Skip to main content

Guitarmike posted:
Big_Boy_4005 posted:
J 611 posted:
Byrdie posted:

Just a thought provoking question - how many of you use the self checkout at the many, many stores that are now implementing them?

I use it all the time. If you only have a few items with bar codes its much faster and more convenient than waiting for a cashier. Automation is coming. What we have to figure out is how to take care of the millions that will be out of their current jobs. Truck drivers are on the front lines here in my opinion. It will be interesting to see how society will react. How governments will react. This will be a much larger change than when we went from a mostly agrarian work force to an industrialized one. The next 50 to 100 years will be very interesting. 

Well, when things get screwed up at the self checkout, guess what, the human comes over to fix it.

Same for trains and just about everything else.

Just wait, the self checkout in the McDonald's in the train station will get screwed up and the autopilot trains will crash through the bumpers at the end of the tracks in the station while asking you if you want your receipt.....

Almost sounds like the end of the movie Silver Streak. That train had no engineer when it crashed into Union Station.

Guitarmike posted:
Big_Boy_4005 posted:
J 611 posted:
Byrdie posted:

Just a thought provoking question - how many of you use the self checkout at the many, many stores that are now implementing them?

I use it all the time. If you only have a few items with bar codes its much faster and more convenient than waiting for a cashier. Automation is coming. What we have to figure out is how to take care of the millions that will be out of their current jobs. Truck drivers are on the front lines here in my opinion. It will be interesting to see how society will react. How governments will react. This will be a much larger change than when we went from a mostly agrarian work force to an industrialized one. The next 50 to 100 years will be very interesting. 

Well, when things get screwed up at the self checkout, guess what, the human comes over to fix it.

Same for trains and just about everything else.

Just wait, the self checkout in the McDonald's in the train station will get screwed up and the autopilot trains will crash through the bumpers at the end of the tracks in the station while asking you if you want your receipt.....

I take your response as tongue in cheek. But, people did have similar fears when the new technology of the "horseless carriage" made its initial debut. Those fears turned out to be baseless. However, that new technology did put people out of work. How many farrier's and carriage makers do you see around today?  Over our heads right now there are thousands of airplanes flying themselves. It's only a matter of time before automation takes over more and more aspects of our lives and the labor force. What needs to happen now are discussions about how we take care of people that currently have jobs that simply wont exist in the future. If machines do the labor of a large low skilled work force, what do those millions of people do now? Do governments allow this automation? If they do allow automation on a mass scale do we implement a universal basic income? The list of questions goes on and on. 

Yes, airplanes are flying themselves to some extent, but they do have pilots on board, in case something goes wrong. They don't take off or land themselves.

Trains are really no different. They have to be made up and broken down, in order to get their loads to their destinations. You're just not going to send hazmat loads out there unattended. No matter how hard the railroads try, there will always be a human on board.

J 611 posted:

I take your response as tongue in cheek. But, people did have similar fears when the new technology of the "horseless carriage" made its initial debut. Those fears turned out to be baseless. However, that new technology did put people out of work. How many farrier's and carriage makers do you see around today?  Over our heads right now there are thousands of airplanes flying themselves. It's only a matter of time before automation takes over more and more aspects of our lives and the labor force. What needs to happen now are discussions about how we take care of people that currently have jobs that simply wont exist in the future. If machines do the labor of a large low skilled work force, what do those millions of people do now? Do governments allow this automation? If they do allow automation on a mass scale do we implement a universal basic income? The list of questions goes on and on. 

But those fears did NOT turn out to be baseless:  the carnage on the roadways that was predicted has come to pass.  We have simply accepted the carnage as the cost of doing business.  (Over the years, some means have been found to reduce it, but more people die on the highways every year than probably died in "horse wrecks" for the total of recorded history.)

Those planes flying overhead have people at the controls if necessary.

Oddly enough, I believe there are more farriers in the county where I live than there were 150 years ago

But do we really want to be slaves to HAL and lose our humanity?

And some sort of "income" will make us all slaves of Big Brother.

The social disruption this will cause will make the Yellow Jacket situation in France look like a NHL fight on the ice.

Do people in Boardrooms think of these things?  Today, with email snd social media?  If the social framework goes down the sewer, they will, at best, be broke.  At worst dead.  History does repeat itself.  

Last edited by Dominic Mazoch
Dominic Mazoch posted:

But do we really want to be slaves to HAL and lose our humanity?

And some sort of "income" will make us all slaves of Big Brother.

The social disruption this will cause will make the Yellow Jacket situation in France look like a NHL fight on the ice.

Do people in Boardrooms think of these things?  Today, with email snd social media?  If the social framework goes down the sewer, they will, at best, be broke.  At worst dead.  History does repeat itself.  

HAL's already here, except she's called Alexa...

Rusty

Last edited by Rusty Traque
Dominic Mazoch posted:

But do we really want to be slaves to HAL and lose our humanity?

And some sort of "income" will make us all slaves of Big Brother.

The social disruption this will cause will make the Yellow Jacket situation in France look like a NHL fight on the ice.

Do people in Boardrooms think of these things?  Today, with email snd social media?  If the social framework goes down the sewer, they will, at best, be broke.  At worst dead.  History does repeat itself.  

Amen. 

Yes, it comes from the boardrooms and - dare I say, as a University employee - from the Ivory Towers. We help industry drive out costs from the equation and sadly that begins with what is usually the second largest line-item: Wages and Salaries.

J 611 posted:
Guitarmike posted:
Big_Boy_4005 posted:
J 611 posted:
Byrdie posted:

Just a thought provoking question - how many of you use the self checkout at the many, many stores that are now implementing them?

I use it all the time. If you only have a few items with bar codes its much faster and more convenient than waiting for a cashier. Automation is coming. What we have to figure out is how to take care of the millions that will be out of their current jobs. Truck drivers are on the front lines here in my opinion. It will be interesting to see how society will react. How governments will react. This will be a much larger change than when we went from a mostly agrarian work force to an industrialized one. The next 50 to 100 years will be very interesting. 

Well, when things get screwed up at the self checkout, guess what, the human comes over to fix it.

Same for trains and just about everything else.

Just wait, the self checkout in the McDonald's in the train station will get screwed up and the autopilot trains will crash through the bumpers at the end of the tracks in the station while asking you if you want your receipt.....

I take your response as tongue in cheek. But, people did have similar fears when the new technology of the "horseless carriage" made its initial debut. Those fears turned out to be baseless. However, that new technology did put people out of work. How many farrier's and carriage makers do you see around today?  Over our heads right now there are thousands of airplanes flying themselves. It's only a matter of time before automation takes over more and more aspects of our lives and the labor force. What needs to happen now are discussions about how we take care of people that currently have jobs that simply wont exist in the future. If machines do the labor of a large low skilled work force, what do those millions of people do now? Do governments allow this automation? If they do allow automation on a mass scale do we implement a universal basic income? The list of questions goes on and on. 

Universal Basic Income = Unemployment Benefit.

As usual, we are simply resorting to calling it something else to make more, ahem, palatable.  My father, in all his years in Hollywood, never once took unemployment. The Greatest Generation saw it as a handout (although, to be fair, it really wasn't). 

Rusty Traque posted:

HAL's already here, except she's called Alexa...

Rusty

My wife bought one. I nicknamed it Voldemort so we could talk about it "behind its back". Last night I finally unplugged the stupid thing.

There's a great short story by Ray Bradbury called "The Murderer" about a guy in a futuristic world plagued by automation, who takes out his frustration on the modern "conveniences".

Are we there yet?

Rule292 posted:

We've HAD automated railroads since the early 1970's, the BM&LP. 

And both BART and the DC Metro are automated, with the "engineer" acting as an attendant when in automatic train operation mode.  Which is the normal method of operation. 

Not an argument for or against, just an example of the fact that the basic technological issues of starting, running and stopping trains have been solved 40 years ago. 

What to do when all of our jobs are replaced by automation is the government's problem since the biggest losers will be those whose livelihood depend on taxes. 

The Washington DC area Metrorail system has actually been operating in manual mode since the 2009 accident at Fort Totten where the automation failed (the system failed to detect the presence of a train in the block ahead), eight people were killed and eighty were injured.  Information on more recent developments in this situation is at: https://wamu.org/story/17/06/0...ncies-worker-stress/

I think a better example of early automation is the Port Authority Transit Corporation's (PATCO's) Lindenwold Line from Philadelphia to Lindenwold, NJ.  

Thete is something else no one is thinging about:  electrons.

A ham friend of mine ordered some burgers via the drive in.  Had a 100 watt HF/shortwave radio in the car.  Hit mike.  All the electronics in the store went crazy.....!

Lightning, radio waves, EMP, spark gap generator, solar storms..... can knock out the radio communications system and the electronics itself.  In the 1850's there was a solar storm so strong parts of the telegraph grid did not need human produced electricity of operate.  A few WU stations actually caught fire because of tje power of the solar storm.  One can go from Jetsons to Flintstones in a blink of an eye.

Plus radio signals, electronics, and programming can be jammed or hacked.

One day Engineer David is on a train, and does sometimg which needed to be done.  And the controlling computer days, "I cannot allow you to do that, Dave...."  

Last edited by Dominic Mazoch
Big_Boy_4005 posted:

Yes, airplanes are flying themselves to some extent, but they do have pilots on board, in case something goes wrong. They don't take off or land themselves.

Trains are really no different. They have to be made up and broken down, in order to get their loads to their destinations. You're just not going to send hazmat loads out there unattended. No matter how hard the railroads try, there will always be a human on board.

They actually could take off and land themselves, that technology has been around for a long time, and in many cases might make for more efficient operations, especially in bad weather conditions, but there are political and other reasons why it isn't done, and I suspect even if they automate that they would keep a pilot on board as backup, the way the BART system has an engineer as a backup to the automated system that  runs it (or at least they used to).  I don't know if trains will ever be fully automated, it is going to depend on context, much the same way that in crowded harbors a pilot takes over for the captain when bringing a big ship in, in denser areas at the very least there would be someone on board.

 

LaramieJoe posted:
Dominic Mazoch posted:

But do we really want to be slaves to HAL and lose our humanity?

And some sort of "income" will make us all slaves of Big Brother.

The social disruption this will cause will make the Yellow Jacket situation in France look like a NHL fight on the ice.

Do people in Boardrooms think of these things?  Today, with email snd social media?  If the social framework goes down the sewer, they will, at best, be broke.  At worst dead.  History does repeat itself.  

Amen. 

Yes, it comes from the boardrooms and - dare I say, as a University employee - from the Ivory Towers. We help industry drive out costs from the equation and sadly that begins with what is usually the second largest line-item: Wages and Salaries.

Depends on the ivory tower, the sociologists and such have been worried about this for a long time....on the other hand, the business schools love the idea of getting rid of labor in the name of driving profits......and today where everything is based around stock price and also in who owns large blocks of stock and their goals , it has sent this into hyperdrive. It isn't new, when mass production made the cost of production low, the owners of those newly efficient factories paid their workers less and less, often cutting salaries to make more money in good times, then wondered why the economy routinely cratered (such a situation was one of the problems with the 1920's and the crash that followed) and the parallel to today, none of them seemed to ask themselves the question, who would buy their product? They want a world where labor costs are minimized but assume that someone will still be there to buy their product, and if in fact we are heading to a world where large swaths of blue collar and white collar jobs are gone, where more and more is produced by less and less people, what happens?

 

In the past, automation replaced older jobs but created more jobs that didn't exist, the problem is now the automation that is replacing older jobs is also stopping the creation of new jobs since automation is taking them before they are created, pure and simple.  Kurt Vonnegut tackled this in his book "Player Piano" almost 60 years ago, in his world if you weren't one of the managers, you either worked in the army or a giant public works kind of thing.....thing is, most people want to be useful, to do things and earn what they get, and no one I have seen has a clue how to make sure people can have a decent living, feel like they are doing something useful, in the kind of world we see happening, the real question (to me) isn't who will do the work, but rather, in a world where automation takes over, what will drive and motivate people? I doubt the answer would be, as some propose, to go backwards, forbid automation and the like, just not practical, I suspect the real answer is finding things for people to do that leaves them fulfilled and allows them to be paid for it without it being traditional employment...and don't have a clue how that will happen. 

palallin posted:
J 611 posted:

I take your response as tongue in cheek. But, people did have similar fears when the new technology of the "horseless carriage" made its initial debut. Those fears turned out to be baseless. However, that new technology did put people out of work. How many farrier's and carriage makers do you see around today?  Over our heads right now there are thousands of airplanes flying themselves. It's only a matter of time before automation takes over more and more aspects of our lives and the labor force. What needs to happen now are discussions about how we take care of people that currently have jobs that simply wont exist in the future. If machines do the labor of a large low skilled work force, what do those millions of people do now? Do governments allow this automation? If they do allow automation on a mass scale do we implement a universal basic income? The list of questions goes on and on. 

But those fears did NOT turn out to be baseless:  the carnage on the roadways that was predicted has come to pass.  We have simply accepted the carnage as the cost of doing business.  (Over the years, some means have been found to reduce it, but more people die on the highways every year than probably died in "horse wrecks" for the total of recorded history.)

Those planes flying overhead have people at the controls if necessary.

Oddly enough, I believe there are more farriers in the county where I live than there were 150 years ago

You can't make a comparison like that, that is like saying "less people die in Montana in a given year then die in a given week in NYC"...given that NYC's population is like 20 times that of Montana the state. In the day of the horse and carriage, a relatively few people got around that way, most people spent most of their time walking likely, for longer distance they took trains or maybe a carriage for hire..and it showed. Most people back in the pre car day lived in a world where they spent most of their lives within 5 or 10 miles of where they were born, and if you factored the death rate of travel by horse and carriage versus the relatively few miles people travelled, would be an interesting study (yep, 35,000 people die on the roads each year in the US, but that is travelling 3.2 trillion miles a year! Another thing to think of, back in the days of the horse and carriage, how many people died because of the lack of fast transportation? How many people died because they couldn't get to a doctor in time, because of some other characteristic, something today routinely because of ambulances and good roods, doesn't happen? All technology changes bring consequences, for the good and bad, and in the end you hope a)the good is greater than the bad and b)there are people who care enough to try and minimize the bad consequences, rather than 'seeing it as the cost of doing business', whether it is those who fought for the safety standards on cars that have made them light years more safe than cars were in the 1950's or those who are asking questions about what to do with the people left behind, rather than those who shrug their shoulders and say "that is life, there are winners and losers" and walk away, then wonder why people get angry. 

bigkid posted:
Big_Boy_4005 posted:

Yes, airplanes are flying themselves to some extent, but they do have pilots on board, in case something goes wrong. They don't take off or land themselves.

Trains are really no different. They have to be made up and broken down, in order to get their loads to their destinations. You're just not going to send hazmat loads out there unattended. No matter how hard the railroads try, there will always be a human on board.

They actually could take off and land themselves, that technology has been around for a long time, and in many cases might make for more efficient operations, especially in bad weather conditions, but there are political and other reasons why it isn't done, and I suspect even if they automate that they would keep a pilot on board as backup, the way the BART system has an engineer as a backup to the automated system that  runs it (or at least they used to).  I don't know if trains will ever be fully automated, it is going to depend on context, much the same way that in crowded harbors a pilot takes over for the captain when bringing a big ship in, in denser areas at the very least there would be someone on board.

 

 

As a former airline captain, I disagree.  The automation is there to assist the pilots with mundane tasks (believe it or not, actually flying the aircraft) so they can better make decisions, which no machine will ever be able to do.  It is not there to replace the pilots.  An automated aircraft could not even get off the gate because the crew makes countless decisions that the passengers never see.

We normally only used our autoland system when the visibility was very low (around 600 feet) to allow the pilot not flying to concentrate on looking for the visual cues necessary to continue the approach while the pilot "flying" monitored the instruments and the autopilot system.  The autoland system had severe crosswind limitations (around 10 knots) while the normal crosswind limitation was 39 knots for pilot-flown landings.  Additionally, I am not aware of any aircraft that can accomplish an automated takeoff.  If you have an engine failure or fire, blown tire, runway incursion, or any other issue during takeoff, the captain will decide very quickly whether to abort or continue the takeoff.  A machine cannot make those types of decisions--they require thought, experience, and situational awareness.

The process of simply holding an altitude and navigating between GPS defined waypoints is what most people think about when this discussion arises.  There is far more to it than that.  Similarly, the process of operating a train from point A to point B is only a minor part of the entire process.

PGentieu posted:
Rule292 posted:

We've HAD automated railroads since the early 1970's, the BM&LP. 

And both BART and the DC Metro are automated, with the "engineer" acting as an attendant when in automatic train operation mode.  Which is the normal method of operation. 

Not an argument for or against, just an example of the fact that the basic technological issues of starting, running and stopping trains have been solved 40 years ago. 

What to do when all of our jobs are replaced by automation is the government's problem since the biggest losers will be those whose livelihood depend on taxes. 

The Washington DC area Metrorail system has actually been operating in manual mode since the 2009 accident at Fort Totten where the automation failed (the system failed to detect the presence of a train in the block ahead), eight people were killed and eighty were injured.  Information on more recent developments in this situation is at: https://wamu.org/story/17/06/0...ncies-worker-stress/

I think a better example of early automation is the Port Authority Transit Corporation's (PATCO's) Lindenwold Line from Philadelphia to Lindenwold, NJ.  

I actually forgot about the Lindenwold Line.  I was part of a system tour in the 1970s which included the dispatching center as well as a pretty extensive shop tour. 

WMATA has a lot of issues and money isn't necessarily at the root of it.   I remember watching one of the "motormen" sneak his girlfriend into the cab in the 80s when I worked in the area.    Most of the rides were boring but on one occasion (in the days before the windows allowing you to see into the cab and watch the displays were blacked out) on a weekend run to Shady Grove we were treated to a good ride with a good "engineer" who was authorized to operate his run manually.  And that he did!     In any event, much of the Metro was obsolete before it started due to the limits of two tracks.  

Not unlike the current conundrum of the Broadway and the Hudson River tunnels. 

Anyway, automation is coming.  How well some of us (ME included) survive it remains to be seen.

 

Last edited by Rule292
IC EC posted:
bigkid posted:
Big_Boy_4005 posted:

Yes, airplanes are flying themselves to some extent, but they do have pilots on board, in case something goes wrong. They don't take off or land themselves.

Trains are really no different. They have to be made up and broken down, in order to get their loads to their destinations. You're just not going to send hazmat loads out there unattended. No matter how hard the railroads try, there will always be a human on board.

They actually could take off and land themselves, that technology has been around for a long time, and in many cases might make for more efficient operations, especially in bad weather conditions, but there are political and other reasons why it isn't done, and I suspect even if they automate that they would keep a pilot on board as backup, the way the BART system has an engineer as a backup to the automated system that  runs it (or at least they used to).  I don't know if trains will ever be fully automated, it is going to depend on context, much the same way that in crowded harbors a pilot takes over for the captain when bringing a big ship in, in denser areas at the very least there would be someone on board.

 

 

As a former airline captain, I disagree.  The automation is there to assist the pilots with mundane tasks (believe it or not, actually flying the aircraft) so they can better make decisions, which no machine will ever be able to do.  It is not there to replace the pilots.  An automated aircraft could not even get off the gate because the crew makes countless decisions that the passengers never see.

We normally only used our autoland system when the visibility was very low (around 600 feet) to allow the pilot not flying to concentrate on looking for the visual cues necessary to continue the approach while the pilot "flying" monitored the instruments and the autopilot system.  The autoland system had severe crosswind limitations (around 10 knots) while the normal crosswind limitation was 39 knots for pilot-flown landings.  Additionally, I am not aware of any aircraft that can accomplish an automated takeoff.  If you have an engine failure or fire, blown tire, runway incursion, or any other issue during takeoff, the captain will decide very quickly whether to abort or continue the takeoff.  A machine cannot make those types of decisions--they require thought, experience, and situational awareness.

The process of simply holding an altitude and navigating between GPS defined waypoints is what most people think about when this discussion arises.  There is far more to it than that.  Similarly, the process of operating a train from point A to point B is only a minor part of the entire process.

No argument on the complexity of flying a plane ,especially in difficult or emergency situations,right now autonomous vehicles have real problems with handling the unexpected. However, AI and machine learning are growing exponentially, they are doing things like a system that can infer a solution from a single point of data,something once thought impossible (in a human being they call that intuition).  With that,it isn't far off when an automated system could handle it better than a human being,among other things no emotions to get in the way,like panic...could also be found out to be impractical but i suspect not.

As far as automation creating new jobs, that is to be blunt self serving propoganda based on the past. The amount of new jobs created supporting automation are a small fraction of what they replace, an automated system to manage financial portfolios can be written and supported by a small team of people,replacing many more people,the analysts,portfolio managers and traders that traditionally did the work. With machine learning the system can improve its own operation,requiring even less people.

There are problems that cannot be solved by automation, that was proven about 80 yrs ago by Alan Touring, but the reality is that that limitation applies in a relatively small class of problems,the rest no matter how complex can eventually be solved.

 

bigkid posted:
IC EC posted:
bigkid posted:
Big_Boy_4005 posted:

Yes, airplanes are flying themselves to some extent, but they do have pilots on board, in case something goes wrong. They don't take off or land themselves.

Trains are really no different. They have to be made up and broken down, in order to get their loads to their destinations. You're just not going to send hazmat loads out there unattended. No matter how hard the railroads try, there will always be a human on board.

They actually could take off and land themselves, that technology has been around for a long time, and in many cases might make for more efficient operations, especially in bad weather conditions, but there are political and other reasons why it isn't done, and I suspect even if they automate that they would keep a pilot on board as backup, the way the BART system has an engineer as a backup to the automated system that  runs it (or at least they used to).  I don't know if trains will ever be fully automated, it is going to depend on context, much the same way that in crowded harbors a pilot takes over for the captain when bringing a big ship in, in denser areas at the very least there would be someone on board.

 

 

As a former airline captain, I disagree.  The automation is there to assist the pilots with mundane tasks (believe it or not, actually flying the aircraft) so they can better make decisions, which no machine will ever be able to do.  It is not there to replace the pilots.  An automated aircraft could not even get off the gate because the crew makes countless decisions that the passengers never see.

We normally only used our autoland system when the visibility was very low (around 600 feet) to allow the pilot not flying to concentrate on looking for the visual cues necessary to continue the approach while the pilot "flying" monitored the instruments and the autopilot system.  The autoland system had severe crosswind limitations (around 10 knots) while the normal crosswind limitation was 39 knots for pilot-flown landings.  Additionally, I am not aware of any aircraft that can accomplish an automated takeoff.  If you have an engine failure or fire, blown tire, runway incursion, or any other issue during takeoff, the captain will decide very quickly whether to abort or continue the takeoff.  A machine cannot make those types of decisions--they require thought, experience, and situational awareness.

The process of simply holding an altitude and navigating between GPS defined waypoints is what most people think about when this discussion arises.  There is far more to it than that.  Similarly, the process of operating a train from point A to point B is only a minor part of the entire process.

No argument on the complexity of flying a plane ,especially in difficult or emergency situations,right now autonomous vehicles have real problems with handling the unexpected. However, AI and machine learning are growing exponentially, they are doing things like a system that can infer a solution from a single point of data,something once thought impossible (in a human being they call that intuition).  With that,it isn't far off when an automated system could handle it better than a human being,among other things no emotions to get in the way,like panic...could also be found out to be impractical but i suspect not.

As far as automation creating new jobs, that is to be blunt self serving propoganda based on the past. The amount of new jobs created supporting automation are a small fraction of what they replace, an automated system to manage financial portfolios can be written and supported by a small team of people,replacing many more people,the analysts,portfolio managers and traders that traditionally did the work. With machine learning the system can improve its own operation,requiring even less people.

There are problems that cannot be solved by automation, that was proven about 80 yrs ago by Alan Touring, but the reality is that that limitation applies in a relatively small class of problems,the rest no matter how complex can eventually be solved.

 

 

I still don't understand the ultimate goal of this level of automation.  Is it

1. to make travel safer?  U. S. airlines have had zero fatalities for several years now.  It's hard to improve upon that record.  I have also witnessed aircraft automation make several mistakes that required immediate crew intervention.  See my post in the PTC thread.

The city of Scottsdale, Arizona, recently stopped allowing the testing of "driverless" cars on its streets due to a fatal accident.

My PC is fairly reliable, yet it still crashes occasionally or locks up.  Not a good thing if this happens to the flight control system during an approach to minimums.

2. to make travel cheaper?  I would argue that the cost to equip a single airliner would likely exceed the lifetime earnings of both pilots.  Additionally, a single engineering salary for the developmental staff  is not that far off the average airline pilot salary, not to mention salaries for the technicians that have to maintain it.  Where's the savings?

IC EC posted:
bigkid posted:
IC EC posted:
bigkid posted:
Big_Boy_4005 posted:

Yes, airplanes are flying themselves to some extent, but they do have pilots on board, in case something goes wrong. They don't take off or land themselves.

Trains are really no different. They have to be made up and broken down, in order to get their loads to their destinations. You're just not going to send hazmat loads out there unattended. No matter how hard the railroads try, there will always be a human on board.

They actually could take off and land themselves, that technology has been around for a long time, and in many cases might make for more efficient operations, especially in bad weather conditions, but there are political and other reasons why it isn't done, and I suspect even if they automate that they would keep a pilot on board as backup, the way the BART system has an engineer as a backup to the automated system that  runs it (or at least they used to).  I don't know if trains will ever be fully automated, it is going to depend on context, much the same way that in crowded harbors a pilot takes over for the captain when bringing a big ship in, in denser areas at the very least there would be someone on board.

 

 

As a former airline captain, I disagree.  The automation is there to assist the pilots with mundane tasks (believe it or not, actually flying the aircraft) so they can better make decisions, which no machine will ever be able to do.  It is not there to replace the pilots.  An automated aircraft could not even get off the gate because the crew makes countless decisions that the passengers never see.

We normally only used our autoland system when the visibility was very low (around 600 feet) to allow the pilot not flying to concentrate on looking for the visual cues necessary to continue the approach while the pilot "flying" monitored the instruments and the autopilot system.  The autoland system had severe crosswind limitations (around 10 knots) while the normal crosswind limitation was 39 knots for pilot-flown landings.  Additionally, I am not aware of any aircraft that can accomplish an automated takeoff.  If you have an engine failure or fire, blown tire, runway incursion, or any other issue during takeoff, the captain will decide very quickly whether to abort or continue the takeoff.  A machine cannot make those types of decisions--they require thought, experience, and situational awareness.

The process of simply holding an altitude and navigating between GPS defined waypoints is what most people think about when this discussion arises.  There is far more to it than that.  Similarly, the process of operating a train from point A to point B is only a minor part of the entire process.

No argument on the complexity of flying a plane ,especially in difficult or emergency situations,right now autonomous vehicles have real problems with handling the unexpected. However, AI and machine learning are growing exponentially, they are doing things like a system that can infer a solution from a single point of data,something once thought impossible (in a human being they call that intuition).  With that,it isn't far off when an automated system could handle it better than a human being,among other things no emotions to get in the way,like panic...could also be found out to be impractical but i suspect not.

As far as automation creating new jobs, that is to be blunt self serving propoganda based on the past. The amount of new jobs created supporting automation are a small fraction of what they replace, an automated system to manage financial portfolios can be written and supported by a small team of people,replacing many more people,the analysts,portfolio managers and traders that traditionally did the work. With machine learning the system can improve its own operation,requiring even less people.

There are problems that cannot be solved by automation, that was proven about 80 yrs ago by Alan Touring, but the reality is that that limitation applies in a relatively small class of problems,the rest no matter how complex can eventually be solved.

 

 

I still don't understand the ultimate goal of this level of automation.  Is it

1. to make travel safer?  U. S. airlines have had zero fatalities for several years now.  It's hard to improve upon that record.  I have also witnessed aircraft automation make several mistakes that required immediate crew intervention.  See my post in the PTC thread.

The city of Scottsdale, Arizona, recently stopped allowing the testing of "driverless" cars on its streets due to a fatal accident.

My PC is fairly reliable, yet it still crashes occasionally or locks up.  Not a good thing if this happens to the flight control system during an approach to minimums.

2. to make travel cheaper?  I would argue that the cost to equip a single airliner would likely exceed the lifetime earnings of both pilots.  Additionally, a single engineering salary for the developmental staff  is not that far off the average airline pilot salary, not to mention salaries for the technicians that have to maintain it.  Where's the savings?

Primarily cost, the cost of automated flight systems would in the end be less than human pilots, airplanes already have sophisticated flight systems, the fly by wire systems on control surfaces for example. As far as reliability goes, most modern military aircraft are basically aerodynamically unstable,they cannot be flown without the fly by wire system, so they have multiple levels of redundancy. Aircraft flight systems are not a pc, and comparisons don't hold. A truly automated piloting system would be specific to that application (won't be running windows 11 on a dell pc), would have levels of testing and certification that a pc could only dream of, upgrade procedures a lot more rigid than commercial software, would require redundancy, maintenance procedures requiring replacement every x hours and the unit taken out and throughly tested before reuse, this is standard w regular aviation components,would be stricter w autonomous flight systems. 

Not saying fully automated flight would be practical this moment, saying it will be someday in the not near future. Automated cars likewise right now have problems, but if you think they aren't going to happen read up on what the automakers are doing, they are either buying companies developing those systems or investing heavily in doing it themselves.

In the end that kind of automation will save money, one of the things you have to remember is maintenance functions are heavily automated, and newer generations are more and more able to fix themselves.  Eventually air traffic control systems will require few people as well.

It isn't that I want this to happen, I know only too well that economic forces are pushing to automate as many jobs as possible in the guise of efficiency but few are truly figuring out how to handle the consequences,it already is happening at an accelerating pace.

 

AmeenTrainGuy posted:
IC EC posted:
bigkid posted:
IC EC posted:
bigkid posted:
Big_Boy_4005 posted:
 

 

I still don't understand the ultimate goal of this level of automation.  Is it

1. to make travel safer?  U. S. airlines have had zero fatalities for several years now.  It's hard to improve upon that record.  I have also witnessed aircraft automation make several mistakes that required immediate crew intervention.  See my post in the PTC thread.

The city of Scottsdale, Arizona, recently stopped allowing the testing of "driverless" cars on its streets due to a fatal accident.

My PC is fairly reliable, yet it still crashes occasionally or locks up.  Not a good thing if this happens to the flight control system during an approach to minimums.

The computer they use on airplanes are not your average PC. An average consumer PC has 16gb of RAM and an i7 processor which would not be enough processing power to even begin to run the automation software.

2. to make travel cheaper?  I would argue that the cost to equip a single airliner would likely exceed the lifetime earnings of both pilots.  Additionally, a single engineering salary for the developmental staff  is not that far off the average airline pilot salary, not to mention salaries for the technicians that have to maintain it.  Where's the savings?

 

Which proves the point--if a relatively simple PC running Windows crashes, why would a much more complex system be any more reliable?  Additionally, complexity does not equate to capability; the Apollo Guidance Computer (AGC) used for navigation during the Apollo program had 64K of memory and ran at 0.043 MHz (Computer Weekly).

IC EC posted:
AmeenTrainGuy posted:
IC EC posted:
bigkid posted:
IC EC posted:
bigkid posted:
Big_Boy_4005 posted:
 

 

I still don't understand the ultimate goal of this level of automation.  Is it

1. to make travel safer?  U. S. airlines have had zero fatalities for several years now.  It's hard to improve upon that record.  I have also witnessed aircraft automation make several mistakes that required immediate crew intervention.  See my post in the PTC thread.

The city of Scottsdale, Arizona, recently stopped allowing the testing of "driverless" cars on its streets due to a fatal accident.

My PC is fairly reliable, yet it still crashes occasionally or locks up.  Not a good thing if this happens to the flight control system during an approach to minimums.

The computer they use on airplanes are not your average PC. An average consumer PC has 16gb of RAM and an i7 processor which would not be enough processing power to even begin to run the automation software.

2. to make travel cheaper?  I would argue that the cost to equip a single airliner would likely exceed the lifetime earnings of both pilots.  Additionally, a single engineering salary for the developmental staff  is not that far off the average airline pilot salary, not to mention salaries for the technicians that have to maintain it.  Where's the savings?

 

Which proves the point--if a relatively simple PC running Windows crashes, why would a much more complex system be any more reliable?  Additionally, complexity does not equate to capability; the Apollo Guidance Computer (AGC) used for navigation during the Apollo program had 64K of memory and ran at 0.043 MHz (Computer Weekly).

The answer is you are comparing apples to oranges. A windows pc is pretty much off the shelf hardware,there is nothing exotic in it per se, but windows itself is complex. Where it gets crazy is pc's support a wide variety if hardware component (video boards, sound drivers, hard drives), you have a wide variety if applications that can run on it,including support for older programs, and there are a lot of ill written programs out there written for windows..an OS like windows has to try and run all those, and the permutatiins are ridiculous...and a pc is a 500 buck or so piece of gear w windows installed, so microsoft,while it spends time and resources w testing, can't hope to pick up all the gotchas (it is why windows reports back crashes,then they patch it)

A system designed to fly an airliner is not general purpose,you are talking a system w likely a customized operating system (if not completely custom, you likely would want a realtime OS or something close to it) with specific applications written to it, the whole thing is designed to run as one thing.  

And given the nature of how critucal it is, regulations and yes fear of consequences means that unlike a pc they spend years and millions of dollars making sure it doesn't crash,testing it in ways microsoft couldn't do, they spend millions writing programs to test such a system alone, not to mention flight testing. Put it thus way,if you had a business and your pc crashed,causing you some kind of loss,good luck trying to sue microsoft, they would laugh you out of court (that and warrantee that basically only requires ms to fix the problem). You simply can't compare them, both are computers but after that,light years different, like comparing a Yugo to a Toyota or other modern car.

While not on the same order of magnitude as a self flying system, the sophisticated control systems in a car don't fail much either,yet that is a full bore computer system that these days is usingg 64 bit technology (and may require soon a 48 volt power system). 

 

 

 

bigkid posted:

The answer is you are comparing apples to oranges. A windows pc is pretty much off the shelf hardware,there is nothing exotic in it per se, but windows itself is complex. Where it gets crazy is pc's support a wide variety if hardware component (video boards, sound drivers, hard drives), you have a wide variety if applications that can run on it,including support for older programs, and there are a lot of ill written programs out there written for windows..an OS like windows has to try and run all those, and the permutatiins are ridiculous...and a pc is a 500 buck or so piece of gear w windows installed, so microsoft,while it spends time and resources w testing, can't hope to pick up all the gotchas (it is why windows reports back crashes,then they patch it)

A system designed to fly an airliner is not general purpose,you are talking a system w likely a customized operating system (if not completely custom, you likely would want a realtime OS or something close to it) with specific applications written to it, the whole thing is designed to run as one thing.  

And given the nature of how critucal it is, regulations and yes fear of consequences means that unlike a pc they spend years and millions of dollars making sure it doesn't crash,testing it in ways microsoft couldn't do, they spend millions writing programs to test such a system alone, not to mention flight testing. Put it thus way,if you had a business and your pc crashed,causing you some kind of loss,good luck trying to sue microsoft, they would laugh you out of court (that and warrantee that basically only requires ms to fix the problem). You simply can't compare them, both are computers but after that,light years different, like comparing a Yugo to a Toyota or other modern car.

While not on the same order of magnitude as a self flying system, the sophisticated control systems in a car don't fail much either,yet that is a full bore computer system that these days is usingg 64 bit technology (and may require soon a 48 volt power system).

You know, all it really takes is one faulty sensor to tell the computer the wrong condition on which to act. Without human override, disaster could happen.

There was crash in the Java Sea, not that long ago, in which even the pilots couldn't recover from the stall the computer had created, due to a bad sensor.

Pilots have since been given additional training to learn to overcome this glitch. The solution was simple, but at the time the pilots didn't know, and fought the computer, and the computer won. The sensor was probably not an expensive part, but it ended up costing 189 people their lives.

Now you think it's an OK idea to send trains loaded with chemicals careening blindly through the countryside? Yikes!

Have you ever been at the throttle of a train? I have, and it scared the crap out of me. It was an old diesel switcher, with three loaded passenger cars in tow. The run was downgrade all the way back to the station. I felt a little bad for those folks, because they didn't get the smoothest ride. I was a little rough on the brakes. Oh, but it got better, as we were approaching the station, a car suddenly turned in front of us and crossed the tracks. I turned white. I was happy when that day ended and nobody died.

The guys who operate trains are true professionals, and know how to do it safely. The railroads, in their quest for the almighty dollar, should never be allowed to replace them with technology.

I actually predicted that it would be the air traffic controllers who would be the ones to force the end to the government shutdown. They have their hands on the throat of the country, and it just took a small squeeze on NYC to prove it. The shutdown ended within hours. They need all the technology they can get, but we still need them and probably always will.

clem k posted:

Question :  Out of all the means of transport, airplanes, cars, trucks, boats, and rail.  Seems like rail would be the easiest to automate safely ? 

One would think, but looking at everyone's responses to this post I am concluding that automation with human oversight (in the cab) is essential. Insurance companies, for their part, will want to know how risk is mitigated.

I don't think that we will see totally automated trains, planes or trucks in widespread use, likely the first phase will be like the rules on driverless cars, where a person has to be in the drivers seat and aware, read to take over if something happens (it is why I doubt you will see Uber and the like with driverless fleets anytime soon, you can't expect a passenger  to sit in the driver's seat and take over, might not even be licensed to drive). The first thing likely would be where the automated flight systems are tested with both pilots on board on a trial basis, where they could take over if something happens, then likely it would be single pilot as backup to the automated system for (a relatively) long while. Same with trucks and trains, although it will happen on trains in specific applications a lot sooner (like the freight route in australia going through desolate areas, or something like perhaps even industrial switching, which is not unlike the automated systems they are putting in warehouses and the like). With planes, it could be cargo flights would be the first test of this, where they might have operators on the ground ready to take over for a plane in distress operating it remotely, the way they do drones today, could be the same with passenger flights. Not to mention there obviously is going to take time for the public to get used to the idea of driverless vehicles, in the early days of commercial travel it took a long, long time before people felt safe flying, in the early days of railroads likewise a lot of people were afraid to ride them. 

And just to be contrary, no system is foolproof, automated or human, airline travel has been extremely safe for a long time (well, outside of the conditions on the planes, being stuck on the tarmac for hours, being crowded in like sardines in a can, etc), but obviously when accidents happen it is horrible because it involves hundreds of people. Computer systems, no matter how well designed and tested, are going to have problems in them, and for example something simple like a sensor going bad could cause what we saw  with the airliner (me, personally, that is a sign of bad design, sensors should be constantly monitored and tested, even on cars sensors are checked all the time by the ECU, and on something as critical as aircraft should have multiple sensors, where they check against each other and agree (this is done routinely on military aircraft, spacecraft and the like..the Shuttle had literally 8 levels of backup computer, and multiple levels of sensor input for redundancy), so that was bad design...but bad design happens all the time that leads to crashes, anyone remember the engines on the DC10 falling off the plane, the batteries catching fire on the 757, the engines on one of the airbuses where the turbine blades failed and took out the engines.....

The other thing to keep in mind with piloted aircraft is that a lot of accidents have been caused by human error, and over the years they have changed procedures and policy and trained pilots so they don't happen again, same thing will happen with automated flight. In the 1970's, in an analog to the sensor that reputedly caused the plane's autopilot to crash the system,an indicator bulb was blown out that indicated whether the landing gear had deployed. The flight crew (they had 3 in the cockpit back then), all obsessed about the bulb being out, and while one of the flight crew went to look below deck on the plane to see if the landing gear was out visually, the other two obsessed about the bulb..and the plane ended up crashing in the everglades because it was descending rapidly and by the time they noticed, it was too late. Cultural issues on some airlines caused crashes, where a pilot was doing something wrong and the other pilot wouldn't say anything, because the pilot was senior to him,  pilots have accidentally turned off the auto pilot and not realized it and planes ended up doing all kinds of nasty things......every time this has happened, the accident is analyzed and the airlines and regulators come up with ways to prevent it, and it shows, the crash/accident rate for aircraft globally is at an all time low, even notorious airlines like KAL have cleaned up their act, the same thing will happen with automated technology, to argue that automated systems aren't perfect is leaving out that a lot of the accidents with manned vehicles and such are caused by human error, because humans aren't perfect either. Could an automated piloting system do what Captain Sully did, or the guy who landed the United plane in Iowa where the steering system had pretty much failed, using engines alone? Not sure, but I would also argue that likely a large percentage of pilots faced with the same situation might not do so well *shrug*.

And again, I am not pushing this kind of thing, I don't know how I would feel if I was on a plane with no pilot on board (I would hope they had remote control capability in case of a problem), and I am not pushing to replace people, as much for the reason that IMO no one has figured out what will replace the jobs lost if in fact this happens as for questions about the safety, but I also am telling you that the kind of technology out there, with things like machine learning that is advancing at a rapid pace, this kind of thing is likely to happen more and more. 

 

 

 

 

 

 

LaramieJoe posted:
clem k posted:

Question :  Out of all the means of transport, airplanes, cars, trucks, boats, and rail.  Seems like rail would be the easiest to automate safely ? 

One would think, but looking at everyone's responses to this post I am concluding that automation with human oversight (in the cab) is essential. Insurance companies, for their part, will want to know how risk is mitigated.

Actually Doc, if I'm not mistaken, the railroads self insure. At least the big ones do. If that's the case, it presents a real disincentive to risk full automation.

I still find the whole proposition highly impractical for North America, given the geography, the weather and the commodities being hauled. But then I've been going on about that for two pages now.

Elliott, I think you are right. They do self-insure. Still, they are required to buy coverage atop what they can cover by themselves from so-called "re-insurers."

@bigkid .. I think you will enjoy this article, which came out today. More to the point of automated systems not yet ready for prime time: 

Lion Air crash shows cockpit computers are no substitute for pilot skills

From the Los Angeles Time, February 4, 2019

© LA Times, 2019

When an altitude sensor failed on a Turkish Airlines Boeing 737 flight to Amsterdam in 2009, the jetliner’s computerized flight controls erroneously cut the engine thrust. The pilots didn’t understand what happened in time to prevent a crash.

The accident had striking similarities to the recent Lion Air tragedy in Indonesia, which took the lives of 189 people. A failed sensor led flight computers to put the 737 MAX jetliner into a series of dives, based on the erroneous calculation that it was losing lift and about to stall. The crew didn’t diagnose the problem, which could have been remedied with the flip of a switch, and the plane fell into the Java Sea. The investigation into the crash is ongoing.

Aviation experts say automated systems have made planes safer than ever and are a major reason why crash rates have declined all over the world. The push to automate — which also reduces airlines’ training costs — is only growing stronger. Boeing has a research project in the works to develop a fully automated jetliner. A company spokesman said last month that testing was ongoing.

But automated flight systems are also implicated in a series of incidents in which they made the wrong decisions and pilots did not fully understand the complex software that adjusts flight controls constantly during automated takeoffs, landings and high-altitude cruising.

“A lot of the optimization that the computer is doing is not made clear to the pilot,” said Douglas Moss, an instructor at USC’s Viterbi Aviation Safety and Security Program. He is a former United Airlines captain and before that, an Air Force test pilot, as well as an attorney. “The pilot is sitting there for 10 or 15 seconds trying to figure out why the computer is pitching up the nose or adjusting the throttle. I can think of thousands of times when the autopilot or flight management system would do something that caught me by surprise. Almost always, it is the right thing to do, but it is the pilot who is responsible for the safety of the flight.”

The two accidents also highlight the potential risks of basing automated flight control decisions on readings by only two sensors — which can create uncertainty when one fails.

Boeing and Airbus, the dominant international manufacturers of large jetliners, declined to provide detailed information about how many sensors their aircraft models have for various critical measurements, such as altitude, airspeed and angle of attack, citing the ongoing Lion Air investigation.

But Federal Aviation Administration documents reviewed by The Times, along with interviews of union officials and aviation experts, point out that some, though not all, aircraft have three sensors for critical readings, allowing a computerized voting system to eliminate a discrepant sensor.

Aviation experts say the pilots’ authority, certainly outside of the U.S. and Western Europe, is being gradually encroached on by automated control systems that offer air carriers lower training costs and crew expenses in an increasingly competitive international industry. It has led to a decline in basic manual flying skills, the ability to use the stick, rudder and throttle to keep a plane at the correct speed, pitch and altitude, a wide range of safety experts say.

“Pilots are not being told or taught everything they need to know about their airplanes,” said Chesley Sullenberger, the renowned pilot who made an emergency landing on the Hudson River a decade ago that saved every person on board.

“It is not easier or cheaper or requires less training to fly an automated airplane. It frequently requires more, because you have to have a deep understanding of how a system works, including the dark corners, the counterintuitive things it might do in certain circumstances. Many foreign carriers are trying to take people with zero flying experience, put them in simulators and quickly put them in the right seat of a jetliner. They don’t have the experience, knowledge, skills and confidence to be the absolute master of the aircraft start to finish.”

“It is not easier or cheaper or requires less training to fly an automated airplane," he says. "It frequently requires more." (Todd Sumlin/MCT)

Dennis Tajer, an American Airlines captain and spokesman for the Allied Pilots Assn., agrees automated systems should result in more pilot training, not less.

“It makes the aircraft a bit more complicated, so understanding that, being able to utilize it and making it a part of your safety standard is critical,” he said. When a piece of equipment fails, it’s incumbent on the pilot to keep control of the plane, he said, adding, “stick and rudder skills save lives.”

The U.S. airline industry agrees, and Boeing doesn’t dispute the value of pilots.

Company executives have questioned where global airlines are going to get all the pilots needed to fly the planes that are on order and expected over coming decades.

As air travel increases rapidly around the world, many foreign carriers are coming to depend on automated controls to help flight crews that do not routinely have the deep experience, military background and intensive training that is common among major U.S. and Western European airlines. American experts are growing increasingly concerned that such crews are reluctant to fly aircraft manually and lack the skills necessary to intervene when computers make the wrong decisions.

Those concerns were at the forefront in the crash of an Asiana Airlines Boeing 777 at San Francisco in July 2013, which killed three and injured 187. The crew had inadvertently disengaged the auto throttle, which is akin to cruise control on an automobile, as it made its final approach. The airspeed and altitude varied widely over the prescribed glide slope and the plane’s landing gear and tail hit the concrete sea wall at the far edge of the runway. Federal investigators blamed the accident on improper speed and altitude control, noting that Asiana Airlines emphasizes the use of cockpit automation. As a result of similar policies, many crews from nations around the world have limited hands-on manual flying skills.

In the Turkish Airlines and Lion Air accidents, the pilots should have been able to manually fly the planes out of harm’s way, experts say.

In the Turkish Airlines flight, one of the aircraft’s two radar altimeters reported that the plane was eight feet below the ground, leading the computer to think it was about to land and triggering an automatic reduction in the throttle. In fact, the aircraft was at about 2,000 feet and the crew was trying to reduce airspeed on the approach. But the crew did not realize the power settings had been cut to idle until it was too late, and the plane crashed a mile short of the runway. Nine people were killed.

In the Lion Air accident, one of the plane’s two angle-of-attack sensors, which measure the angle at which the wings are moving through the air, failed on takeoff from Jakarta. The aircraft’s autopilot disregarded the good sensor and followed readings of the discrepant left side, or captain’s sensor. As a result, it triggered software meant to offset the aircraft’s tendency for the nose to pitch up.

Such a nose-high attitude can reduce lift and potentially stall the jetliner. But in this case, the software — the maneuvering characteristics augmentation system, or MCAS — was responding to wrong data and putting the plane into a dive. It is similar to what is known as a “runaway trim,” which can be caused by as many as five problems. Some experts say the crew should have known how to flip off the MCAS and manually fly the plane, which is what another crew had done the day before in the same plane. Instead, the captain repeatedly attempted to pull up the nose and never correctly diagnosed the problem.

“Automation complacency is a huge issue,” said Andrew Skow, founder of Tiger Century Aircraft, which develops cockpit safety systems, and a former Northrop Grumman chief engineer and flight test director. “If an automation system has control and is right most of the time, you become complacent.”

Skow said automation systems often fail to consider unintended effects, which occurred in the Turkish Airlines and Lion Air accidents. He also faults Boeing’s design of the MCAS, saying it trimmed the nose down at too fast a rate, because it was designed to be engaged at a lower speed. At 250 knots, it was generating violent nose-down movements that the crew could not counteract.

Boeing declined to comment on the matter, citing the pending investigation.

A third angle-of-attack sensor may have allowed the flight computers to determine which sensor was giving a bad reading, Skow said, a view shared by other experts.

Tajer, the pilot’s union spokesman, said airline flight manuals indicate Boeing planes have two AOA sensors, while Airbus models have three. Could three sensors have made a difference in the Lion Air accident?

“I think Boeing is very much asking itself that question,” said Steve Wallace, the former director of the FAA’s office of accident investigation and now an aviation safety consultant.

But Wallace and others don’t make the case that three sensors would have necessarily changed the outcome. William Tuccio, a former aerospace engineer at the National Transportation Safety Board, said engineers rely on what’s called failure mode and effects criticality analysis as one way to assess risk probabilities and determine the proper amount of protections. 

Additional sensors may be unnecessary and could be a waste of money or an increase in complexity, he said: “If you’re going to answer a single question about reliability or engineering resilience, it is an incomplete question to just ask how many AOA sensors there are.”

Wallace pointed out that no jetliner has crashed while on an automated landing, though there have been accidents because of the crew’s incomplete understanding of the automation. “Fundamentally, automation is a huge plus for aviation safety,” he said. “You still need to know the basics of stick and rudder flying.”

And then there is the cost versus benefit. “How many backups to your backups do you need?” said Michael Barr, the former director of the USC safety program. “Nobody wants to talk about cost.”

No one in aviation expects automation to slow down, though there are sharp disagreements among experts about how far it should go.

As far as a fully automated aircraft, Sullenberger, who retired as a commercial pilot in 2010, is skeptical that it will happen for a long time and doubts he would fly in one. “Unless it is actually capable of learning in the moment and improvising, unless it could be essentially human, I don’t see how you ever have the same level of confidence.”

Robert Ditchey (a funny last name, if you ask me), a safety consultant, expert witness and former airline executive, says the day there is no pilot in the cockpit is when he will stop flying. “We have the ability to autonomously land a spacecraft on Mars, but do we want to put people on board? I don’t think so.”

But Wallace is a big fan of automation and where it is headed. “I drive my Tesla with autopilot on all the time. It keeps a safe distance.”

Milestones in aviation automation

1917 — Sperry Corp. develops the first rudimentary autopilot that uses gyroscopes to reduce pilot workload.

1932 — First instrument landing at Berlin-Tempelhof Central Airport, using the Lorenz beam system.

1949 — Bill Lear wins the Collier Trophy, aviation’s most coveted award, for development of the Automatic Pilot and Automatic Approach Control System at Santa Monica. Two years earlier a U.S. Air Force C-54 made a transatlantic flight, including takeoff and landing, completely under the control of the Lear autopilot.

1964 — First fully automatic landing using an instrument landing system, at Bedford Airport, United Kingdom.

1982 — The first modern flight management system, including an autothrottle that operates like cruise control in a car, introduced in the Boeing 767.

1987 — The first use of a fly-by-wire system on a passenger jet on the Airbus 320. Fly-by-wire takes pilot input and calculates the control-surface movements required to deliver the result.

1995 — Boeing’s new 777 has a fly-by-wire system with flight envelop protection, which is supposed to prevent stalls or excessive stressful movements, though a pilot can override them.

2016 — Boeing introduces the 737 MAX, which incorporates the maneuvering characteristics augmentation system, or MCAS. It helps control an aircraft’s tendency to pitch up in certain conditions.

2018 — Boeing announces a program to research a pilotless commercial jetliner.

bigkid posted:
IC EC posted:
bigkid posted:
Big_Boy_4005 posted:

Yes, airplanes are flying themselves to some extent, but they do have pilots on board, in case something goes wrong. They don't take off or land themselves.

Trains are really no different. They have to be made up and broken down, in order to get their loads to their destinations. You're just not going to send hazmat loads out there unattended. No matter how hard the railroads try, there will always be a human on board.

They actually could take off and land themselves, that technology has been around for a long time, and in many cases might make for more efficient operations, especially in bad weather conditions, but there are political and other reasons why it isn't done, and I suspect even if they automate that they would keep a pilot on board as backup, the way the BART system has an engineer as a backup to the automated system that  runs it (or at least they used to).  I don't know if trains will ever be fully automated, it is going to depend on context, much the same way that in crowded harbors a pilot takes over for the captain when bringing a big ship in, in denser areas at the very least there would be someone on board.

 

 

As a former airline captain, I disagree.  The automation is there to assist the pilots with mundane tasks (believe it or not, actually flying the aircraft) so they can better make decisions, which no machine will ever be able to do.  It is not there to replace the pilots.  An automated aircraft could not even get off the gate because the crew makes countless decisions that the passengers never see.

We normally only used our autoland system when the visibility was very low (around 600 feet) to allow the pilot not flying to concentrate on looking for the visual cues necessary to continue the approach while the pilot "flying" monitored the instruments and the autopilot system.  The autoland system had severe crosswind limitations (around 10 knots) while the normal crosswind limitation was 39 knots for pilot-flown landings.  Additionally, I am not aware of any aircraft that can accomplish an automated takeoff.  If you have an engine failure or fire, blown tire, runway incursion, or any other issue during takeoff, the captain will decide very quickly whether to abort or continue the takeoff.  A machine cannot make those types of decisions--they require thought, experience, and situational awareness.

The process of simply holding an altitude and navigating between GPS defined waypoints is what most people think about when this discussion arises.  There is far more to it than that.  Similarly, the process of operating a train from point A to point B is only a minor part of the entire process.

No argument on the complexity of flying a plane ,especially in difficult or emergency situations,right now autonomous vehicles have real problems with handling the unexpected. However, AI and machine learning are growing exponentially, they are doing things like a system that can infer a solution from a single point of data,something once thought impossible (in a human being they call that intuition).  With that,it isn't far off when an automated system could handle it better than a human being,among other things no emotions to get in the way,like panic...could also be found out to be impractical but i suspect not.

As far as automation creating new jobs, that is to be blunt self serving propoganda based on the past. The amount of new jobs created supporting automation are a small fraction of what they replace, an automated system to manage financial portfolios can be written and supported by a small team of people,replacing many more people,the analysts,portfolio managers and traders that traditionally did the work. With machine learning the system can improve its own operation,requiring even less people.

There are problems that cannot be solved by automation, that was proven about 80 yrs ago by Alan Touring, but the reality is that that limitation applies in a relatively small class of problems,the rest no matter how complex can eventually be solved.

 

I work in IT and as you may know the big push is to get everything to the "cloud" and automate everything possible. The future is here and it is called the software defined datacenter. The days of earning a living as a Windows server admin are done, you need to know how to work on the cloud.

In IT, automation is killing off traditional jobs and creating new ones, pretty much one for one. There is a massive back end needed to deliver automated services, I'd imagine the same can be said when you automate anything. That massive back end will require care and feeding for years to come.

The thing is, will people want to learn new skills and adapt? I see some people not wanting to change. That means there will be casulties as tomorrow becomes today.

AmeenTrainGuy posted:
Dominic Mazoch posted:

Ok, what happens if the cloud gors down?  Or you cannot get to it?

The cloud is run through the internet, and the internet is impossible to take down. Think about it like a mainline if it goes down trains will have to wait until it gets fixed. Not all the computation gets down through the cloud. The cloud only takes care of the tasks that require supercomputers. An automated train should theoretically be able to control itself to a stop if it gets offline. A deeper explanation can be done, but I believe it is too complex for this topic since this is not an artificial intelligence/machine learning forum.

I think the better question is,  what if the cloud gets hacked?

We already have vulnerable infrastructure in the form of water and power systems. Should we add trains and planes to the list?  Nuclear weapons are soooo last century. That won't be where the next attack will come from.

Big_Boy_4005 posted:
AmeenTrainGuy posted:
Dominic Mazoch posted:

Ok, what happens if the cloud gors down?  Or you cannot get to it?

The cloud is run through the internet, and the internet is impossible to take down. Think about it like a mainline if it goes down trains will have to wait until it gets fixed. Not all the computation gets down through the cloud. The cloud only takes care of the tasks that require supercomputers. An automated train should theoretically be able to control itself to a stop if it gets offline. A deeper explanation can be done, but I believe it is too complex for this topic since this is not an artificial intelligence/machine learning forum.

I think the better question is,  what if the cloud gets hacked?

We already have vulnerable infrastructure in the form of water and power systems. Should we add trains and planes to the list?  Nuclear weapons are soooo last century. That won't be where the next attack will come from.

No technology connected to the internet is perfectly secure (the only nearly secure systems are not on the Net, and if networked are only accessible through a local network that is not connected to the internet or via a direct leased line connection that  otherwise is not easily accessible (leased lines in theory can be tapped, but require local access to the cable and knowing which one is carrying the traffic). 

Comparing the cloud to for example a water system getting hacked is not a valid comparison IMO. A water system likely is running their applications on their own environment (or a contracted host environment, same difference), they wrote the application, they wrote the application security, the database structure, the database calls in the application, basically the whole thing. Most of the hacks you hear about are at this level, while operating systems have their security flaws, it is why they continually update phone and desktop systems OS, that isn't where most of the breaches  happen.

The problem is that on this level a lot of factors make these systems vulnerable. With things like the water system or electrical grid, you likely have legacy systems that have been out there a long time, that have been patched and changed, and security quite frankly wasn't as big an issue back when they were written, and also with things like a water system or electrical grid, they don't exactly have huge IT budgets or the willingness to spend money on teams of people to test the code for vulnerabilities, then fix them, both of which can be time consuming and expensive. 

The real problem IMO quite frankly is that the consequences of a breach to a company is so low, that even now they don't really care. When Anthem got hacked and critical personal information was breached in the process, the cost to them was the CEO (who along with the CIO kept their jobs) sitting up there and saying "I feel your pain" and offering a free year of credit reporting to the people affected; Target had a massive breach, and within a couple of months of it happening, they had no ill effects and even had growth.  Despite all the furor, there were no real consequences to the firms in terms of regulations, what regulations that exist are toothless and the fines are so small the companies are often willing to risk that. Basically, if it doesn't give a RIO or doesn't preclude a major loss, they won't spend money on it. I work in an industry on the other hand where data and systems safety is heavily regulated, and you better believe they take security seriously, because the cost of failure would be huge, unlike places like Anthem and Experien and Target and so forth. 

Cloud computing is a bit different. First of all,  applications running in the cloud are removed from a lower level threats  you have running on a traditional hardware farm, the are abstracted from that layer from what I understand, and the cloud  infrastructure was designed with security from the beginning, in part because it was so new (and for example Amazon, built the cloud to handle their online retail business, which they knew security would be a major factor if they had a major hack, so they designed it with a major level of security, IWS is if not the biggest, one of the biggest players in cloud computing companies run their applications on). More importantly, the application vendors who use the cloud are writing applications from scratch, and unlike the patched together legacy systems that routinely get hacked, they are designed with security in mind, too. Plus, from what I understand about cloud computing, the cloud service has fairly strict requirements when applications are hosted there and have their own set of tests that people using the cloud have to have their applications pass, plus the cloud people run all kinds of tests themselves for things running in their environment, do security penetration testing and the like from what I have been told.

More importantly, the cloud vendors themselves would pay dearly if loopholes in the way the cloud was set up allowed major hacking of the clients apps running in there, it would destroy them; unlike the empty suits at Anthem or Experien or Target, the cost of failure is so high that they can't take security lightly, and they don't, it would drive applications to switch to another vendor and they could face serious financial liability for what happened to any vendor on there. 

You would figure something like Experien, that has the information of hundreds of million of people in it, or a water plant or a electrical grid, would be considered critical systems, but they aren't. I would be concerned about security around automated control systems on aircraft and trains, but I would be a lot more concerned about how well these systems can respond to out of the norm situations than being hacked, security , as difficult as it can be, is a lot easier thing to build in than being able to handle out of the norm situations, it is why I tend to look at automated flight systems (pilotless ones) or fully automated trains on any large scale as not being a great  idea. 

Dominic Mazoch posted:

Or an EMP at the right place.  Say where those supercomputers are!

An EMP of the scale required would wipe out so much modern infrastructure the cloud farms would be the least of it, it would fry the power grids, it would wipe out most of the cars and trucks on the road (modern trucks and buses rely on computers, too), many appliances would fail, the bank teller networks and atm's would be toast. The thing about the cloud is that it doesn't run in some data center somewhere, it runs across data centers located all over, it isn't centralized all in one place, if an EMP wiped out a particular data center the applications can be restarted in another, if  I understand cloud architecture correctly, it has a level of redundancy in it traditional data centers don't really have (I am just speculating, but I suspect with cloud architecture, which I am not an expert on, they could implement what used to be called fault tolerance, have applications running in multiple farms that run the cloud and if one goes down, it simply keeps processing out of the other ones); if it is a widespread EMP, like some sort of "EMP" bombs detonated in the atmosphere, it would wipe out a lot more than just cloud computing. 

Add Reply

Post

OGR Publishing, Inc., 1310 Eastside Centre Ct, Suite 6, Mountain Home, AR 72653
800-980-OGRR (6477)
www.ogaugerr.com

×
×
×
×
Link copied to your clipboard.
×
×