PILOT SPIN
Spin Zone => Spin Zone => Topic started by: Rush on March 28, 2019, 02:06:17 PM
-
Oh that makes me feel good.
-
Oh that makes me feel good.
I read earlier today where Boeing said the 737 MAX is now "safer".
SAFER???? Safer than what?
How about SAFE!
-
I read earlier today where Boeing said the 737 MAX is now "safer".
SAFER???? Safer than what?
How about SAFE!
You must have detected my sarcasm.
-
FLY the plane.
-
FLY the plane.
I do realize mechanical connections can fail, but at least I can wrap my mind around mechanical connections. "Software" flying a plane is like a soulless box of fog with no apparent relationship between input and output.
-
That software will fly the plane far more efficiently than a human.
-
That software will fly the plane far more efficiently than a human.
Assuming that the inputs are correct and are interpreted correctly. The Blue Screen of Death quite literally on some unforseen fault. The problem with software and aircraft is that there are a LOT of unknowns. It all goes smoothly until it doesn't. When it gets into a corner it can't understand, suddenly going "I give up, it's your airplane" is not a good thing.
-
Assuming that the inputs are correct and are interpreted correctly. The Blue Screen of Death quite literally on some unforseen fault. The problem with software and aircraft is that there are a LOT of unknowns. It all goes smoothly until it doesn't. When it gets into a corner it can't understand, suddenly going "I give up, it's your airplane" is not a good thing.
Or as in the recent cases, "I can't do it, but I won't let you do it either".
-
That software will fly the plane far more efficiently TO THE SCENE OF THE CRASH than a human.
FTFY.
I write software sometimes. I am afraid.
-
Assuming that the inputs are correct and are interpreted correctly. The Blue Screen of Death quite literally on some unforseen fault. The problem with software and aircraft is that there are a LOT of unknowns. It all goes smoothly until it doesn't. When it gets into a corner it can't understand, suddenly going "I give up, it's your airplane" is not a good thing.
Large aircraft have been flown by progressively more sophisticated software for the past 38 years and millions upon millions of hours. When you compare the millions of hours flown versus the accidents, it's quite phenomenal.
Today's aircraft have tremendous redundancy. The latest problem with the Max8 has more to do with a poor design compounded with a manufacturer desire to do something it should have not done to satisfy customer demands, and an agency oversight that permitted it to happen.
-
FTFY.
I write software sometimes. I am afraid.
You don't write software for major aircraft manufacturers. Not even close.
-
You don't write software for major aircraft manufacturers. Not even close.
That is correct. I actually don’t write much in the way of any software anymore, it’s more about making sure the developers do it right. That is difficult enough.
Your point is....???
-
That software will fly the plane far more efficiently than a human.
Until it doesn’t. I haven’t seen a glitch-free software program yet.
-
Until it doesn’t. I haven’t seen a glitch-free software program yet.
Have you ever seen a glitch-free human flying an airplane, driving a car or typing on PS?
-
Have you ever seen a glitch-free human flying an airplane, driving a car or typing on PS?
That's a true statement, and there have been many instances of pilot error causing issues, and accidents. However, I'd rather have a trained, experienced human brain dealing with a problem than the software flying the plane for me, that think there's a problem that isn't. But I get what you are saying, and that's a valid point.
-
Until it doesn’t. I haven’t seen a glitch-free software program yet.
How do you explain the millions of hours flown by Airbus and Boeing using fly by wire? IIRC, Airbus or Boeing have yet to have an aircraft have to land using direct law. The redundancy is built in and has proven to be rather effective.
Like stated earlier, this has been in progression since the early 80's, and everything (large transport) is using it. Even the modern business jets are very heavily software controlled.
-
I don't deny the efficiency and benefits of computerized technology. There's no turning back, we're all seduced. The benefits are overwhelming. But there's a dark side: It's turning us, as humans, into devolving lumps of useless flesh.
-
I don't deny the efficiency and benefits of computerized technology. There's no turning back, we're all seduced. The benefits are overwhelming. But there's a dark side: It's turning us, as humans, into devolving lumps of useless flesh.
:( Good morning to you, too. Reminds me of the Star Trek episode where a planet of beings had evolved into nothing but brains resting on pedestals. They snatched the Enterprise crew so they could occupy their bodies and start over.
Humans don’t seem to recognize their own march of devolution until it is a smidge too late. :(
It’s almost as if we are inexorably drawn to create existential challenges for ourselves to overcome.
-
So are we saying the technology used today in large transports should devolve back to say 1950's technology? Should we go back to cables and hydraulics and manually controlled engine fuel systems? Single analog autopilots?
-
Artificial Intelligence is way farther along than many of us know. It is scary smart, and yes ultimately a threat to humans. We are becoming obsolete. Yeah, Yeah, I know Skynet and all that. Talk to some people high up in the corporate structure in Silicon Valley. Even they are scared of it, and holding back a lot of the facts about it.
-
Scary story.
Have a Phd friend that is doing some REALLY high level AI stuff. Interestingly they are using GPU's to do the heavy processing.
Anyway, they developed a system that could analyze MRI's, etc. It very quickly got to the point that it was way better than a doctor at detecting anomalies. Then the scary part. It has started exhibiting behaviours (it is a learning algorithm) that they can not explain. i.e. it is doing things it shouldn't be. Small scale at this point, but Skynet stuff...
-Dan
-
How do you explain the millions of hours flown by Airbus and Boeing using fly by wire?
They usually get it right, there are millions of hours flown within predicted parameters.
The MCAS programmers found a way to get it wrong, but superior pilot training and a great US reporting system has prevented anything from happening in the US yet.
-
Have you ever seen a glitch-free human flying an airplane, driving a car or typing on PS?
Nope, but I haven’t seen a computer reboot or fix a glitch in a human. The opposite is quite true on a daily basis however.
Why did they bother putting a front window and stick in the space shuttle, and require a pilot to be in the left seat? Certainly a computer should have been able to bring one of those flying bricks onto a runway, right?
-
Nope, but I haven’t seen a computer reboot or fix a glitch in a human. The opposite is quite true on a daily basis however.
Why did they bother putting a front window and stick in the space shuttle, and require a pilot to be in the left seat? Certainly a computer should have been able to bring one of those flying bricks onto a runway, right?
Space shuttle had ancient technology. A Boeing 757/767 had far more advance flight technology (with regards to avionics) than the shuttle. At the time you could program the 75/76 to do a complete flight via the FMS's (2) and the autopilots (3) and the auto thrust system. All the pilot really needed to do was the takeoff with the gear and flap retraction, and on approach put the gear down with flaps. The plane would auto land and keep the nosewheel within 6 inches of the centerline and apply the brakes. The shuttle? It's ancient avionics had to have the flight profile updated via a ground link before the flight (which meant if the winds had shifted at launch time it was a no go because the flight plan had to be updated and uploaded via ground link).
With what we have today, the original configuration of the B75/76 is outdated, tremendously. Gyro technology is far superior as is flat panel displays. Integration of GPS into FMS has the IRS accurate down to several feet, rather than thousands of feet. Engine FADEC is so reliable now that we have ETOPS out to 330 minutes.
-
How do you explain the millions of hours flown by Airbus and Boeing using fly by wire? IIRC, Airbus or Boeing have yet to have an aircraft have to land using direct law. The redundancy is built in and has proven to be rather effective.
Like stated earlier, this has been in progression since the early 80's, and everything (large transport) is using it. Even the modern business jets are very heavily software controlled.
What you said is true, and does not invalidate my statement.
There are compromises in everything. Make everything fly by wire, and you can save weight by making components thinner. They didn’t help AA Flight 587, an Airbus A300 that has the vertical stabilizer tear off because of excess rudder forces.
I have been involved in some Safety Stand Downs in the CAF. One lesson that keeps popping up is a four-part video of NASA astronaut Mike Mullane talking about “Normalization of Deviance.” Put simply, that’s normalizing actions because you haven’t had a problem in the past. Example: not wearing your seat belt because you haven’t been in an accident. In Mike’s case, he was demonstrating that the Challenger and Columbia crashes weren’t accidents; they were predicable events.
This is part 1 and well worth the 16 minutes. The other 3 parts can be found on YouTube and are great.
https://youtu.be/Ljzj9Msli5o
-
What you said is true, and does not invalidate my statement.
There are compromises in everything. Make everything fly by wire, and you can save weight by making components thinner. They didn’t help AA Flight 587, an Airbus A300 that has the vertical stabilizer tear off because of excess rudder forces.
I have been involved in some Safety Stand Downs in the CAF. One lesson that keeps popping up is a four-part video of NASA astronaut Mike Mullane talking about “Normalization of Deviance.” Put simply, that’s normalizing actions because you haven’t had a problem in the past. Example: not wearing your seat belt because you haven’t been in an accident. In Mike’s case, he was demonstrating that the Challenger and Columbia crashes weren’t accidents; they were predicable events.
This is part 1 and well worth the 16 minutes. The other 3 parts can be found on YouTube and are great.
https://youtu.be/Ljzj9Msli5o
I've sat through several case studies on Columbia and Challenger. It's absolutely astounding the lapses that led up to those accidents.
If you want some good reading, find A Human Error Approach to Aviation Accident Analysis: The Human Factors Analysis and Classification System. I've attended seminars by Dr. Shappell and his insight is very interesting.