Boeing 737 Max and Airbus Controls Discussion

[pilotSteve]

This redesign scares me. Rather the new steering "wheel", not really bothered by the rest. I can't fathom what Tesla is thinking. There is a reason the steering wheel is mostly round. In fact, I thought it was required by law. The only way a yolk style steering wheel has any practicality is if it only is quarter turn center to lock. Much like an airplane. Very dangerous to have such fast steering on a non-race car. I absolutely hate it.

“steer by wire”, where ratio is variable based on speed. Also likely non-limear response to steering angle. Similar to a sidestick in modern aircraft vs old “control wheel.

Also steer by wire could have “envelope protection” where if driver mode is “standard” (not advanced) the vehicle would not “let” the driver get too wild (unsafe). Crazy? maybe, maybe not.....
Airbus vs Boeing. Each has advantages and fans/haters.

 

[jbcarioca]

“steer by wire”, where ratio is variable based on speed. Also likely non-limear response to steering angle. Similar to a sidestick in modern aircraft vs old “control wheel.

Also steer by wire could have “envelope protection” where if driver mode is “standard” (not advanced) the vehicle would not “let” the driver get too wild (unsafe). Crazy? maybe, maybe not.....
Airbus vs Boeing. Each has advantages and fans/haters.

Yes! I am solidly an Airbus person too. Also, don't forget that maintenance is lower and 'recalls' are fewer. Of course one must write ones own code. Third parties might not quite get there. VW id3 and B737MAX/B787 batteries have in common third party software. Desperate attempts to recycle 1960's technology really don't work well with airplanes or cars.

We already have Tesla proving over and over the merits of in-house software.

 

[jbcarioca]

Speaking as a pilot, I think Elon missed a great opportunity. The yoke was a half-measure. Ditch the yoke for a side-mounted stick.

That would really free up some real estate in front of the driver.

Do you remember all the furor about side stick controls. I recall Boeing people screaming Unsafe! Then FlyBY Wire- Unsafe! the pilot always knows best. That kind of luddite view held Boeing back decades.
A good deal of the Tesla vs legacy choices seem eerily reminiscent of B vs A.
Thus far we do seem to know it is a very rough transition for legacy, just as VW is experiencing with software.
Staying safe often means abandon 'tried and true' when 'tried' becomes 'tired'.

 

[Altair7]

Do you remember all the furor about side stick controls. I recall Boeing people screaming Unsafe! Then FlyBY Wire- Unsafe! the pilot always knows best. That kind of luddite view held Boeing back decades.
A good deal of the Tesla vs legacy choices seem eerily reminiscent of B vs A.

I don't think Boeings every did anything like this:

Designing out human decision making is an inherently bad idea.

 

[ucmndd]

I don't think Boeings every did anything like this:

They did, they just happened to have 346 people aboard 2 different jets when it happened (see: 737 MAX MCAS)

 

[Altair7]

They did, they just happened to have 346 people aboard 2 different jets when it happened (see: 737 MAX MCAS)

Interesting you mention that, since I fly 737's, including the Max.

It's just a little bit more complicated. But still, a few relevant points:

1) Automation overriding human input (i.e., the Software Engineers Know Best mentality);
2) Poorly trained and lightly experienced pilots; and
3) Faulty system design relying on only one channel (air data source).

Throw poor maintenance into the mix, the stage was set.

Essentially, there was a very deeply seeded booby trap, and it took a couple of real boobies to find it. Absolutely shocking that they were even in that area of the performance envelope in VMC conditions.

Come to think of it, Tesla's FSD has already handed out at least one Darwin Award:

Apple engineer killed in 2018 Tesla crash had complained about Autopilot

 

[Just a Reader]

Interesting you mention that, since I fly 737's, including the Max.

It's just a little bit more complicated. But still, a few relevant points:

1) Automation overriding human input (i.e., the Software Engineers Know Best mentality);
2) Poorly trained and lightly experienced pilots; and
3) Faulty system design relying on only one channel (air data source).

Throw poor maintenance into the mix, the stage was set.

Essentially, there was a very deeply seeded booby trap, and it took a couple of real boobies to find it. Absolutely shocking that they were even in that area of the performance envelope in VMC conditions.

Come to think of it, Tesla's FSD has already handed out at least one Darwin Award:

Apple engineer killed in 2018 Tesla crash had complained about Autopilot

Nice, I mean, don't hold back. I suppose that you were one of the pilots who were invited by Boeing to prove that it was solely operator error?
I suppose that Boeing knows their customers and Boeing was happy enough to take their Dollars. Their customers seem to be decently capable operators of other B 737s and it was Boeing's specific sales pitch that the Max basically needed no additional pilot training. If it had been only about training, then why didn't Boeing manage to convince anyone that some additional hours of training would do?
Boeing should feel lucky that it's from the US and that regulators don't want to leave Airbus as the only supplier of aircraft in that field.

 

[Altair7]

I suppose that you were one of the pilots who were invited by Boeing to prove that it was solely operator error?

Sorry, I am just a worker-bee. No VIP Invites for me. If I had that kind of clout I would be on the Roadster Forums, not the Model S Forums.

But I do read accident investigations, and this is as far off-topic as I care to go into such a tragedy with a non-pilot:

For if you were a pilot (and Germany has produced some of the worlds finest pilots dating back to Otto Lillienthal, so you would know this) there is still this thing called "airmanship". When the the automation is doing something un-commanded, the well-trained pilot turns it all off and flies the airplane - by hand. Using airmanship and principles of aerodynamics.

Such as "in a stabilzer trim runaway at 500 knots and the thrust levers still in the take-off position, there is too much air-loading on the horizontal stabilizer to control - reduce thrust and speed immediately". And maybe - just maybe - in clear weather looking outside when the instruments don't make sense?

BTW, 500 knots IAS (indicated airspeed) is 160 knots over the red-line (maximum certificated) speed of a 737. Since a 737 does not accelerate as quickly as a Model S, that took some time to reach that state of affairs.

And that is exactly the state of things when Ethiopian contacted the ground and made a 50 foot crater into the ground (that would be 16 meters auf Metric).

Runaway Stabilizer Trim used to be a Memory Item for most of the worlds airlines. And the Very Second thing on that Memory Item List can be summarized for the layman as "Turn off the Automation".

Thanks to the hubris of "software engineers know better than pilots" it is once again a required Memory Item - for most carriers.

And to as your ill-advised comment that the operators being "decently capable" - the first officer of the Ethiopian plane had just 200 hours time - apparently much of that was simulator, not in an actual airplane. In the United States, he would be barely qualified for the most minimally paying single-pilot single-engine flying job (something like towing banners), let alone flying a major airline passenger jet.

Good Day Sir. Next time,

 

[Just a Reader]

Sorry, I am just a worker-bee. No VIP Invites for me. If I had that kind of clout I would be on the Roadster Forums, not the Model S Forums.

But I do read accident investigations, and this is as far off-topic as I care to go into such a tragedy with a non-pilot:

For if you were a pilot (and Germany has produced some of the worlds finest pilots dating back to Otto Lillienthal, so you would know this) there is still this thing called "airmanship". When the the automation is doing something un-commanded, the well-trained pilot turns it all off and flies the airplane - by hand. Using airmanship and principles of aerodynamics.

Such as "in a stabilzer trim runaway at 500 knots and the thrust levers still in the take-off position, there is too much air-loading on the horizontal stabilizer to control - reduce thrust and speed immediately". And maybe - just maybe - in clear weather looking outside when the instruments don't make sense?

BTW, 500 knots IAS (indicated airspeed) is 160 knots over the red-line (maximum certificated) speed of a 737. Since a 737 does not accelerate as quickly as a Model S, that took some time to reach that state of affairs.

And that is exactly the state of things when Ethiopian contacted the ground and made a 50 foot crater into the ground (that would be 16 meters auf Metric).

Runaway Stabilizer Trim used to be a Memory Item for most of the worlds airlines. And the Very Second thing on that Memory Item List can be summarized for the layman as "Turn off the Automation".

Thanks to the hubris of "software engineers know better than pilots" it is once again a required Memory Item - for most carriers.

And to as your ill-advised comment that the operators being "decently capable" - the first officer of the Ethiopian plane had just 200 hours time - apparently much of that was simulator, not in an actual airplane. In the United States, he would be barely qualified for the most minimally paying single-pilot single-engine flying job (something like towing banners), let alone flying a major airline passenger jet.

Good Day Sir. Next time,

Well, Mr. Altair7,

isn't it remarkable that the only crew that survived this malfunction was a crew who had an off-duty captain on board who could concentrate on analyzing the problem, while the regular crew tried to keep the plane under control? Isn't it equally surprising that no regulator - not even the soft-hearted FAA - was willing to accept Boeing's argument that it was all the crew's fault?
As you claim to be a pilot of both the NG and the Max you will be well aware of the fact that Boeing changed the functioning of overriding the auto-trim - not bothering to tell anyone about it - and by the time the pilots realized what was going on the plane was most likely in an unrecoverable situation.
Finally, Boeing knows their customers. If it takes two Chuck Yaegers in the cockpit to provide the necessary airmanship to keep a shoddily engineered Product in the sky, then they shouldn't have sold it to all and sundry (with no simulators for the Max being available , btw).

Good day to you

 

[emailforbrett]

Well, Mr. Altair7,

isn't it remarkable that the only crew that survived this malfunction was a crew who had an off-duty captain on board who could concentrate on analyzing the problem, while the regular crew tried to keep the plane under control? Isn't it equally surprising that no regulator - not even the soft-hearted FAA - was willing to accept Boeing's argument that it was all the crew's fault?
As you claim to be a pilot of both the NG and the Max you will be well aware of the fact that Boeing changed the functioning of overriding the auto-trim - not bothering to tell anyone about it - and by the time the pilots realized what was going on the plane was most likely in an unrecoverable situation.
Finally, Boeing knows their customers. If it takes two Chuck Yaegers in the cockpit to provide the necessary airmanship to keep a shoddily engineered Product in the sky, then they shouldn't have sold it to all and sundry (with no simulators for the Max being available , btw).

Good day to you

American Airlines had two separate MCAS events prior to these two crashes. In each event, the American crews performed the Stabilizer Runaway procedure, hit the CUTOUT switches, landed uneventfully, and just wrote it up to be fixed.

These two crashes were pilot error, lack of training, and lack of experience. Add some political pressure from all over the globe and a dash of social media and the end result is an extended parking of the airplane to add a second AOA input and some software tweaks. Admittedly, that second AOA input should have been included when the plane was designed. Boeing really screwed that up, as well as not telling the airlines MCAS existed.

 

[Just a Reader]

American Airlines had two separate MCAS events prior to these two crashes. In each event, the American crews performed the Stabilizer Runaway procedure, hit the CUTOUT switches, landed uneventfully, and just wrote it up to be fixed.

These two crashes were pilot error, lack of training, and lack of experience. Add some political pressure from all over the globe and a dash of social media and the end result is an extended parking of the airplane to add a second AOA input and some software tweaks. Admittedly, that second AOA input should have been included when the plane was designed. Boeing really screwed that up, as well as not telling the airlines MCAS existed.

Really? Well, they should have told their own union about it, because it was scathing about Boeing:

American Airlines’ pilots’ union has hit back at Boeing for insinuating that some responsibility for the two crashes of its 737 Max jets lies with the pilots,...Tajer said the Ethiopian Airlines pilots did what they were instructed to do, but that Boeing’s controversial anti-stall software (MCAS) forced the plane into such an aggressive nosedive that the pilots could not recover. “They had wired that thing so that it was irrecoverable. It just blew us away,” Tajer said.
American Airlines union: blaming pilots for Boeing 737 Max crashes 'inexcusable'

If only those pilots had had that superior American airmanship. I don't want to hijack this thread with this topics but the arrogance that is oozing out of your comments is spectacular.

 

[rooter]

Hate to say it because Boeing is a major economic force in my area, but if a pilot is not -told- about MCAS, how can s/he know how to switch it off?

Also, I have personal experience as an enterprise infosec consultant in Boeing's software division, and... ugh. Then again, alot of software consulting companies (Luxoft, Booz-Allen, Slalom, Deloitte), ugh.

Ah yes, every time you come to a complete stop, don't forget to look at the screen whether the car has correctly guessed whether or not to continue in the same direction or go in the opposite direction. You are right, that is so much simpler than assuming the car will go in the direction you tell it to go until you tell it otherwise. That is of course if the screen is actually showing something, not rebooting itself or simply dead.

It is important to try and exercise comprehension, if it is within your abilities.

When the feature comes, of the car automatically choosing gear, remember it will rely on the sensors. If a fence is behind you, it will na-ha-hot choose reverse. If a garage door is in front of you, it will na-ha-hot choose forward. And (needless to say?), it will not zoom off as you sit down; you must step on the brake for anything to activate. (uh- der)

 

[jbcarioca]

...

It is important to try and exercise comprehension, if it is within your abilities.

When the feature comes, of the car automatically choosing gear, remember it will rely on the sensors. If a fence is behind you, it will na-ha-hot choose reverse. If a garage door is in front of you, it will na-ha-hot choose forward. And (needless to say?), it will not zoom off as you sit down; you must step on the brake for anything to activate. (uh- der)

This should be part of any persons basic training. Both in aircraft and cars such basic training is often not done.

I took two advanced driving courses. The first one was mandated by my employer at the time. I just checked, they no longer have such training. BTW, that was a US Government agency with what they then considered special needs. The second was my choice. Both made me much more aware of vehicle and operator limitations.

Tesla high speed accidents often happen, it seems. With 'P' versions and now "Plaid' even more ludicrous than "Ludicrous" the question might be whether driver training and education should be a part of Tesla. There are insurance discounts for all sorts of education/training. Why not have Tesla insurance incentives to teach people how to cope with 1.44G acceleration, or even any Tesla at all?

Early in my piloting career I kept upgrading before typical pilots did. When I bought a Lear with less than 1000 total flight hours the insurance requirement was Flight Safety plus 90 day recurrent training. They also mandated aerobatic training in a taildragger. I did that. Soon my V1 cuts were hardly noticeable except for reduced power. I was not exceptional as a pilot, but exhaustive training served a valuable purpose.

Now, as people point out people with 200 total hours are copilots in 737's. In addition all major airframes try to minimize training requirements because training is expensive. After all an A320 can be flown by most 15 year old video gamers. Ditto for B737...unless something goes wrong, such as a pitot tube icing in an A330 or an autopilot malfunction in a B737. Training, intensive training could have helped prevent those accidents.

What happens with a blowout in a Tesla, or any given system malfunction? We don't really know. It is reasonable to think that some 1-2 day long training might help people make better choices while operating these very complex machines. At a minimum such should be available for people who choose to do it. Tesla insurance surely would give a preferred rate for that.

 

[UncleCreepy]

I don't think Boeings every did anything like this:

Designing out human decision making is an inherently bad idea.

While I agree with the last statement, I'm surprised that you posted this anti Airbus biased video but you also claim

But I do read accident investigations, [...]

If you believe the plane was to blame for that accident, you may want to read the official report from the BEA.

 

[whitex]

Hate to say it because Boeing is a major economic force in my area, but if a pilot is not -told- about MCAS, how can s/he know how to switch it off?

Also, I have personal experience as an enterprise infosec consultant in Boeing's software division, and... ugh. Then again, alot of software consulting companies (Luxoft, Booz-Allen, Slalom, Deloitte), ugh.


It is important to try and exercise comprehension, if it is within your abilities.

When the feature comes, of the car automatically choosing gear, remember it will rely on the sensors. If a fence is behind you, it will na-ha-hot choose reverse. If a garage door is in front of you, it will na-ha-hot choose forward. And (needless to say?), it will not zoom off as you sit down; you must step on the brake for anything to activate. (uh- der)

If you stopped waiting for the construction worker holding a stop sign while in a detour on the "wrong side" of the highway, it's likely the car will take into account that you are in opposite direction lane, with car in front and behind you, so the logical direction to go is backwards. When the car is parallel parked between 2 cars, will it pick forward or backwards based on which of the car, the one in front of behind me, gives me more room to pull out?

Even if Tesla guesses 99.9% correctly (which I sincerely doubt will be the accuracy when it releases), that means 1 in 1000 times it will guess wrong, potentially causing an incident. Sorry, I really don't care for a car which will try to crash once every 1-3 years (assuming daily commuting). It's like the stupid FSD which drives fine most days, excepts every few months gets an update which tries to impale the car onto the median. All those closely bunched up buttons on the steering wheel, horn vs. wipers vs auto-pilot, etc - not for me either, way to easy to hit the wrong one.

I get that some people like to drive experimental cars, all the power to them, I'm just not that crazy about paying Tesla to test their car technology anymore (I originally foolishly believed Elon would actually deliver on his promised on the cars he made the promises on).

 

[HVM]

I don't know controls, but do you think the Blue Loctite would fix this or is it to better use red?