The Future Of Artificial Intelligence
#41
I think it's funny people point to cargo as being the first to adopt pilotless cockpits. Look at their history: they have been the last to park their planes still with FEs, with some still flying around with FEs.
The next gen aircraft being developed now are still 2 pilot, and even cargo trains are still two "pilot." Given that the next gen are still designed to be two pilot, and airliners pushed beyond 30-40 years... we don't have anything to worry about for a while. Maybe when I retire 29 years from now we'll see that beginning to take hold...
The next gen aircraft being developed now are still 2 pilot, and even cargo trains are still two "pilot." Given that the next gen are still designed to be two pilot, and airliners pushed beyond 30-40 years... we don't have anything to worry about for a while. Maybe when I retire 29 years from now we'll see that beginning to take hold...
#42
Gets Weekends Off
Joined APC: Dec 2005
Posts: 8,939
I think it's funny people point to cargo as being the first to adopt pilotless cockpits. Look at their history: they have been the last to park their planes still with FEs, with some still flying around with FEs.
The next gen aircraft being developed now are still 2 pilot, and even cargo trains are still two "pilot." Given that the next gen are still designed to be two pilot, and airliners pushed beyond 30-40 years... we don't have anything to worry about for a while. Maybe when I retire 29 years from now we'll see that beginning to take hold...
The next gen aircraft being developed now are still 2 pilot, and even cargo trains are still two "pilot." Given that the next gen are still designed to be two pilot, and airliners pushed beyond 30-40 years... we don't have anything to worry about for a while. Maybe when I retire 29 years from now we'll see that beginning to take hold...
#43
Gets Weekends Off
Joined APC: May 2014
Posts: 1,681
It will be decades (at least.)
If all the silicon valley brainiacs can't get something like this figured out:
https://arstechnica.com/gadgets/2017...y-bot-drowned/
They will have quite a ways to go before automating anything that actually matters.
This dumbass thing had one job.... One extremely simple job. Unfortunately, it didn't go so well.
(Most people don't understand AI. It's a lot like the early days of PCs. A lot of people were intimidated about trying to learn computer stuff because they assigned way, way more capabilities to a PC than it actually had. Today, people do the same thing. AI is very much just a bar trick at the present. People who are honest about the technology understand that it is very, very limited.)
If all the silicon valley brainiacs can't get something like this figured out:
https://arstechnica.com/gadgets/2017...y-bot-drowned/
They will have quite a ways to go before automating anything that actually matters.
This dumbass thing had one job.... One extremely simple job. Unfortunately, it didn't go so well.
(Most people don't understand AI. It's a lot like the early days of PCs. A lot of people were intimidated about trying to learn computer stuff because they assigned way, way more capabilities to a PC than it actually had. Today, people do the same thing. AI is very much just a bar trick at the present. People who are honest about the technology understand that it is very, very limited.)
#44
Banned
Joined APC: Feb 2011
Position: 756 Left Side
Posts: 1,629
Many of you might want to pick up (better yet, get a subscription to) Aviation Week & Space Technology.
Unmanned aircraft have been running supply missions for years overseas to test it's feasibility.
It was successful. Now they are taking it to the next step-
https://www.wired.com/2016/11/lockhe...ghting-drones/
Then you have this. They have taken the Piaggio Aerospace P.180 Avanti II business aviation aircraft and turned it into-
Piaggio Aerospacespace - P.1HH HammerHead
People need to learn the difference between Autonomous, AI, Remote control, Drone and reduced manning.
I do not believe we will EVER see a passenger flight flown without a pilot onboard. However, I fully expect to see both passenger and cargo being flown by reduced manning with a remote pilot sitting in some control station.
The question is when?
Let's be honest. We already have Autonomous in most of our aircraft. We have autopilots. Different levels based on age of aircraft, and what the airline decides to buy.
Keep in mind, even the newest A350/B787-10 can't rely on it's automony (CatIII Autoland) if it's landing at an airfield that doesn't have a CatIII certified ILS.
At the same time, we can all remember times (in this past year?) where something was MEL'd. Center Autopilot?, Auto throttles?, etc..
Yes, there are many problems to overcome. But then again, the way technology continues to evolve and make major gains, it would be foolish to believe change will not happen within the next 10-15 years.
Just my opinion.
Motch
Unmanned aircraft have been running supply missions for years overseas to test it's feasibility.
It was successful. Now they are taking it to the next step-
https://www.wired.com/2016/11/lockhe...ghting-drones/
Then you have this. They have taken the Piaggio Aerospace P.180 Avanti II business aviation aircraft and turned it into-
Piaggio Aerospacespace - P.1HH HammerHead
People need to learn the difference between Autonomous, AI, Remote control, Drone and reduced manning.
I do not believe we will EVER see a passenger flight flown without a pilot onboard. However, I fully expect to see both passenger and cargo being flown by reduced manning with a remote pilot sitting in some control station.
The question is when?
Let's be honest. We already have Autonomous in most of our aircraft. We have autopilots. Different levels based on age of aircraft, and what the airline decides to buy.
Keep in mind, even the newest A350/B787-10 can't rely on it's automony (CatIII Autoland) if it's landing at an airfield that doesn't have a CatIII certified ILS.
At the same time, we can all remember times (in this past year?) where something was MEL'd. Center Autopilot?, Auto throttles?, etc..
Yes, there are many problems to overcome. But then again, the way technology continues to evolve and make major gains, it would be foolish to believe change will not happen within the next 10-15 years.
Just my opinion.
Motch
#45
Gets Weekends Off
Joined APC: May 2014
Posts: 1,681
Let's be honest. We already have Autonomous in most of our aircraft. We have autopilots. Different levels based on age of aircraft, and what the airline decides to buy.
Keep in mind, even the newest A350/B787-10 can't rely on it's automony (CatIII Autoland) if it's landing at an airfield that doesn't have a CatIII certified ILS.
At the same time, we can all remember times (in this past year?) where something was MEL'd. Center Autopilot?, Auto throttles?, etc..
Yes, there are many problems to overcome. But then again, the way technology continues to evolve and make major gains, it would be foolish to believe change will not happen within the next 10-15 years.
Just my opinion.
Motch
Keep in mind, even the newest A350/B787-10 can't rely on it's automony (CatIII Autoland) if it's landing at an airfield that doesn't have a CatIII certified ILS.
At the same time, we can all remember times (in this past year?) where something was MEL'd. Center Autopilot?, Auto throttles?, etc..
Yes, there are many problems to overcome. But then again, the way technology continues to evolve and make major gains, it would be foolish to believe change will not happen within the next 10-15 years.
Just my opinion.
Motch
I have flown several generations of aircraft. I find myself solving much more opaque and goofy problems in an airbus than in any previous airplane.
Automation is a double-edged sword. Airbus pretty much took stick-and-ridder skills out of the equation, but they created a computer network that you better know what you are doing to manage-and you better watch like a damn hawk.
Look how many crashes in the last decade have involved Airbii..... There are reasons for that, and screwy automation is a very big component.
Machines are simple to manage. They do machine stuff. Networks sometimes fail in very strange ways. Personally, I have seen some things in an A320 which have no explanation in the book, from check airmen or from anyone else. When I flew purely mechanical planes, every failure always made logical sense-and was very easy to figure out. I have to jump in and make the plane do something right at least a couple times a month. With older generations, that wasn't even a thing. They would do exactly what you told them to do, every time.... No surprises.
An airbus will fix some poor stick and rudder problems, but it will eat your ass up if you are a poor manager of automation.
#46
:-)
Joined APC: Feb 2007
Posts: 7,339
AI won't cut out the pilot, it will cut the current pilot's salaries in half, just like the last round of automation did. Consult your current pay check for proof of that. Purchasing power is down 50% or more for all pilots, from as little as 20 years ago.
#47
My purchasing power is the same as my dad at NW in the early 90s. He was a 320 captain then, and I'll be one later this year.
#48
No it's not that plain and simple: you're not going to cut that other pilot, you'll send him to a ground station somewhere to help in case the 1 pilot you have or the "computer" goes TU. I'm assuming he will still have a salary. Second, any cost savings by eliminating the cost of salaries for FOs would be easily offset by the initial costs of fielding and mitigating the risks, legal, regulatory demands, etc. The only real driver for when (which it probably will but not in our lifetimes) it happens is cost. Air traffic demand is pretty constant (barring a black swan event) so even if there were cost savings, how much of a lower ticket price would pax really see if the execs making these decisions, motivated by cost wanted to increase profit margin?
I can't believe we are still talking about this.
I can't believe we are still talking about this.
#50
Gets Weekends Off
Joined APC: Jul 2016
Posts: 463
A Robot Copilot Just Flew?and Landed?a 737 Sim
I think its very likely you will see these AI / robot copilots within 5-10 years. First in military capacity like the C17's and tankers then I would guess it would move to civilian cargo before passenger service. You might see the robot replace the relief pilot first. You know they call that "dozing for dollars"
Imagine the Robot pilot, Never complains, calls in sick, works 24 / 7. If the captain is doing something non SOP it will call him / her on it. And it will email the Chief pilot right away with a video and data of what the captain did. You can't hurt HAL's feelings.
I read Aviation Week and Popular Mechanics and other business trades on a regular basis. Its coming a lot faster than you think. Just 20 years ago I had to call my answering machine from a payphone to see if I had any messages. Now I'm texting my friend In Europe and have the whole world of answers in my hand.
I have about 16 years left. I don't think I'll see it at my airline, but definitely in Cargo by the time I retire.
I think its very likely you will see these AI / robot copilots within 5-10 years. First in military capacity like the C17's and tankers then I would guess it would move to civilian cargo before passenger service. You might see the robot replace the relief pilot first. You know they call that "dozing for dollars"
Imagine the Robot pilot, Never complains, calls in sick, works 24 / 7. If the captain is doing something non SOP it will call him / her on it. And it will email the Chief pilot right away with a video and data of what the captain did. You can't hurt HAL's feelings.
I read Aviation Week and Popular Mechanics and other business trades on a regular basis. Its coming a lot faster than you think. Just 20 years ago I had to call my answering machine from a payphone to see if I had any messages. Now I'm texting my friend In Europe and have the whole world of answers in my hand.
I have about 16 years left. I don't think I'll see it at my airline, but definitely in Cargo by the time I retire.
I've seen this article thrown around: First, it didn't land the plane, it interfaced with the autopilot during an autoland; it literally pushed buttons. Second, it managed a simulator during CAVOC conditions. What happens when you have you have to hand fly an approach to landing in gusty crosswinds and on an icy runway.
Also, not sure you know anything about the C-17 mission if you think the autopilot just drones from point A to point B.
Thread
Thread Starter
Forum
Replies
Last Post