797 to possibly be built for one pilot?
06-05-2019, 05:06 AM
#61
Gets Weekends Off
Joined APC: Feb 2009
Posts: 1,267
Originally Posted by rickair7777
Apples to oranges.
Correct that the "problem set" for autonomous car navigation is more complicated than for airplane navigation, since the car must interpret and react to the environment and unexpected obstacles.
In aviation the navigation environment is already known and in a database, both routes and terrain. Pop-up obstacles consist of weather and other aircraft. With ADS-B and ATC radar, plus some optical see and avoid technology, other aircraft are not too big of a challenge in positively controlled airspace. They might have to keep autonomous aircraft in A/B/C/D airspace, with added transition corridors so as not to routinely "test" autonomous see and avoid, ie reserve it as an additional backup layer of defense if ATC separation fails, as opposed to just turning them loose with all the bug-smashers on a Sunday afternoon.
That's the easy part. The hard part is all the stuff we do which requires judgement, and that includes WX avoidance. Autonomous cars can just revert to human driver mode or pull over if they get confused or malfunction... aircraft will have to complete the mission to a full stop on a runway. The really hard part is doing that in a cost-effective manner, and to make matters worse it will take a MASSIVE up front investment which will gradually pay off over many decades. The airlines don't think that far ahead. The government has no real incentive (airplanes don't vote but pilots do). Boeing has more pressing matters on their plate.
And all that assumes you have some sort of AI which can respond to stuff nobody thought of when they programmed it. Which they don't, and it's not even on the horizon yet.
Correct that the "problem set" for autonomous car navigation is more complicated than for airplane navigation, since the car must interpret and react to the environment and unexpected obstacles.
In aviation the navigation environment is already known and in a database, both routes and terrain. Pop-up obstacles consist of weather and other aircraft. With ADS-B and ATC radar, plus some optical see and avoid technology, other aircraft are not too big of a challenge in positively controlled airspace. They might have to keep autonomous aircraft in A/B/C/D airspace, with added transition corridors so as not to routinely "test" autonomous see and avoid, ie reserve it as an additional backup layer of defense if ATC separation fails, as opposed to just turning them loose with all the bug-smashers on a Sunday afternoon.
That's the easy part. The hard part is all the stuff we do which requires judgement, and that includes WX avoidance. Autonomous cars can just revert to human driver mode or pull over if they get confused or malfunction... aircraft will have to complete the mission to a full stop on a runway. The really hard part is doing that in a cost-effective manner, and to make matters worse it will take a MASSIVE up front investment which will gradually pay off over many decades. The airlines don't think that far ahead. The government has no real incentive (airplanes don't vote but pilots do). Boeing has more pressing matters on their plate.
And all that assumes you have some sort of AI which can respond to stuff nobody thought of when they programmed it. Which they don't, and it's not even on the horizon yet.
06-12-2019, 11:44 AM
#62
On Reserve
Joined APC: Jun 2017
Posts: 13
Originally Posted by rickair7777
Management would love that. Free negotiating capital for something which would never affect them personally. Way too soon IMO.
And if it comes to that, trying to indefinitely stall inevitable technological progress is a losing battle. Better to have language that simply ensures full pay to age 65 for pilots displaced by automation. Airlines will need to do that anyway, otherwise nobody in their right mind would invest in the training and dues paying to get into the career. They'll have to bridge the valley once it becomes obvious that there's a realistic timeline for the deployment of significant automation.
And if it comes to that, trying to indefinitely stall inevitable technological progress is a losing battle. Better to have language that simply ensures full pay to age 65 for pilots displaced by automation. Airlines will need to do that anyway, otherwise nobody in their right mind would invest in the training and dues paying to get into the career. They'll have to bridge the valley once it becomes obvious that there's a realistic timeline for the deployment of significant automation.
Thread
Thread Starter
Forum
Replies
Last Post
