Airlines Blame GA for Holiday Flight Delays

Fully automated or autonomous does not have to mean no humans to mind the store.
 
TCABM said:
Fully automated or autonomous does not have to mean no humans to mind the store.
Click to expand...
Yeah. I'm sure only the best and brightest pilots will sit in an airplane and watch a computer do the flying.
 
LoLPilot said:
I'm not advocating single pilot ops or replacing cargo pilots. Far from it. What I'm saying is that when you look at transportation mishaps, the majority of them are down to human error. When you look at road vehicles, the current crop of autonomous road vehicles are safer than human drivers, even accounting for the accidents that have been attributable to the autonomous cars.
Click to expand...

How do we know ? In 2021 in the U.S., there was only one motor vehicle death per 73 million miles traveled. Has there really been 73 million miles of autonomous road-vehicle testing?

https://www.iihs.org/topics/fatality-statistics/detail/state-by-state

Same thing with aviation. There is a reason that airlines would prefer that pilots fly their jets using the autopilot. You're correct that from a robust statistical standpoint, mishaps involving things like cargo planes are unusual enough that it would take years and years of testing to have a robust statistical test say that yes, the computer flown plane is safer than the human pilot. However we can compare rates of computer failure like the MCAS incidents to human errors dealing with problems in judgment, fatigue, drug use, improper training, etc you'll almost certainly find that the computer brain is more reliable.
Click to expand...

A good autopilot is a great tool, but one of the most important issues is knowing when to turn it off. The biggest concern I have about AI is how it will do in evaluating and responding to bad weather and emergency situations. That's why I want to see statistically significant data from real-world flying, not plausibility arguments.

THAT BEING SAID...

As I said above, I wouldn't want to ride in a jet that didn't have a human pilot (or two) up front. Studies of consumer preference found that most people wanted a person driving the bus, operating the subway, etc. I have a feeling that most people feel better knowing that in the front of every train there is a conductor and an engineer who have people or things that they want to go home to.
Click to expand...

I want whoever or whatever is piloting the aircraft to have as much to lose as I do.
 
brien23 said:
If Boeing and Airbus are working on fully automated aircraft, it's not if but when it's going to happen. Now will the people fly on them or not, given the price of a ticket if the price is a lot cheaper people will chose the cheaper flight. If you are one that won't fly on one plan on the bus or train that is unless they also are fully automated.
Click to expand...
Big corporations have had many successes, but they have also worked on things that turned out not to be successful. So I wouldn't consider the fact that they're working on something to be proof that it will necessarily succeed.
 
Salty said:
Yeah. I'm sure only the best and brightest pilots will sit in an airplane and watch a computer do the flying.
Click to expand...

Pay me enough bux and I’d do it. Besides, there’s always the up & comers willing to give their time away for free just to build hours.
 
TCABM said:
Pay me enough bux and I’d do it. Besides, there’s always the up & comers willing to give their time away for free just to build hours.
Click to expand...
Yeah, and they’ll be well trained and sharp as tacks I’m sure.
 
Zeldman said:
A little thread drift...but the Safe Return Emergency Autoland system seems to work so far.


I wish to buy a Cirrus Vision Jet but I am $2,999,999.99 short...
Click to expand...
Driving by TKI on Sunday I noticed the Cirrus facility had left a couple on the ramp and the place was closed. There's a chance one might be unlocked.

If you're up to a 200 yard sprint across a winter wheat field and scaling the 8' chain link fence topped with razor wire, you could get lucky.

:rofl:
 
I mean, lack of handflying isn't the bug, it's the feature, for those attracted to that kind of aviation work. I don't think increased automation in that already automated space would substantively impact recruitment and retention, beyond whatever outright reduction in crewed positions that technology brings or doesn't bring.
 
You don't need AI for autonomous flight. In a controlled environment, rules-based algorithms are entirely capable of controlling aircraft. AI would actually make flight less safe by introducing uncertainty into computational results.
 
Ed Haywood said:
You don't need AI for autonomous flight. In a controlled environment, rules-based algorithms are entirely capable of controlling aircraft. AI would actually make flight less safe by introducing uncertainty into computational results.
Click to expand...
I can see the fully automated cockpit of the future having a human and a dog seated in front. The human is there to make the passengers feel safe and the dog is there to attack the human if he reaches for any of the controls.
 
Ed Haywood said:
You don't need AI for autonomous flight. In a controlled environment, rules-based algorithms are entirely capable of controlling aircraft. AI would actually make flight less safe by introducing uncertainty into computational results.
Click to expand...
Whatever you call it, my skepticism remains about its ability to make appropriate decisions in challenging situations.
 
LoLPilot said:
Where do you draw the line, though? Let's say that you are an older person who owns something like a 182 that you used for personal transportation, but you either lost your medical or don't feel like you can safely fly it solo anymore. So, you hire a commercial pilot that you know to fly the airplane either with you as a copilot to do things like operate the radios and help with navigation or just with you as a passenger so that you can continue to travel in your own airplane. That's really not appreciably different from a Part 91 corporate operation where they own something like a Citation or a Gulfstream and hire a type rated pilot to fly it.



I mean look at rail. Railroads, especially light rail like subway systems could be automated, and that technology has existed for years. Yet we will have subway operators and trains still have a conductor and an engineer in the cabs of locomotives. Some of that is due to union agreements from my understanding, but I can't believe that part of that is not also due to the fact that you are hardly alone in your assessment that you would not want to ride in something that wasn't operated by a person who had family / a pet / hobbies / a favorite TV show / some other reason to entice them to want to go home from work safely.
Click to expand...
I drew the line.
 
Palmpilot said:
How do we know ? In 2021 in the U.S., there was only one motor vehicle death per 73 million miles traveled. Has there really been 73 million miles of autonomous road-vehicle testing?

https://www.iihs.org/topics/fatality-statistics/detail/state-by-state



A good autopilot is a great tool, but one of the most important issues is knowing when to turn it off. The biggest concern I have about AI is how it will do in evaluating and responding to bad weather and emergency situations. That's why I want to see statistically significant data from real-world flying, not plausibility arguments.



I want whoever or whatever is piloting the aircraft to have as much to lose as I do.
Click to expand...

Do you think that I am disagreeing with you here? I'm not.

As far as autonomous vehicle testing, AV Industry Association reports 44 million miles traveled by member vehicles, with 1 serious injury: https://theavindustry.org/resources/blog/data-44million-miles

A NHTSA report found that 45% of fatal crashes involved a driver who was either drunk, speeding, or not restrained. AVs couldn't do anything about passenger restraints but they won't drive drunk. I worked on a study of commercial vehicle traffic crashes that identified fatigue and speeding as two of the main contributing factors to how bad an accident was. Simply put, a computer isn't going to engage in distracted or impaired driving, which gives it a leg up over us. That doesn't mean I advocate for autonomous vehicles. In fact, I'm against them aside from the mobility opportunities they offer to people who cannot drive for some reason, whether that is an injury, age, etc. But thems the facts.
 
Clip4 said:
I drew the line.
Click to expand...

So last night I did night currency work. I hadn't flown at night in many months (I resisted the urge to say since last year) and wanted a second pilot, so I paid a friend who is a CFI to go along with me in the right seat while I flew off some landings. I have also hired friends who are commercial rated to fly as safety pilots for me to do practice approaches. By definition, I hired a commercial pilot. Were we no longer general aviation in these circumstances?
 
LoLPilot said:
We have had automated trains for several years. If anything, they are the easiest vehicle to automate, and yet even freight rail maintains two crewmembers in the cab.
Click to expand...
that's because, as we've seen with the automated cars being tested in major cities, automation can't handle the edge cases. Humans can.
 
Half Fast said:
Why would watching a computer fly count toward flight experience?
Click to expand...

You seem to think 90+% of part 121 flying isn’t monitoring computers already, but I never discussed qualifications.

Salty asked “who would want to do that?”. For enough money, I’d do it. Why? Easy money.
 
TCABM said:
You seem to think 90+% of part 121 flying isn’t monitoring computers already, but I never discussed qualifications.

Salty asked “who would want to do that?”. For enough money, I’d do it. Why? Easy money.
Click to expand...
Not really. I was really asking who would want to trust their lives to people who would want to do that.

I don't mean that personally.

I suspect that you would get bored of that REAL quick, and I wouldn't trust even a skilled pilot that was willing to let their skills atrophy while watching a computer fly.

There's a huge difference between what we have now and an "automated" solution where you are just there if all else fails.
 
TCABM said:
You seem to think 90+% of part 121 flying isn’t monitoring computers already, but I never discussed qualifications.
Click to expand...

And lack of hand flying experience should be a concern, something we should be correcting, not making worse. When the fit hits the shan, will an autopilot-watcher who hasn’t flown much by hand for years really be able to save the day?

So what’s the advantage of AI? Going from 90% to 99.99999% while still paying someone to be in the cockpit?
 
Salty said:
Not really. I was really asking who would want to trust their lives to people who would want to do that...
Click to expand...
Have you seen the airlines people fly today? “The flying public” cares a lot less about what’s going on upfront than we, as a community, like to believe. They just know flying is safe enough for them to pay for the trip at a certain price point. Proof in point is a youtube full of gasping-in-fear pax during light chop at FL350.

Half Fast said:
….So what’s the advantage of AI? Going from 90% to 99.99999% while still paying someone to be in the cockpit?
Click to expand...
Whoever said AI? Full automation doesn’t require AI, just look at autoland systems. PF/PM are both just monitoring from AP engage on climbout until taxi to the gates already.
 
LoLPilot said:
That doesn't mean I advocate for autonomous vehicles. In fact, I'm against them aside from the mobility opportunities they offer to people who cannot drive for some reason, whether that is an injury, age, etc.
Click to expand...
I'm curious; you argue that AVs are likely safer than human drivers, but also state that you're generally against them. What are the reasons you oppose them?
 
LoLPilot said:
Do you think that I am disagreeing with you here? I'm not.

As far as autonomous vehicle testing, AV Industry Association reports 44 million miles traveled by member vehicles, with 1 serious injury: https://theavindustry.org/resources/blog/data-44million-miles

A NHTSA report found that 45% of fatal crashes involved a driver who was either drunk, speeding, or not restrained. AVs couldn't do anything about passenger restraints but they won't drive drunk. I worked on a study of commercial vehicle traffic crashes that identified fatigue and speeding as two of the main contributing factors to how bad an accident was. Simply put, a computer isn't going to engage in distracted or impaired driving, which gives it a leg up over us. That doesn't mean I advocate for autonomous vehicles. In fact, I'm against them aside from the mobility opportunities they offer to people who cannot drive for some reason, whether that is an injury, age, etc. But thems the facts.
Click to expand...

Don't mind me; I was just nit-picking some details.

I didn't realize that the testing total had gotten that high. Thanks for the info.
 
TCABM said:
Whoever said AI? Full automation doesn’t require AI, just look at autoland systems. PF/PM are both just monitoring from AP engage on climbout until taxi to the gates already.
Click to expand...

Okay, then rephrase what I wrote.

So what’s the advantage of to full automation? Going from 90% to 99.99999% while still paying someone to be in the cockpit?​
 
TCABM said:
Whoever said AI? Full automation doesn’t require AI, just look at autoland systems. PF/PM are both just monitoring from AP engage on climbout until taxi to the gates already.
Click to expand...

Flying safely requires the ability to make intelligent decisions in hazardous situations. Witmo said that the airlines won't need human pilots anymore.
 
Last edited:
Half Fast said:
Okay, then rephrase what I wrote.

So what’s the advantage of to full automation?​
Click to expand...
No idea the economics of it, but it still doesn’t change my position: pay me enough money and I’ll watch an airplane fly itself along, given I can override it when it’s gone stupid.

That’s the problem with every shiny object syndrome: proponents don’t recognize implementing “innovation” is incremental after you get to critical mass and in the back side of the bell curve.[/i][/indent]
 
Palmpilot said:
Flying safely requires the ability to make intelligent decisions in hazardous situations. Witmo said that the airlines won't need human pilots anymore.
Click to expand...

I’m not @Witmo and am not arguing in support of his position.

I find it highly unlikely that in my lifetime (I’m 50) we’ll get to mass transit with no human at the front of the transportation appliance.
 
TCABM said:
I find it highly unlikely that in my lifetime (I’m 50) we’ll get to mass transit with no human at the front of the transportation appliance.
Click to expand...
We already have it in London (Docklands Light Rail). I'm sure it's in plenty of other places around the world too.
 
TCABM said:
I’m not @Witmo and am not arguing in support of his position.
Click to expand...

You asked "Whoever said AI?". I was attempting to answer that question.

TCABM said:
I find it highly unlikely that in my lifetime (I’m 50) we’ll get to mass transit with no human at the front of the transportation appliance.
Click to expand...
You could be right. [Oops. I just saw Katamarino's datapoint.]
 
Enormous difference between a train on rails with basically one control Stop <-----> Go and an aircraft with dozens, perhaps hundreds of things to control. Even if you add in automated switching logic, it's still a pretty simple finite state machine with no steering, compared to controlling flight.
 
Half Fast said:
I would point out that speed can also be controlled, but then @Ed Haywood would accuse me of nitpicking, so I won’t.
Click to expand...
That’s why there’s dashes between the two. But it’s basically just a slider.
 
Katamarino said:
I'm curious; you argue that AVs are likely safer than human drivers, but also state that you're generally against them. What are the reasons you oppose them?
Click to expand...

I had an earlier post in the thread about how we need to decide if the point of life is to exist or if the point of life is to live it. I think there is a happy medium but I'm firmly against removing someone's agency because it's good for them. I fear that if AVs proliferate that we are a short time away from either someone arguing that we should outlaw human drivers or insurance companies making human driven cars so onerous to insure that it is virtually impossible to drive oneself. As I said in the earlier post I also ride motorcycles and obviously fly airplanes. It is entirely possible to admit that something is safer / more efficient / objectively better and still be against it for personal or philosophical reasons. Drinking alcohol is bad for you, and I doubt I'm alone on here in saying I enjoy throwing one back.
 
You must log in or register to reply here.
Back
Top