It seems the problem is the interface. If the problems could be shown on a (laptop?) screen one or two dozen at a time, with individual checkboxes and a check-all-checkboxes button meaning "I have read and acknowledged this problem/error/message" with of course the ability to go back and read through them again later, they could have gotten through with everything a lot faster. It seems the manufacturer would want to (and regulatory agencies would want them to) implement such a change to make it easier and faster to deal with such a major situation.
They'll fix it only after a black box recording will show a crash was caused because the pilots were too busy clearing warning messages to actually fly the damn plane.
Meanwhile those of us in human factors will shake our heads. Disclosure: I worked with Aerospatiale for a brief time in 1989, and nothing one of their products did since then, from Airbuses to Arianes, has surprised me in the least. They have a touching faith in the ability of software to correct every ill, touching as in "touched in the head", and I speak as someone with decades in the software biz.
It's like Tesla marketing their driving assist as "Autopilot", which leads to people thinking the car will drive itself under any circumstance. Several crashes later... Tesla would've been far better off naming it "Copilot", a better description of the thing's capabilities.
Meanwhile those of us in human factors will shake our heads. Disclosure: I worked with Aerospatiale for a brief time in 1989, and nothing one of their products did since then, from Airbuses to Arianes, has surprised me in the least. They have a touching faith in the ability of software to correct every ill, touching as in "touched in the head", and I speak as someone with decades in the software biz.
It's like Tesla marketing their driving assist as "Autopilot", which leads to people thinking the car will drive itself under any circumstance. Several crashes later... Tesla would've been far better off naming it "Copilot", a better description of the thing's capabilities.
My conclusion on that A380 is that at least a CNC probing device should be used to check the last step between the un-machined pipe and the machined steps all the way around the pipe.
Better still get a filter made that fits without all that machining.
That engine part sawed through al awful lot or cables making the number of error messages blindingly obvious.
Better still get a filter made that fits without all that machining.
That engine part sawed through al awful lot or cables making the number of error messages blindingly obvious.
Those glass screens, they're called MFDs, are very specific. They are not computers in the sense of what Airbus puts in front of the pilots. The MFDs are operated with embedded software, with single board computers and real time operating systems - under control of the Mission Computers. Hardly a "laptop computer".
In a Boeing, the MFDs warnings do not flow to every display, nor across multiple MFDs, they are displayed just when specific to the systems specific to that display and do not interfere with the workload of the pilot: The flow of data and control is extremely constrained to let the pilots concentrate at their task without overriding them.
The input menus are very simple and complex interface stuff is relegated to the data input keyboard (and display) on the center pedestal.
Something like the C17 is actually a lot more complex than a 380/747/787/etc because it has a bunch of extra subsystems you don't need in a commercial bus. And guess what, it has a stick.. yep... the C17 has a stick. I learned how to fly it in the simulators. The best thing was taking off, very lightly loaded, at FULL power from Long Beach, up to 5000 feet and then doing a sharp turn West and diving on the 405 and flying it up past LAX ( no LAX traffic programmed into my dive bomb scenarios ).. then UP, UP, UP, over Mulholland and doing a near tactical landing on Edwards.
No Airbus jockey has any idea about such things. Dive bomb The Police... OMG, how antisocial. Quick, call Le Inspecteur!
My buddy actually got certified for twin engine aircraft with just the bare minimum of actual flying hours with an instructor. We both had "flown" the C17. A Cessna is nothing compared with an AF Heavy doing an airdrop from 10,000 feet.. or flying in formation.
I didn't bother to get my pilot certification. I preferred to dive bomb the CHP on the 405 at El Segundo Blvd.... No instructor would let me fly a plane with my penchant to hit the airbrakes and flaps and hold it at 100 feet over the deck....
Besides, the Airbus computers would freak out.Boring airplanes, actually.
Last edited:
It's a computer but hardly a computer? Sorry does not compute. But I think you are trying to say that Boeing is superior because a vital human computer interface is off to the side and can only be operated with one hand where as Airbus have the option of a keyboard in front of you in the place most people are used to typing on a keyboard?
And thank you for being captain obvious and stating that military aircraft have some different things on them. I would never ever have guessed that
BTW last time I checked Airlines would still employ retired military pilots so I am sure there are plenty of 'airbus jockeys' who have a really good idea about these things.
Oh and as a passenger I like boring airplanes. When I want an exciting ride I'll get in a shorts skyvan for take off. Don't fancy landing in one though...
And thank you for being captain obvious and stating that military aircraft have some different things on them. I would never ever have guessed that
BTW last time I checked Airlines would still employ retired military pilots so I am sure there are plenty of 'airbus jockeys' who have a really good idea about these things.
Oh and as a passenger I like boring airplanes. When I want an exciting ride I'll get in a shorts skyvan for take off. Don't fancy landing in one though...
^ Nope, you are not getting my point... and indeed trying to ridicule me by being cute ( a logically invalid way to try to argue a point )....
An embedded computer driving firmware is a much simpler device than a general purpose computer driving a bunch of programs that allow dynamic "upgrades" and further "installations".
Let me give you an example.... take the Honda built in NAVI vs an Android Phone connected via an Android Auto Interface.
The Honda built in NAVI is very limited, it's is simple, giving you enough a well defined set of interface(s) and data items and it takes a MINIMUM amount of time, effort and distraction to read while you're driving. It also can not be changed -unless you get a new image at the dealer- so it's predictable. It's is very deterministic, reliable and robust as well.
The Android Auto interface can be a huge distraction. It supports a lot more applications, it's constantly wanting to update itself, it's more complicated and ultimately, although it gives you a lot more information, it is distracting and it takes a lot more effort to use -while you're driving.
Boeing approach to cockpit displays is like the Honda built in NAVI. Simple. Deterministic. Just the data you need, no more. Very easy to use.
Airbus is in between Android Auto and the built in. It gives you a lot more information... indeed more information that you need. It could be said it give you a lot of spurious data that ultimately distracts from the operation of the aircraft. It's more complicated to use.
Speaking of distraction.... Since Boeing expects the pilots to actively FLY the plane, they want a simple system. Airbus, OTOH, expects the computer to fly the plane and the pilots to shut up and play with the joystick... so the pilots can be distracted without affecting the operation of the aircraft.
My comment about Airbus jockeys... simple... don't be cute... in an Airbus the computer flies the plane... so even Evel Knievel would be prevented from doing any stunt. This goes for the best Luftwaffe ex-jockeys as well. The last time an Airbus crew tried to fly the computers they all ended up at the bottom of Jamaica Bay after the empennage of their plane broke off on take off from JFK.
My comment about cockpit complexity... the C17 uses the same methodology towards Human Interfaces as the Commercial planes do... Keep It Simple. That's why it can support so many complex to use subsystems that would overwhelm the much more complex world view of the Airbus. So, you missed my point in there... ( trying to be cute with ridicule just won't fly here... ). It's a somewhat more complex cockpit than the 747, 787... etc... yet it supports a lot more stuff without confusing the pilots.
I'd rather have my pilots fly the plane than a computer. And for that, you need a simple to use cockpit. That's all.
Oh you would have loved dive bombing the 405.... and being in the jump seat of the cockpit while we were refueling in mid air.... and... hey... I got paid for doing that. ;-) In reality, passengers should fly with a four point harness. That silly seat belt is just to keep you alive long enough to kiss your a$$$$ goodbye.
The only fly in the ointment was that the headsets we were issued were made by Bose. Which are actually quite good for their aviation purposes but the stereo soundstage is nowhere... they're.... MONO.
An embedded computer driving firmware is a much simpler device than a general purpose computer driving a bunch of programs that allow dynamic "upgrades" and further "installations".
Let me give you an example.... take the Honda built in NAVI vs an Android Phone connected via an Android Auto Interface.
The Honda built in NAVI is very limited, it's is simple, giving you enough a well defined set of interface(s) and data items and it takes a MINIMUM amount of time, effort and distraction to read while you're driving. It also can not be changed -unless you get a new image at the dealer- so it's predictable. It's is very deterministic, reliable and robust as well.
The Android Auto interface can be a huge distraction. It supports a lot more applications, it's constantly wanting to update itself, it's more complicated and ultimately, although it gives you a lot more information, it is distracting and it takes a lot more effort to use -while you're driving.
Boeing approach to cockpit displays is like the Honda built in NAVI. Simple. Deterministic. Just the data you need, no more. Very easy to use.
Airbus is in between Android Auto and the built in. It gives you a lot more information... indeed more information that you need. It could be said it give you a lot of spurious data that ultimately distracts from the operation of the aircraft. It's more complicated to use.
Speaking of distraction.... Since Boeing expects the pilots to actively FLY the plane, they want a simple system. Airbus, OTOH, expects the computer to fly the plane and the pilots to shut up and play with the joystick... so the pilots can be distracted without affecting the operation of the aircraft.
My comment about Airbus jockeys... simple... don't be cute... in an Airbus the computer flies the plane... so even Evel Knievel would be prevented from doing any stunt. This goes for the best Luftwaffe ex-jockeys as well. The last time an Airbus crew tried to fly the computers they all ended up at the bottom of Jamaica Bay after the empennage of their plane broke off on take off from JFK.
My comment about cockpit complexity... the C17 uses the same methodology towards Human Interfaces as the Commercial planes do... Keep It Simple. That's why it can support so many complex to use subsystems that would overwhelm the much more complex world view of the Airbus. So, you missed my point in there... ( trying to be cute with ridicule just won't fly here... ). It's a somewhat more complex cockpit than the 747, 787... etc... yet it supports a lot more stuff without confusing the pilots.
I'd rather have my pilots fly the plane than a computer. And for that, you need a simple to use cockpit. That's all.
Oh you would have loved dive bombing the 405.... and being in the jump seat of the cockpit while we were refueling in mid air.... and... hey... I got paid for doing that. ;-) In reality, passengers should fly with a four point harness. That silly seat belt is just to keep you alive long enough to kiss your a$$$$ goodbye.
The only fly in the ointment was that the headsets we were issued were made by Bose. Which are actually quite good for their aviation purposes but the stereo soundstage is nowhere... they're.... MONO.
Last edited:
That's 600 words failing to make a coherent point. I really do not understand what you are trying to say here. I also believe you are getting a lot of things wrong. Military aircraft in general take as much of the flying off the pilot as possible so the pilot can concentrate on the mission. Airbus follow that strategy . If you believe the father of a friend of mine from school days then boeing lost the plot with the 757 which he described as 'push button eat banana'. But concorde pilots were an interesting bunch.
But on your 'KISS' principle this is all well and good but a disc burst that puts a number of holes in the aircraft and severs a good chunk of the wiring harness on one side is NOT a simple failure. What would the dreamliner have told the pilots for that same sort of failure? Some examples would really help make some sense of your posts.
But on your 'KISS' principle this is all well and good but a disc burst that puts a number of holes in the aircraft and severs a good chunk of the wiring harness on one side is NOT a simple failure. What would the dreamliner have told the pilots for that same sort of failure? Some examples would really help make some sense of your posts.
If Boeing knows better, and still produced defective airplanes, then they are criminals, not just incompetent.
I don't want to pass judgement.
But in the West, including Europe, pushing out products with inferior manufacturing just to make a buck ( and a quarterly bonus ) is not considered criminal.
If we held stock brokers, salesmen, executives, bureaucrats and politicians criminally liable for their business and regulatory decisions, we'd be a better place.
It hurts when you design a good product and then some bean counter or regulatory jerk or politician decides to short change it.
But in the West, including Europe, pushing out products with inferior manufacturing just to make a buck ( and a quarterly bonus ) is not considered criminal.
If we held stock brokers, salesmen, executives, bureaucrats and politicians criminally liable for their business and regulatory decisions, we'd be a better place.
It hurts when you design a good product and then some bean counter or regulatory jerk or politician decides to short change it.
Negligence is criminal. Plenty of people are convicted of negligence.
neg·li·gence / noun failure to take proper care in doing something. "some of these accidents are due to negligence"
Criminal negligence refers to conduct in which a person ignores a known or obvious risk, or disregards the life and safety of others. Federal and state courts describe this behavior as a form of recklessness, where the person acts significantly differently than an ordinary person under similar circumstances.
Criminal negligence is when a person acts with a disregard for obvious risks to human life and safety. The State must prove that a defendant acted with criminal negligence to convict that person of certain offenses.
In these cases, a prosecutor must show the following to prove that an accused acted with criminal negligence:
neg·li·gence / noun failure to take proper care in doing something. "some of these accidents are due to negligence"
Criminal negligence refers to conduct in which a person ignores a known or obvious risk, or disregards the life and safety of others. Federal and state courts describe this behavior as a form of recklessness, where the person acts significantly differently than an ordinary person under similar circumstances.
Criminal negligence is when a person acts with a disregard for obvious risks to human life and safety. The State must prove that a defendant acted with criminal negligence to convict that person of certain offenses.
In these cases, a prosecutor must show the following to prove that an accused acted with criminal negligence:
- he/she acted recklessly and created a high risk of death or great bodily injury, and
- a reasonable person would have known that these acts would create such a risk.1
- knowledge of a danger, and
- more than a mistake or excusable accident.
If that's the case, why weren't people in the EPA found criminally liable for their decisions that led to all that cadmium that got released on the Colorado River a few years ago?
How about the political bias and Yellow Journalism in the Mainstream "Media" Machine?
The Wall Street people in those firms pushing for regulations that benefit themselves at the expense of the pensions and jobs of ordinary working Americans?
Etc, etc...
In our system, many people, properly politically connected, do not take responsibility for their actions.
Do I agree with that? He!!! no. If it were up to me, I'd have many people hanging from light poles imported from Milan.
Heck, the City of London and most of the capitals in Europe would be filled with "previously important people" dangling from tall lights posts imported from Milan.
And the World would be a better place for it.
How about the political bias and Yellow Journalism in the Mainstream "Media" Machine?
The Wall Street people in those firms pushing for regulations that benefit themselves at the expense of the pensions and jobs of ordinary working Americans?
Etc, etc...
In our system, many people, properly politically connected, do not take responsibility for their actions.
Do I agree with that? He!!! no. If it were up to me, I'd have many people hanging from light poles imported from Milan.
Heck, the City of London and most of the capitals in Europe would be filled with "previously important people" dangling from tall lights posts imported from Milan.
And the World would be a better place for it.
I believe you and that's really disappointing. I only know an average amount of human factors, but this event surely indicates a change is needed, even if such an extraordinary event never happens again. It's bad enough Boeing and the FAA were agreeing to let Boeing self-inspect (or not) and whatever else went on.
For this purpose (and assuming the original code is well-written, which we never really know but that's one purpose in having a simpler device) I would hope it doesn't allow dynamic "upgrades" but most if not all the embedded processors and stuff I look at thesedays assume there's going to be some at least part-time upstream connection and the code will be able to phone home and if a new version is available do self-updating. Many have encryption and thus can talk directly over the Internet, others don't but can easily get a code update from another processor/controller that does talk to the Internet and (allegedly) does have encryption.
But it's easy to look at MCAS that only had cursory testing and wonder how many other systems like that have only cursory testing.
There's perhaps some legal definition of the words "high risk" (there are so many "ordinary" words that in law, engineering and other areas have specific and probably different meaning than in ordinary English), but I can only wonder exactly what that means.
There are so many things that cause real harm, someone clearly knew about it and didn't fix it, but "for whatever reason" it doesn't get prosecuted. The DOJ is only now looking at criminally charging Boeing because promises were made to the FAA and such that certain things would change since the two MAX crashes, but they didn't change, and it's only after these latest incidents that this was looked into.
Here's a video news thing from a couple days ago:
Ever done DO-178A/B/C ?
It's NOT allowed.
The firmware needs to go through extensive testing and validation before it's ever allowed on the aircraft. It must satisfy the requirements and you can not do this in a field installation.
Also, the code is measured to verify it's well written by many metrics -static analysis being one of them.
Same thing with equipment for Telcos (internetworking and voice), Medical devices, metrology, ham radio, etc, etc... or for any type of device that must go through qualification.
Would you want your oscilloscope to update itself?
Only in consumer products do they allow that.. like my LG TV that updated its firmware and now it won't connect automatically with the Roku....
Any device that is designed with any seriousness and with attention to robustness, reliability, correctness and deterministic behavior ( ie: real time embedded stuff ) will NOT allow such self-updating.
The only time I recall such a device updating was one of JPL's NASA Mars units that somehow had a defect that escaped testing (how? I can not believe they were so sloppy). It's firmware was updated, manually... the risk being that they risked losing the lander.
OK, another time, Lucent decided to field update the ATT5ESS switches in the Chicago Metropolitan Loop. They lost it... and the Chicago Board of Trace and banks lost their telco connections for a week!
You just don't update such things
It's just not something you'd give to some outsourced hackers who thing Python is a language.
BTW- the avionic hardware in an airliner must also satisfy DO-254 (software is handled by DO-178). I'm sure the mechanical construction of the aircraft must go through similar standards. So, I'm really disappointed at the lack of manufacturing standards of quality at Boeing.
+++
MCAS: No the problem was not that it wasn't fully tested and certified... it was that IT WASN'T DISCLOSED to the pilots. This change of behavior was part of the design and it required re-certification of the pilots, but Sales and Marketing decided to keep the cost down for the airlines..... Boom!
"...As Boeing and the FAA advanced the 737 MAX toward production, they limited the scrutiny and testing of the MCAS design. Then they agreed not to inform pilots about MCAS in manuals, even though Boeing’s safety analysis expected pilots to be the primary backstop in the event the system went haywire..."
Sure, blame the engineers... you know this decision came from above.... Sales and Marketing responding to customers to keep the price down...
https://www.seattletimes.com/seattl...-max-system-gained-power-and-lost-safeguards/
Last edited:
It's all BS.
The FAA was fully onboard with this.
If anything the FAA, its inspectors and regulators, are the primary actors at fault. Their job is to ensure that quality standards are kept up.
Just like a metrology device needs to have greater accuracy than what it's used to measure, the inspectors and regulators must ensure their standards are greater than those of their 'inspected'. The failed their tasks miserably and should all be held criminally liable FIRST and FOREMOST.
The Mainstream Propaganda Machine won't say this because they have become the mouthpiece of the Gov.
BTW... the reason why airbus never encounters such issues?
Again, from the Seattle Time article quoted above:
"In the midst of a wind-up turn, the software would automatically swivel up the leading edge of the plane’s entire horizontal tail, known as the horizontal stabilizer,"
"Philosophically, Boeing had long opposed efforts to create automated actions such as a stick-pusher .."
In an Airbus, the computer controls the plane and pilots are used to that... in a Boeing, pilots are used to being in control... hence, this new nose up behavior -the problem that MCAS fixes- should have been disclosed to them!
Again, from the Seattle Time article quoted above:
"In the midst of a wind-up turn, the software would automatically swivel up the leading edge of the plane’s entire horizontal tail, known as the horizontal stabilizer,"
"Philosophically, Boeing had long opposed efforts to create automated actions such as a stick-pusher .."
In an Airbus, the computer controls the plane and pilots are used to that... in a Boeing, pilots are used to being in control... hence, this new nose up behavior -the problem that MCAS fixes- should have been disclosed to them!
I can't see that in any of this, one methodology is right and another wrong. I see that, though all these planes are in some ways "evolutionary" in many they are "revolutionary". The work being expended to get the next 5% of fuel saving, the extra 3% of seats, the 7% lower downtime, strains all engineers as well as practical and theoretical envelopes.
To me, it all points to one thing, objective high powered oversight, be it in design stages, in manufacturing plants and at testing and approval stages is paramount, and that this information is fed back into the manufacturers processes.
That there seems to be little (maybe I'm wrong here) rigorous testing of software by external bodies is troubling. The profit motive is always a temptation for companies, in every industry or walk of life. Look at VW and the emissions scandal as a lesson for all. Doing it “on the cheap” is not an option.
That we are able to design and make these things in the first place is a testament to the engineers involved. Maybe, we have to get used to paying a bit more for our air tickets, and make sure that money gets passed to the approvals teams.
To me, it all points to one thing, objective high powered oversight, be it in design stages, in manufacturing plants and at testing and approval stages is paramount, and that this information is fed back into the manufacturers processes.
That there seems to be little (maybe I'm wrong here) rigorous testing of software by external bodies is troubling. The profit motive is always a temptation for companies, in every industry or walk of life. Look at VW and the emissions scandal as a lesson for all. Doing it “on the cheap” is not an option.
That we are able to design and make these things in the first place is a testament to the engineers involved. Maybe, we have to get used to paying a bit more for our air tickets, and make sure that money gets passed to the approvals teams.
A fantastic article - the details around the cause and ultimately how they arose back in design and manufacture eye opening to say the least. Clearly having an experienced crew at the time was the major factor the plan landed safely.
- Status
- Not open for further replies.