "effective aircraft" is concerned totally, involving the HMI with observability.
roulishollandais is offline  
Old 7th May 2012, 01:53
  #62 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,123
roulishollandais

From Norman paper:

4. TWO THOUGHT EXPERIMENTS
Consider two thought experiments. In the first, imagine a captain of a plane who turns control over to the autopilot, as in the case studies of the loss of engine power and the fuel leak. In the second thought experiment, imagine that the captain turns control over to the first officer, who flies the plane 'by hand'. In both of these situations, as far as the captain is concerned, the control has been automated: by an autopilot in one situation and by the first officer in the other. But in the first situation, if problems occur, the autopilot will compensate and the crew will notice only by chance (as in the case study of the fuel leak). When automatic devices compensate for problems silently and efficiently, the crew is 'out of the loop', so that when failure of the compensatory equipment finally occurs, they are not in any position to respond immediately and appropriately.


Here is the major portion of the problem, aisi. Delegating command is (become) not an enhancement, but a degradation of efficiency. In a delicate system that needs constant monitoring, it is unsafe to lose currency in the dynamic system. More is lost into the automation than is gained by mere "reduced workload", and in an emergent recapture of man control, the system has become functionally unknown to the operator! This is not a major knock on the automatic, instead it points out a glaring deficiency in the fundamental design of the system. Instead of a fluent and fluid "intermodal", we at times have an "extra modal".

This extra (outlaw) mode can be described loosely, IMHO, as: "Loss Of Control"....

Best regards,
Lyman is offline  
Old 7th May 2012, 18:22
  #63 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 756
Snoop

@Lyman & others

Forget to put human left, and machine right : they are both together in the effective aircraft. We do not manage a pile of plates, plate human, plate machine, separatly. Otherwise you will always have an interface problem.

Observability and controllability are minimum required. I learned that when learning first year automation in Paris. They concern the total dynamic system.

Describing human as a system or a part of the system is not a negation of menkind... !
roulishollandais is offline  
Old 9th May 2012, 12:43
  #64 (permalink)  
 
Join Date: Jan 2012
Location: Not far from the edge of the Milky Way Galaxy in the Orion Arm.
Posts: 510
Bl---y machines.(Interface)

Y`know, (sigh) this is wind up city, the thing with machines is that they are fabulous calculators and also precision instruments. They can right an aircraft inito S & L quickly and accurately, they can warn you of this and alert you to that, they can disengage (saying "not my prob, you deal with it") but they cannot "WANT". They do not "WANT". It is womans and mens wanting for survival and civilisation that has created . . .survival and civilisation basically.
Machines do not want to live, do not want to eat a great cheesburger, savour the delights of a perfectly chilled Sauternes with the local pud. they do not want to go out with Britney Spears, or George Clooney (each to therir own gender) They do not want to preserve life. they do not want to save the situation, they do not want to be at home or down at the pub later on that evening, they do not want to take out a real foxy chick. All of these wants is what made man fly in the first place, to create flying, then some shmuck created the flying licences, fair enough, then those wonderful people to administer the licences (which only stinks when they become an authority or ceases to be an administration or organinsation) also, machines are not political, which was nothing to do with this post. At which point - a fellow human could have said to me - "why are you writing such c--p, get the hell out of here" and take control of this . . .post.

Whereas, the computor, This actual computor, will keep on spurting out the garbage as I input the garbage (In). Until a moderator - a man or a woman comes along and deletes or edits it.

Point? Machines have no "sense" no commonsense, no higher sense, no sense of preportion, no sense of humour (thank God!) no sense of living or dying, nor any sense of direction.

Even at basic high school level we study biology and machines.

The difference between machines and humans are two-fold:- firstly, we have a respiratory system a digestion system and, and, a nervous system.(ok, threefold, so I cannot count, what do you think I am a machine??) All of which a machine has not.

So, to use an accurate tool like a calculator and control inputter (A/P)(A/T)(NAV)(INS) and a reader (EFIS) and a screamer and shouter (ECAS) is fine
but bear in mind, "IT" does not know what it is doing nor why it is doing it - in fact, essentially, they (the machines) do not exist, they are just there because we put them there. They have no sense. They cannot be hurt, just bits of plastic and metal bits with wire in them.

Last edited by Natstrackalpha; 9th May 2012 at 13:00. Reason: Can`t afford a secretary.
Natstrackalpha is offline  
Old 9th May 2012, 13:16
  #65 (permalink)  
 
Join Date: Jan 2012
Location: Not far from the edge of the Milky Way Galaxy in the Orion Arm.
Posts: 510
Smile

I learned from an old Navy pilot. His brand of heresy never left me. An aircraft in flight is not lifted by the upper surface, it is pushed, from beneath. The upper surface is not relevant to flight, except to say that it provides a place for the underside to rest against.
It is obvious I am in . . hallowed or indeed sacred company here. e the company of pilots or engineers.

The above is my theory too - problem is that if the upper surface gets rough with hoar frost say, then the thing does not work and stuffs the aircraft into the hedge, during takeoff. Therefore, the airflow must be smooooth for some reason. . . and, say just if the lift is the "push" from below, and that the aerofoil surfs along then, for some reason, only appreciated by engineers, the lift drag thang must be more delicate than we can see. . .meaning, even the slightest . . interruption of the airflow over the upper surface is going to send it all down the Swanny. (or into the hedge)
Natstrackalpha is offline  
Old 9th May 2012, 20:21
  #66 (permalink)  
 
Join Date: Jun 2011
Location: Devonshire
Age: 92
Posts: 300
Going back in time... In the 1952-3 era, a York's calculator was made by Dalton ( not yet a Jeppeson), the A/P when it worked, tended to cycle the elevator about 30 - 40 times per minute, the A/T would be manually adjusted. The Navigator was human and the ECAS was in Morse Code at 20-25 wpm by another human, using a key.No CVR invented yet.( Two Pilots, one Navigator, one Radio Officer and one Flying Spanner, plus a stewardess for 36 passengers.)

My fourth and last flight in a York to KIN was from the UK via Lages, ( two 8 1/2 hour flights to Gander(N/S), then Bermuda to Kingston ). Usually we had routed via Keflavik.

Skyways York G-AHFA was the third airliner to have ditched in the N.Atlantic, on 2nd. Feb. 1953, sending out a position report followed quickly by an interrupted "SOS GA- " ( Like AF447 people knew very soon, WHEN and, to some extent, WHERE, whatever happened had occurred...) The SOS was some 6 hours and five minutes after T/O.

After about 6 hours night flight from Lages, fuel which had been in the 540 gallon Centre tank at Take Off and then subsequently transferred to the two inner tanks, when their fuel had been consumed ( each of which fed both the two engines on their own wing) would start to be used.

The water tank filler (40 gallons ?) was close to the Centre tank filler. I cannot recall whether they were marked differently. Both had similar flaps hinged at the front then fastened down with a Dzuz fastener. ( It had been part of my job as a Second Officer to check that they were screwed down.)

IF, and just if, water HAD got into the centre tank, the four fuel pressure warning lights on the flight deck would have remained out... All four engines would splutter to a stop...

An oil slick was found after a day or two 120 nm. SW of LKP. ( A number of oil slicks were found when first looking for AF447, but not related.)

120 nm. from LKP...? The aircraft was not equipped with LORAN... The sky might have been too overcast for ASTRO. It sometimes is, below F/L 10.0.

At the Inquiry, the Solicitor General " Ruled out Contaminated Fuel..."

How could he have been so certain ? The aircraft was never found...

As IGh in #230 on 7th April wrote :

" Investigating Authorities can be led astray... "
Linktrained is offline  
Old 9th May 2012, 22:51
  #67 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 756
Snoop Human brain

Originally Posted by LYMAN #63 may 7.2012
roulishollandais

From Norman paper:

4. TWO THOUGHT EXPERIMENTS
Consider two thought experiments. In the first, imagine a captain of a plane who turns control over to the autopilot, as in the case studies of the loss of engine power and the fuel leak. In the second thought experiment, imagine that the captain turns control over to the first officer, who flies the plane 'by hand'. In both of these situations, as far as the captain is concerned, the control has been automated: by an autopilot in one situation and by the first officer in the other. But in the first situation, if problems occur, the autopilot will compensate and the crew will notice only by chance (as in the case study of the fuel leak). When automatic devices compensate for problems silently and efficiently, the crew is 'out of the loop', so that when failure of the compensatory equipment finally occurs, they are not in any position to respond immediately and appropriately.

Here is the major portion of the problem, aisi. Delegating command is (become) not an enhancement, but a degradation of efficiency. In a delicate system that needs constant monitoring, it is unsafe to lose currency in the dynamic system. More is lost into the automation than is gained by mere "reduced workload", and in an emergent recapture of man control, the system has become functionally unknown to the operator! This is not a major knock on the automatic, instead it points out a glaring deficiency in the fundamental design of the system. Instead of a fluent and fluid "intermodal", we at times have an "extra modal".

This extra (outlaw) mode can be described loosely, IMHO, as: "Loss Of Control"....

Best regards
Hi Lyman,
I shall answer to your post, give me ten days please.
I want to show you how we can be closer to system theory. Man's brain doesn't work like logic, also scientific's and engineer's brain : so we must stay very close to system definitions, after we first write and imagine with our imagination, our personnality and our sensibility.
roulishollandais is offline  
Old 12th May 2012, 01:26
  #68 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
Reasons for AI SS introduction to airliners

Hi,

Why AI (Airbus SAS) introduced the SS? What reason(s)?

It was representative (the visible item) of the Airbus new control philosophy?

What are the main advantages? And it's disadvantages?

What about the overall cost to benefit compared to original (proven) solution?

This interface is adequate when dealing with possible anomalies?

Why active feedback was not built in?

A similar device would be eventually adopted by Airbus SAS competitor(s)?

Last edited by RR_NDB; 13th May 2012 at 08:40. Reason: Text impvmt
RR_NDB is offline  
Old 13th May 2012, 22:30
  #69 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
Two separated SS with no feedback

On the other hand, it may be that the determination was simply that it was unimportant for the Pilots to "see" the other's controls.

They could position the Sticks between pilots?

What was the benefit portion of the change in approach? Anyone?
Cost? Marketing "tool" (representing a new concept of a protected plane)?

A possibility of course, exists that single pilot operation,
Just SW modification? Please clarify.
Awaiting BEA's opinion on the matter....
Too late...


What advantages? Do you know? Cost to benefit? Was a good decision?

Who gained?
RR_NDB is offline  
Old 13th May 2012, 22:45
  #70 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,123
Howdy.

I considered between, on the outer frame of the console R/L. That makes for a busy console, but also makes it easy to obtain the controls of the side that is working, if one is inop, also to cage one set, if that pilot is I/C. I think as a whole, in between is a keeper.

Throttle could be incorporated into the Stick Grip, as radial movement is unadvantaged in the current design, ala helo. I honestly cannot come up with sufficient positives to go with fuselage side placement, Honest.

Console Sticking is completely visible, can be acquired by either pilot, and the only downside is throttle location, if single pilot for any reason. "Is there a pilot on board, Miss?" In this case, Console Sticks are an improvement even on the Yoke.

I am a bit cynical, so bear with me in that I think the closer was Marketing. Very different in concept, to go with the FBW design. Change for Change' sake, here, yes, I think so.

Last edited by Lyman; 13th May 2012 at 22:47.
Lyman is offline  
Old 13th May 2012, 23:25
  #71 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
Scientific approach

I am just trying to understand why AI decided for this "two separated no feedback sticks". A decision with important implications.

I am trying to understand the Rationale used. And who gained?


FBW and protected plane (important change) was not enough? As a marketing tool for the new competitor?
RR_NDB is offline  
Old 14th May 2012, 00:01
  #72 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,123
RR_NDB

i am trying to be objective, and generous. The SideStick is evocative of "A breed apart", and with "One Type Rating" family style, it certainly fits the bill. I am sad to say it is this concept (a breed apart), that sealed the design with sticks as far apart as possible.... To retain the Yoke retains the "look", and looks are important, so yes, it was a marketing tool.

Keep in mind, for better or worse, we see here in evidence all manner of "Defense of design". There is little if any of the advantages, not so apparent, are they?

Of the Bus Pilots I have interviewed, they speak highly of the airplanes' FBW. On the Stick I see mostly a shrug. It is as different, within reasonable limits, as it could be, the Stick thing.... Maybe to remind the pilot what he is driving?

Personally? I love the simplicity of the guts of the Stick arrangement, its lightweight (total add to gross) and its feel. It's solid. Placement? Not so much....

You ask the rationale, and I have to guess only. Gain? Also guessing, initial cost, (low), ongoing maintenance, (also easier and also low cost), and feather weight......

also 'Mac' (bear)
Lyman is offline  
Old 14th May 2012, 03:12
  #73 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Originally Posted by RR_NDB
I am just trying to understand why AI decided for this "two separated no feedback sticks". A decision with important implications.
That was also the main reason behind my question here

What are the main advantages? And it's disadvantages?
A : simplicity – overall dimensions - weight
D : suppresses first class information to a PNF

A similar device would be eventually adopted by Airbus SAS competitor(s)?
The Bombardier CSeries is going for the sidesticks but I have no idea if they will be coupled or not ?

CONF iture is offline  
Old 14th May 2012, 04:50
  #74 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
Why SS?

I missed your question in the quick scan i made when returned from a long trip. Indeed is an interesting point.
D : suppresses first class information to a PNF
Why not a coupling (FBW means) between sticks. This could also be retrofitted quite "easily". (Not interferes with other "features")


Cost seems not to be relevant. The sensors (position) on Stick or Yoke are similar. The mechanics of the Yoke, IMHO not too much more compared to Sticks. I really suspect the decision was marketing oriented. And in a serious survey (among pilots) would not be implemented, technically speaking. Something probably happened biasing to a "novelty". We will learn more on that.
RR_NDB is offline  
Old 14th May 2012, 23:56
  #75 (permalink)  
 
Join Date: Dec 2002
Location: UK
Posts: 1,892
Mac, side-sticks; consider a ‘design’ view.
FBW – it’s state of the art; many technical, weight and cost reasons for using it; this has nothing (directly) to do with the choice of SS.

Certification regulations have evolved to require dual independent (separable) control systems. Mechanically these are relatively heavy, complex (mechanical components, reliability, and servicing), and space demanding = higher cost.

With conventional columns in normal operation, only one is used at a time.
In failure conditions – split systems, again only one column should be used at a time. This assumes that one system has jammed; but if not, then dual (opposing) inputs could be hazardous, and such situations requires crew knowledge and procedure to avoid a hazardous condition.
If dual inputs are made with conventional columns in normal operation, then the larger force might dominate (CRM arguments); some (the majority?) of control systems will auto-split, the force disparity being interpreted as a jammed system.

The choice of SS provides a simpler and lighter installation, which should reduce cost and increase reliability - not necessarily safer as this is a different argument. It also avoids the need for co-ordination during failure as an electronic voting and/or override facility can be provided. This should not be seen as something which is unique or of poor design, but an aspect which is a different implementation of existing requirements, and thus requires different understanding and training in aircraft operations – again some CRM views may have misunderstood this.

SS enables space for a ‘management’ table – modern aircraft operating concept.
Either SS or centre sticks (FBW) enable HOTAS, but note the modern industry’s reliance on autothrust systems.
A SS does not enable changing hands during manual flight; but note the recent dominance of autopilot operations - and why should you write and hand-fly at the same time.

Re control position/feedback. For column designs in normal operation, physical feedback is not required (advantages for hands-on during training, but note the increased use of simulators).
Columns enable some visual feedback of hand-wheel position (roll), less so for pitch, but this may be judged as a minor contribution to awareness which should focus on the outcome – what is the aircraft doing.
In abnormal situations, controls split or not, perhaps the rarity of events and inability to define the human contribution in such operations results in a design biased towards the older regulations.
Again, irrespective of the control input, it’s the output which is important: – consider a situation with a jammed feedback-enabled SS, its essentially the same as a non-feedback SS as there is no meaningful information. This point may be arguable w.r.t. recent (post Airbus SS) requirements for the avoidance of error (CS 25.1302), but in the absence of guidance as to what errors will be encountered (foresight), the non-feedback design appears satisfactory from a design view (which would include pilots’ views).

Thus a designer might well choose SS; the advantages outweighing the disadvantages. The initial certification costs might be high – proof of concept, but thereafter identical systems can be used in all future types. SS flight operations require changes in training, but these would consistent with a new type, and marketing will have a new ‘unique’ feature for all future types.
Like most aspects of aviation these choices involve human judgement, which is often subject to bias (including culture), and even where an apparent irrational choice has been taken, it was taken because it was right ‘at that time’.

Something vaguely associated with this – between theory W and theory Z ‘Fostering successes rather than reducing failures.

Last edited by safetypee; 15th May 2012 at 00:04.
safetypee is offline  
Old 15th May 2012, 04:14
  #76 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Airbus had a new concept in mind and thought they would proceed, which they did, whatever the concerns raised by the pilots :
Originally Posted by aguadalte
I had the opportunity to meet Pierre Baud, when I was invited to Toulouse in the early ninety's, to fly one of their A330 testbed aircraft. That was my first experience with a FBW aircraft (before that, only Boeing and A310's) and I personally had the chance to verbalize my worries in this regard (no feed-back on SS, added inputs on SS, lack of need for trimming, ATS in step of Auto-Throttles)
For certain aspects they steped back as for the thrust management system which was initially designed to be operated through pushbuttons only, no thrust lever at all. This was mentioned a few years back by Chris Scott who was one in the early operation of the 320.

But for the independant sidesticks concept, they were not negociable : The pilots would adapt and they would love it, which is not especially untrue.
As PJ2 put it : "We get used to it"

Still, I don't understand how they thought the visual information provided to the PNF by a yoke was of no value anymore and therefore not necessary ?
Every rotation phase, flare phase, as a PNF, I am very consious for being deprived of such nice information, just feeling a bit out of the loop. I have well a few anecdotes to make my point, even if to this day they are only anecdotes ... but I am fully aware that further on the road I could be well more seriously reminded how such information was of interest ...

Has one of the reasons behind such thinking by Airbus just been mentioned by Lyman :
A possibility of course, exists that single pilot operation, contingent on the building of an extensive record of safety thus equipped, might be "just around the corner"?
I think it is a possibility.
As I said earlier, a sidestick for a single crew operation makes a lot of sense, but makes a lot less, IMO, for a multicrew operation.
CONF iture is offline  
Old 15th May 2012, 14:03
  #77 (permalink)  
 
Join Date: Mar 2006
Location: England
Posts: 848
The book linked here has some relevance to the thread: http://www.pprune.org/safety-crm-qa-...ml#post7185320
I’m only on Chapter 2 so far, but the introduction is invigorating and inspires thoughts of improved safety.
PEI_3721 is offline  
Old 16th May 2012, 22:55
  #78 (permalink)  
Thread Starter
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
Studying

I am studying and thinking on the excellent material presented by safetypee, PEI_3721 and others, before posting.

Exciting issue. Will comment ASAP.
RR_NDB is offline  
Old 18th May 2012, 20:21
  #79 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 756
Danger Hard limits = single pilot ; side stick = women pilots with skirts !

Originally Posted by roulishollandais
Quote:
"Originally Posted by bubbers44
"Pilots need to be in control of their aircraft at all times"

It would be necessary that the pilots have control before to start the system, and may reverse initial conditions :

In the (relative) "old" time the pilot opened brackets for the PA/system. The pilot was "MASTER" in the effective aircraft , and the PA/system was "SLAVE" .

In the bad designed today's effective aricraft, the system puts pilots between brackets before he is sitting in the cockpit . The system is "MASTER" in the effective aircraft , and the pilot is "SLAVE" . ECAM throws him biscuits to keep him busy .

In the 8 billions Ariane 501 crash (Arianespace 4.june1996) the initial condition had been added, to compute the ground position of the rocket. At take-off the bug needed only 37 seconds to show it was the master of the rocket.
The perfect inquiry conducted by Jacques-Louis LIONS finished in less than 24 hours, written in 15 days, published in 6 weeks, showed that in computer systems we must not trust the system if we see no failures, but we must take it for faulty so long we have not been able to Proof it is safe, with all the best knowlege at this time
When a Flight control system has hard limits,which is the case for Airbus and not for Boeing, that means that the SYSTEM is MASTER and the PILOT(S) are SLAVE . As soon as you have ONE MASTER SYSTEM the number of pilots is only regulation and look dependant !

As the first woman pilot came in Air France (Danièle Decuré 1974), Ziegler's machism prefered the pilots are slave . And to avoid they wear pants, he choose the side stick ..... Danièle Decuré was wearing a skirt...
roulishollandais is offline  
Old 19th May 2012, 02:20
  #80 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,123
Out here in the west we call that "side-saddle". Only in this case, the cowboy is heading sideways.
Lyman is offline  

Thread Tools
Search this Thread

Contact Us Archive Advertising Cookie Policy Privacy Statement Terms of Service

Copyright © 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.