Man-machine interface and anomalies
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 82
Posts: 1,454
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Dozy
From what I was told, you've got it 180 degrees the wrong way. The original detailed specifications were drawn up by the pilot and aeronautical engineers. The systems engineers simply implemented those specifications.
Pilots are generally a conservative bunch. It promotes longevity.
Join Date: Jul 2002
Location: UK
Posts: 3,093
Likes: 0
Received 0 Likes
on
0 Posts
In my admittedly limited experience (after all, I don't spend all day every day around pilots), the only generalisation that can be made about them as a breed is that they tend to be of above average intelligence and as individuals are what you would call "driven". Other than that it's a crapshoot - I've encountered (and heard second-hand information of and from) pilots who are technophobes, technophiles, liberal, conservative, religious, agnostic - the list goes on and on.
The Airbus FBW system and flight deck layout is not that much of a jump - and actually a fairly logical iterative step, based on historical precedent. The sidestick (or "minimanche" en francais) was initially tested on a Concorde testbed by the French aerospace agency - the idea was to bring the advances used in the space programme and military aviation to the civil arena - it certainly wasn't particularly controversial initially amongst the pilots who were involved. Once the decision is made to go all-hydraulic, a lot of the necessity for interlinked controls with significant physical leverage goes away. As soon as you have proven technology that can keep the aircraft inside its flight envelope, it makes sense to use it. Remember that the head of the pilot engineer group at Airbus was previously D.P Davies' successor at the ARB, and to those who knew him he was pretty much the dictionary definition of a "pilot's pilot". He had around three decades' worth of experience evaluating aircraft designs in terms of safety, and - from what I gather - in spite of being genial, whip-smart and collegial in his approach, not the sort of person who would buckle to demands from outside for the sake of political comity.
He was known for challenging (in the spirit of friendly competition) pilots who had reservations about the design to simulator exercises in which the A320 was pitted against conventional designs, and in every case the simulated A320 outperformed the challenger in terms of safety margins. His untimely death* was a tragedy in every sense, because he above all others was qualified to explain the design to pilots in terms they'd find natural, and his gregarious nature (not to mention his reputational clout) would likely have won more pilots over in the early years had he lived to do so.
The design brief has been misconstrued horribly based on rumour over the years, but at it's heart was this:
This wasn't something that was given to the engineering bods to build and thence to the pilot engineers to tweak - it was a collaborative process from beginning to end.
Of course, once this had gone through the sausage machine of the mainstream press, what was reported was that Airbus were designing a system that would reduce training costs in a radical way, with an unprecedented level of computer technology which would be the safest airliner in the world. At the time FBW was little understood outside of military aviation circles, and the best-known computer presence on other designs was the FMS (incorporating autopilot), so unfortunately the meme that the new Airbus models would be automated to an unprecedented extent (and that this was the reason for the cost savings - at the expense of pilots) took hold, and even more unfortunately has never entirely gone away. The move from large, central control yokes to sidesticks seems to have been interpreted in some quarters as a symbolic indication of diminished pilot authority, rather than the more prosaic attempt to improve ergonomics and ease type conversion that it actually was.
To address RR_NDB's questions, ECAM (and Boeing's response - EICAS), actually presented systems messages in a more concise and relevant way than anything that had come before it, and I'm sure that they've been incrementally improved with each new type that was made (by which I mean that the ECAM system in the A380 is probably a lot more refined than that fitted to the A320). The problem is that UAS specifically is a notoriously difficult failure mode to diagnose without some form of polling over time. Prior to the introduction of these systems, diagnosis was entirely reliant on crew communication - one pilot spotting an incongruous airspeed reading and cross-checking it with his opposite number.
AF447 is an interesting case for several reasons in this regard, because triple pitot tube failure was unheard of until the introduction of the Thales AA pitot tube (which was an optional fit), several years after the A330 and A340 had entered service, and the accident sequence was initiated not because of the UAS situation itself, but because for reasons which will likely be hotly debated for years to come, the designated relief pilot began a series of pitch-up commands before properly assessing the situation. Over 30 other incidents related to the same failure mode had a successful outcome, so the premise that the indications are somehow deficient is at best highly debatable.
[* - due to complications from altitude sickness when he travelled to the Himalayas as part of an accident investigation team - characteristically he requested to lead the technical team in the face of a hazardous assignment despite his seniority.]
The Airbus FBW system and flight deck layout is not that much of a jump - and actually a fairly logical iterative step, based on historical precedent. The sidestick (or "minimanche" en francais) was initially tested on a Concorde testbed by the French aerospace agency - the idea was to bring the advances used in the space programme and military aviation to the civil arena - it certainly wasn't particularly controversial initially amongst the pilots who were involved. Once the decision is made to go all-hydraulic, a lot of the necessity for interlinked controls with significant physical leverage goes away. As soon as you have proven technology that can keep the aircraft inside its flight envelope, it makes sense to use it. Remember that the head of the pilot engineer group at Airbus was previously D.P Davies' successor at the ARB, and to those who knew him he was pretty much the dictionary definition of a "pilot's pilot". He had around three decades' worth of experience evaluating aircraft designs in terms of safety, and - from what I gather - in spite of being genial, whip-smart and collegial in his approach, not the sort of person who would buckle to demands from outside for the sake of political comity.
He was known for challenging (in the spirit of friendly competition) pilots who had reservations about the design to simulator exercises in which the A320 was pitted against conventional designs, and in every case the simulated A320 outperformed the challenger in terms of safety margins. His untimely death* was a tragedy in every sense, because he above all others was qualified to explain the design to pilots in terms they'd find natural, and his gregarious nature (not to mention his reputational clout) would likely have won more pilots over in the early years had he lived to do so.
The design brief has been misconstrued horribly based on rumour over the years, but at it's heart was this:
- To develop a flight deck layout that could be used across a wide range of airliner types, yet retain a consistent feel in order to minimise conversion training costs
- To develop a flight deck environment that would be ergonomically best-of-breed, applying modern (but proven) technology where necessary
- To maximise systems safety by keeping mechanical and electronic complexity to a minimum
This wasn't something that was given to the engineering bods to build and thence to the pilot engineers to tweak - it was a collaborative process from beginning to end.
Of course, once this had gone through the sausage machine of the mainstream press, what was reported was that Airbus were designing a system that would reduce training costs in a radical way, with an unprecedented level of computer technology which would be the safest airliner in the world. At the time FBW was little understood outside of military aviation circles, and the best-known computer presence on other designs was the FMS (incorporating autopilot), so unfortunately the meme that the new Airbus models would be automated to an unprecedented extent (and that this was the reason for the cost savings - at the expense of pilots) took hold, and even more unfortunately has never entirely gone away. The move from large, central control yokes to sidesticks seems to have been interpreted in some quarters as a symbolic indication of diminished pilot authority, rather than the more prosaic attempt to improve ergonomics and ease type conversion that it actually was.
To address RR_NDB's questions, ECAM (and Boeing's response - EICAS), actually presented systems messages in a more concise and relevant way than anything that had come before it, and I'm sure that they've been incrementally improved with each new type that was made (by which I mean that the ECAM system in the A380 is probably a lot more refined than that fitted to the A320). The problem is that UAS specifically is a notoriously difficult failure mode to diagnose without some form of polling over time. Prior to the introduction of these systems, diagnosis was entirely reliant on crew communication - one pilot spotting an incongruous airspeed reading and cross-checking it with his opposite number.
AF447 is an interesting case for several reasons in this regard, because triple pitot tube failure was unheard of until the introduction of the Thales AA pitot tube (which was an optional fit), several years after the A330 and A340 had entered service, and the accident sequence was initiated not because of the UAS situation itself, but because for reasons which will likely be hotly debated for years to come, the designated relief pilot began a series of pitch-up commands before properly assessing the situation. Over 30 other incidents related to the same failure mode had a successful outcome, so the premise that the indications are somehow deficient is at best highly debatable.
[* - due to complications from altitude sickness when he travelled to the Himalayas as part of an accident investigation team - characteristically he requested to lead the technical team in the face of a hazardous assignment despite his seniority.]
Last edited by DozyWannabe; 4th Nov 2012 at 05:26.
Thread Starter
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 875
Likes: 0
Received 0 Likes
on
0 Posts
AF447: Why designated relief pilot began a series of pitch-up commands?
DozyWannabe
I am traveling but will comment now on some points of your good post.
After all the effort we made in trying to understand this case my opinion is:
RR_NDB
I am traveling but will comment now on some points of your good post.
After all the effort we made in trying to understand this case my opinion is:
- UAS events are important and crew must and can know it immediately before any processing of it by the System. Why? Just because pilots don't like complex surprises.
- Complex Systems interact with pilots through Interfaces: Man-machine interfaces. The one used in F-GZCP certainly played an important role to the Human Factors issue.
- Above a certain threshold (of problems) faced by the A/C the current Interface can present important difficulties agravatting Situational Awareness issues. It's characteristics could delay a fast comprehension of "whats going on" something vital in certain situations.
- In AF447 case the "series of pitch-up commands" could not be happened if PF and PM received proper inputs from A/C and System.
- 1 out of 32 previous cases are not low enough to regard AF447 a "black swan" ocurrence.
- The Interface is more important in a Design philosophy where automation is using "hard limits"
- Fault Tolerance and Graceful Degradation (a/c+crew) are important for survivability. The "treshold effect" i commented in an earlier post is extremely dangerous and may appear in situations not predictable by the "mangement/engineers/pilots" who design a complex A/C.
- To have a flight deck layout that could be used across a wide range of airliner types, yet retain a consistent feel in order to minimize conversion training costs
- To have a flight deck environment that would be ergonomically best-of-breed, applying modern (but proven) technology where necessary
- To maximize systems safety by keeping mechanical and electronic complexity to a minimum
RR_NDB
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes
on
0 Posts
Sorry to go all Luddite on us, but I cannot figure out why at least two axes cannot remain inertially selected post AP drop. While the boys get out the page, is it so much to ask of the platform to select 0/.80.? As to roll, why should the rate/select LAW change to DIRECT? Just for excitement?
Join Date: May 2012
Location: Earth
Posts: 88
Likes: 0
Received 0 Likes
on
0 Posts
Most of the stuff is crutches for people that need the automation and SA to lean on or having new toys to play with.
What kills me is that I saw a flight dept crumble after the pilot put the plane down for like 4 months to UPGRADE from an XLS to a UNS1....
or Airlines that could have paid the pilots, but instead upgraded the whole fleet, then 4 years later the company is going banko..
or how about guys that pay millions for a single engine turbine to get the avionics, new leather....and still stick it in the water...when a twin for half the price, but older avionics, would have saved the day.
Don't even get me started on HUDs, TCAS II vs I....
What kills me is that I saw a flight dept crumble after the pilot put the plane down for like 4 months to UPGRADE from an XLS to a UNS1....
or Airlines that could have paid the pilots, but instead upgraded the whole fleet, then 4 years later the company is going banko..
or how about guys that pay millions for a single engine turbine to get the avionics, new leather....and still stick it in the water...when a twin for half the price, but older avionics, would have saved the day.
Don't even get me started on HUDs, TCAS II vs I....
