Jump to content

Recommended Posts

Is this the first step into drone passenger flights?

 

http://money.cnn.com/2016/01/06/technology/ces-2016-ehang-drone/

 

I can hear the announcement now.

 

"Welcome aboard Drone Airlines inaugural flight from Los Angeles to Honolulu.  There is no pilot aboard, but this drone airliner will be controlled during the first half of the flight from Los Angeles.  Halfway across the ocean to Hawaii control will be handed over to a controller in Honolulu.  Relax and enjoy your flight and be assured that nothing can go wrong...can go wring...can go wrong...can go wrong..."

 

Noel 

Link to post
Share on other sites

The thing that makes a pilot perform at his best is having his own skin in the game. Nothing reinforces accountability like being tied to the same fate as everyone else.

 

I hope it never happens...

 

 

Edited by Hobnobs
  • Upvote 1
Link to post
Share on other sites

Doubt we'll ever live to see it.  This one isn't certified for anything yet.  There will be a bunch of regulatory issues to resolve, airspace, blah blah...  But I suspect it's inevitable.

Link to post
Share on other sites

I suspect that eventually there won't be anyone on the ground controlling the plane for take offs and landings either, just good ol' George. 

 

By the time we get to that point, I suspect we will all be so used to our self-driving cars that many people won't even think twice about getting into a plane with no pilot.

 

Eventually people will also literally "plug in" with some kind of direct to brain computer connection

 

Well, that or the zombie apocalypse...I'm not looking forward to any of these choices.  (I may change my mind about self driving cars in another 20 years)

 

 

Link to post
Share on other sites

I wonder if the the first airliners with autopilot and full cat II / III autoland capability were also viewed with such trepidation, even when landing you safely in fog and the passengers being blissfully unaware that technolgy brought them safely back to terra firma...

You can't stop progress, all you can do is ensure it is applied in good order...move with the times... instead of being stuck in the past !

  • Upvote 1
Link to post
Share on other sites

The first airliners with autopilot and full cat II/III autoland capability also had a human pilot as a backup.  Those devices assisted the pilot who was still in command and responsible for the wellbeing of the passengers and himself.  Drone airliners would not have that backup.  If they did what would be use of droning?

 

If a drone jumbojet crashed the legal profession would have a field day assessing blame.  The airline?  The drone manufacturer?  The systems manufacturer?  The flight programmer?  All of the above?

 

Noel

  • Upvote 1
Link to post
Share on other sites

Personally, I don't see it happening for a few reasons.

 

1.  Why build a system that you intend to load 300 people on, then NOT keep 2 other people on it as a Human Backup in case the automated system fails?  You could make the argument today that pilots aren't really needed for anything except takeoffs and landings, 'cos a modern autopilot can be programmed to handle the rest of the flight.  So why not let the pilots just go to sleep during that 14 hour flight half way around the world?  Who needs 'em? Would you want to be on that flight knowing there wasn't anybody in the cockpit for 12 hours if something went wrong?

 

2.  The drone shown is the equivalent of the "personal flying cars" everybody was predicting 70 years ago.  The closest thing we have to that today is a helicopter.  But helicopters aren't in everybody's garages yet being used to commute back and forth to work or the grocery store.

 

It's not the same as the "driverless cars" agenda we are seeing many companies throw huge amounts of money into perfecting.  Driverless cars make sense because of something in Game Theory called "The Tragedy of the Commons".  The only real way to make our road and highway systems ("The Commons") more efficient and SAFER is to take the decision-making out of the hands of the Human driver.  Too many people don't drive on highways with the concept of making the system work at peak efficiency and SAFER for everyone.  There is a "Me First!" attitude that bogs the whole system down for everybody (that's "The Tragedy" part).  That doesn't exist in organized transportation systems like commercial airlines.  Any pilot that cut another airplane off on final approach just to get to the airport 5 seconds sooner would get hung.  There is no "Tragedy of the Commons" need that exists in commercial aviation because the "drivers" (pilots) of those aircraft aren't creating a "Tragedy" of any sort.  There is no compelling reason to REMOVE the pilots from the aircraft, even if the aircraft could do the entire flight automatically.  There IS a reason to keep the Human Factor around in case the automatic system has a failure.

Edited by FalconAF
  • Upvote 1
Link to post
Share on other sites

 

20 minutes ago, birdguy said:

If a drone jumbojet crashed the legal profession would have a field day assessing blame.  The airline?  The drone manufacturer?  The systems manufacturer?  The flight programmer?  All of the above?

 

Noel

The only difference between this and when an airliner crashes today is that the pilot is added in. Airline, systems mfr, flight programmer, maintenance personnel, maintenance facilities, ad infinitum are already part of the legal tangle in such as case. Legally, it would change approximately nothing.

 

11 hours ago, Mickel said:

Doubt we'll ever live to see it.  This one isn't certified for anything yet.  There will be a bunch of regulatory issues to resolve, airspace, blah blah...  But I suspect it's inevitable.

No airspace issues either. They would use the exact same air routes, approaches, SIDs/STARs, etc. just as autopilots are already programmed to follow.

 

11 hours ago, Hobnobs said:

The thing that makes a pilot perform at his best is having his own skin in the game. Nothing reinforces accountability like being tied to the same fate as everyone else.

 

This is probably a bigger factor, although with no pilot, pilot error goes out the window (like the recent landing on a taxiway in Seattle) and airliner crashes would probably be less frequent.

Link to post
Share on other sites

I can certainly see combat aircraft becoming exclusively unmanned in future (I think the F-35 will be the last manned fighter jet ever developed by Western NATO nations), since the reasons to risk the life of a human being over a warzone are diminishing by the day (not to mention the cost savings of not having to design an aircraft to support a human being).

 

But I don't think transportation aircraft will ever become unmanned - for the simple reason that, If you're going to risk the lives of hundreds of passengers in an aircraft, then what does removing two pilots really gain you in terms of reduced risk? Sure you could save money by not employing pilots, but I don't think passengers are ever going to be comfortable climbing inside a metal tube and placing their lives entirely in the hands of a computer. Although airliners can pretty much fly themselves already, ultimately human beings will always be able to do things computers can't do when things go wrong - case in point, Capt. Sullenberger's landing on the Hudson - show me a computer that can do that.

 

What I do think however, is that we'll see the emergence of "optionally manned" airliners - that can fly themselves for cargo transport flights, but can be controlled by a human for passenger flights. And I think we'll see the role of two pilots replaced by a single "systems monitor" - i.e. where the aircraft flies itself, but one human is present on the flight deck, to monitor the flight and be able to take over should something go wrong.

  • Upvote 1
Link to post
Share on other sites

Agree with all you said.  We've already seen that in the past decades with the elimination of Flight Engineers on most commercial airlines.  The automated systems monitor those parameters and present the information to the pilot and copilot directly, who can take the appropriate actions as required.  Whether we go to having just one systems monitor vs two will be determined by whether the Risk Management decision to not have a second Human backup for the first Human backup is worth the risk.  We seem to be headed in that direction though, as many complex aircraft today that would have required two pilots to operate only a decade ago have been certified as one pilot only.  Examples are something like the Cessna Citation Mustang.     

  • Upvote 1
Link to post
Share on other sites

In all honesty, within the next 50 years I can see the role of professional pilot (both military and commercial) being completely eliminated - replaced instead with "flying IT managers", who may in very rare circumstances (i.e. extreme emergencies) get to actually do some flying. I think today's young kids are certainly going to be the last generation of fighter pilots.

 

Maybe in future we'll see engineers being the ones who are in the cockpits - since they will have to understand the aircraft enough to service it, then with some extra training (recoveries/landings etc.), perhaps they could fulfill the role of "systems monitors". Airlines will certainly like it as they will only have to pay one person to perform two roles!

 

It's a sad thought, but future generations with a passion for flying may have to settle for GA aircraft to get their "fix" - a career as a pilot could well become a thing of the past, just as navigators & flight engineers have.

 

Edit:

 

Helicopters may be an exception to this, since sometimes they have to land on fields/roads and other "undefined" areas, where human judgement is required to assess suitability of the landing site, which computers may struggle to do.

Edited by Pete H
Addition
  • Upvote 1
Link to post
Share on other sites
1 hour ago, Pete H said:

In all honesty, within the next 50 years I can see the role of professional pilot (both military and commercial) being completely eliminated - replaced instead with "flying IT managers", who may in very rare circumstances (i.e. extreme emergencies) get to actually do some flying. I think today's young kids are certainly going to be the last generation of fighter pilots.

 

I don't.  To have a completely autonomous aircraft capable of doing everything without Human inputs would require an advance in computer Artificial Intelligence that almost all researchers in that area say probably won't ever happen.  A computer can only produce results based on it's programming, and it simply isn't possible in the foreseeable future we will develop any computer that will be able to learn and "think" on it's own to do everything like that.

 

What's more probable is that the Human pilot won't need to be IN the airplane's cockpit, but can monitor and react to a system failure from a remote location.  Much like the military uses ground based pilots now to fly Unmanned Air Vehicles (UAV's like the Predator).  Those aircraft aren't anywhere near being fully autonomous and still require a ground based "pilot" to provide inputs to them to carry out a successful flight profile.  It's conceivable that an unmanned airliner may not need to have the pilot or systems monitor actually be on the aircraft in the future, but instead the pilot could be the systems monitor and able to take action from the ground location if required.  You still have a paid pilot who has to be trained in that case.  He/she just may not have to be on the aircraft to do the job.  But the likelihood of having any computer system capable of totally replacing the pilots of an aircraft are far out of reach of today's technology, and not very promising even over the next 50 years.  Computer Artificial Intelligence just isn't progressing that way, despite the predictions it would over the PAST 50 years.  We are nowhere near reaching the "last generation" of Human pilots.

 

Your comment about not being able to do it for helicopters is correct, but the same type of "judgement" is still required for the pilot of a fixed wing aircraft on a specified flight plan with a specified destination.  It's even more applicable to a fighter pilot who has to react to an ever-changing combat situation they are involved in, whether it be an air-to-air or air-to-ground engagement.  Taking the pilot out of the real cockpit solves the issue of the pilot possibly getting killed, but it doesn't eliminate the need for the pilot to still be an integral part of the weapons system from a remote location.

 

Edited by FalconAF
Link to post
Share on other sites

The windy thread highlights another thing.  Human pilots are able to handle greater crosswinds than computers are (currently), and those pilots are flying by feel - something they can't have sitting in some office block half a world away.

Edited by Mickel
Link to post
Share on other sites
On 1/8/2016 at 2:43 AM, Mickel said:

The windy thread highlights another thing.  Human pilots are able to handle greater crosswinds than computers are (currently), and those pilots are flying by feel - something they can't have sitting in some office block half a world away.

 

Yes and no.  Currently a human pilot can "technically" land an airplane more safely in high crosswind conditions.  But a "pilot" sitting at a remote location can still do it without having to "physically feel" the crosswind.  As long as the visual cues presented to the pilot at the remote location are sufficient, the pilot develops his/her own "feel"...even without the physical sensations.  You might call it more of an "intuition" about what is happening.  We do it all the time in our own flight sims.  You don't feel ANY real physical sensations while landing in a crosswind in Flight Simulator.  But from experience over time, your visual cues tell you what is happening, and you make the appropriate corrections as needed.  One of the hardest things a student Instrument pilot has to learn is to IGNORE any physical "feeling sensations" when flying without visual cues, and rely ONLY on the visual cues provided by the instruments.  The same thing works for a remote location pilot making a VFR landing looking at a computer screen while landing the aircraft.  They don't need to "feel" anything as long as the visual cues they get looking at the screen presentation out the cockpit window are sufficient. If it didn't work that way, the Air Force would be scraping crashed Predators off runways by the dozens after unsuccessful landing attempts by the remote pilots.

 

In the same vein, a remote pilot doesn't have to "feel" any physical effects of a Wind Sheer on final approach to take the appropriate corrective action.  As long as the pilot is "notified" the wind sheer is happening (by the ground cockpit he/she is operating the aircraft from) the pilot can still make the appropriate corrections.  If the linked systems on the real aircraft detects the wind sheer and the ground cockpit starts screaming "Wind Sheer!  Wind Sheer!"), the remote pilot still becomes aware of it and can take the appropriate action and make appropriate control inputs. Wind sheer alerting systems in the cockpits of real airplanes do the exact same thing for the pilot sitting in the real airplane.

 

Edited by FalconAF
Link to post
Share on other sites

An half meter drone is already plain dangerous, no auto-rotation, no parachute, no safety at all for people around or close to the 8 prop.

 

It's flying on 8 motors, it will go down fast if only one goes defective.

 

Quad prop flying machine have a market because they are stable craft but not this way

 

Ben

 

iu.jpeg

Link to post
Share on other sites
Quote

You can't stop progress,

stupidity more like it, some cant seem to see how "dumb" we are getting. Once upon a time reading a map want an issue, now when your GPS dies in the moddle of a drive you get all confounded because you dont know where you are. Im not liking where all this is going. We are being "told" what we need without out our input. This drone is a nice toy for the rich, i hope im not around when it kills people.

  • Upvote 1
Link to post
Share on other sites
2 hours ago, mjrhealth said:

stupidity more like it, some cant seem to see how "dumb" we are getting. Once upon a time reading a map want an issue, now when your GPS dies in the moddle of a drive you get all confounded because you dont know where you are. Im not liking where all this is going. We are being "told" what we need without out our input. This drone is a nice toy for the rich, i hope im not around when it kills people.

 

I have to say I agree. We are becoming slaves to technology rather than keeping to practical ways for it to serve us. 

 

I work in technology. Technology is inherently flawed because it is created by people who are inherently flawed. The thing humans have though is adaptability. That extra little piece that cannot be programmed, that can deal with the most unexpected of scenarios. Not every human will succeed when the unexpected happens but technology will always fail if the unexpected hasn't been programmed.

 

It scares the crap out of me frankly; the way we as a species think technology is the answer to everything...

  • Upvote 1
Link to post
Share on other sites

I have an old fashioned cell phone Hobnobs.  It doesn't even have a camera.  And I never carry it around with me.  It stays on my desk at home.

 

People ask me what I will do if I have an emergency like a flat tire in a parking lot.  I tell them I'll ask someone to make a call for me.  They will be thrilled to death to whip out their smart phone and use it.

 

Resistance is futile.  You will be assimilated into the collective.  But I intend to be the last old bastard dragged into it.  Hopefully I'll die first.

 

Noel

 

 

Link to post
Share on other sites

The major concern has to be whether the business model would work. Would any airline order very expensive pilotless aircraft in the hope that the public would want to fly in such aircraft? That would be an extremely brave decision.

Then, assuming that an airline introduced such services, the well made points about having a person to cope with unusual situations rears its ugly head. Anyone in the passenger compartment got any idea of how to fly one of these automated beasts?

 

Edited by mcdonar
Link to post
Share on other sites

One could argue there are times when the two or three guys in front don't know how to fly the automated beasts we have now and allow themselves to get into an upset condition...

Edited by Mickel
Link to post
Share on other sites

It's notoriously tricky to predict the future. We tend to overestimate the short term rate of change, but underestimate the long term rate of change. That said, I'd always bet on more automation, not less. Humans aren't getting any cheaper to feed and house, whereas computers are constantly getting more efficient. There's a limit to how week people can be trained to avoid errors. While much cab be said for the capability of the human brain to respond to unique situations, it would be foolhardy to assume past and present limitations of computer intelligence will carry forward into the future. 

 

Flying is a bit of an odd case, at least for those of us willing to devote a great deal of time and effort into simulating the experience. It may seem like automating flight would take away something people want to do rather than classic automation examples, such as the drudgery of production line assembly. While I can see this being an argument for allowing human controlled flight for personal entertainment, I'm not sure it's a good argument against automating most commercial flight undertaken today. Sooner or later computers will be safer than humans, as well as cheaper. At that point, who is going to pay wages to a human to fly their multi-million dollar aircraft, at an increased risk to the plane, passengers, and cargo?

 

As someone who is young enough to hope for at least as many years ahead as I have behind me, I would be surprised if automated commercial flying is not widespread within my lifetime. Not just in terms of pilots either, but also with regards to aircraft maintenance as well. 

Link to post
Share on other sites

One of the issues with automating a system completely is taking into consideration the CONSEQUENCES of what could happen if the automated system fails.  That is the Risk Management consideration that has to be analyzed before deciding whether or not to eliminate the Human Backup component.

 

We automate many things today, like assembly lines, and don't require a Human Backup to be part of the system.  If the computers fail, the assembly line just grinds to a halt (hopefully).  The whole assembly line just stops working, but the building doesn't fall down around it. Then you call in the Humans to see what happened, fix it, and restart the assembly line.  You would use a totally different Risk Management assessment to determine that eliminating all the Humans in a Nuclear Power Plant Control Room wouldn't be a "good idea", regardless of how automated that Nuclear Power Plant was at "running itself".

 

If the computers on a fully automated airplane malfunction and there is no Human pilot, either in the actual airplane or "flying" the airplane from a remote location, the whole airplane could come crashing to the ground with massive consequences and casualties.  Any "remote systems monitor" or "remote pilot" still has to be a "full time" presence in the operation of the airplane, even if they are sitting in a ground-based remote operating location.

 

Noel brought up an interesting issue when he said:

 

16 hours ago, birdguy said:

People ask me what I will do if I have an emergency like a flat tire in a parking lot.

 

Is that flat tire REALLY an EMERGENCY?  The car isn't going anywhere.  It's not gonna crash into anything.  It's just gonna sit there doing nothing.  SOME people might consider a flat tire on a car in a parking lot an "emergency", but it's far from actually BEING one.  Failure to make an ACCURATE Risk Management analysis in not a valid reason to try and justify using available technology as a "safety tool" to avoid a non-tragedy.  

 

I ran into the same thing with one of my Niece's when I bought my touring motorcycle 4 years ago and started doing road trips on it all over the U.S.  She INSISTED I buy a real-time tracking device that would let her know if I ran off a cliff or something and was laying on the ground dying and nobody would know about it.  I told her, "If I run off a cliff 1,000 miles from your home, what in the heck are you gonna be able to do other than call the authorities to tell them to come get my dead body?  The odds of me running off that cliff are about the same as the odds of an airplane falling out of the sky on me.  Sorry, but I'm not spending $500 just to alleviate YOUR fears about ME riding a motorcycle.  If I don't send you a text message at the end of the day saying I arrived safely at my hotel for the night, THEN you can call them and report me missing."  :lol:    

Edited by FalconAF
  • Upvote 1
Link to post
Share on other sites
14 minutes ago, FalconAF said:

One of the issues with automating a system completely is taking into consideration the CONSEQUENCES of what could happen if the automated system fails.  That is the Risk Management consideration that has to be analyzed before deciding whether or not to eliminate the Human Backup component.

 

We automate many things today, like assembly lines, and don't require a Human Backup to be part of the system.  If the computers fail, the assembly line just grinds to a halt (hopefully).  The whole assembly line just stops working, but the building doesn't fall down around it. Then you call in the Humans to see what happened, fix it, and restart the assembly line.  You would use a totally different Risk Management assessment to determine that eliminating all the Humans in a Nuclear Power Plant Control Room would be a "good idea", regardless of how automated that Nuclear Power Plant was at "running itself".

 

If the computers on a fully automated airplane malfunction and there is no Human pilot, either in the actual airplane or "flying" the airplane from a remote location, the whole airplane could come crashing to the ground with massive consequences and casualties.  Any "remote systems monitor" or "remote pilot" still has to be a "full time" presence in the operation of the airplane, even if they are sitting in a ground-based remote operating location.

 

Noel brought up an interesting issue when he said:

 

 

Is that flat tire REALLY an EMERGENCY?  The car isn't going anywhere.  It's not gonna crash into anything.  It's just gonna sit there doing nothing.  SOME people might consider a flat tire on a car in a parking lot an "emergency", but it's far from actually BEING one.  Failure to make an ACCURATE Risk Management analysis in not a valid reason to try and justify using available technology as a "safety tool" to avoid a non-tragedy.  

 

I ran into the same thing with one of my Niece's when I bought my touring motorcycle 4 years ago and started doing road trips on it all over the U.S.  She INSISTED I buy a real-time tracking device that would let her know if I ran off a cliff or something and was laying on the ground dying and nobody would know about it.  I told her, "If I run off a cliff 1,000 miles from your home, what in the heck are you gonna be able to do other than call the authorities to tell them to come get my dead body?  The odds of me running off that cliff are about the same as the odds of an airplane falling out of the sky on me.  Sorry, but I'm not spending $500 just to alleviate YOUR fears about ME riding a motorcycle.  If I don't send you a text message at the end of the day saying I arrived safely at my hotel for the night, THEN you can call them and report me missing."  :lol:    

 

Understanding risk is fundamental to everything we do. From going to the toilet at night without switching any lights on to jumping out of an aeroplane with a parachute. Everything has a potential consequence, some more serious than others. 

 

But; you can't defend what you don't know. We would have to be pretty damn sure you have every conceivable risk covered to apply a full technology solution to this problem. It's a problem that continues to need a human solution because you can't mitigate the potential risks of having an autonomous aircraft flying real passengers.

  • Upvote 1
Link to post
Share on other sites

Falcon...when my wife has a flat tire it's a REAL emergency.  When is the last time anyone here had a flat tire?  And how many current generation drivers actually know how to change a flat?  It doesn't happen as often as it used to anymore, but it does happen.

 

That was one of the first things my father taught me before I ever got behind the wheel and learn how to drive.  Fixing a flat and reading a map.  Lost arts today when people don't know what a lug wrench or jack handle is for let alone find the continent they live in on a world map!  Today's GPS goes on the fritz and the first questions that come to mind are, "Where am I and where am I going?"

 

Noel

Link to post
Share on other sites

While computer controlled flight does clearly present far greater risk than operating a human-free production line, that isn't a fundamental argument against full automation. Rather, it means that it is a more complex automation process that has not been fully implemented yet.

 

Humans fail just like computers do. You don't need perfect computers any more than you need perfect human pilots. You only need the computers to be as good or better than their human equivalents. At that point, having human pilots makes no sense. It's also very unlikely that a human backup to a robot pilot that is proven to be more reliable than said human would make much sense either. If the human pilot did absolutely nothing at all to affect the flight 99.99% of the time, it would be very hard to go from idle watcher of instruments to prepared to do the unexpected for that 00.01% of the time. That's similar to the problem the TSA has. They employ people who are able to harass passengers who take "too much" liquid through checkpoints (a common occurrence) but are useless at detecting weapons when tested, because it's too rare for people to be able to keep paying attention for them. After looking at countless bags with no weapons in them, the human brain stops devoting attention to looking. Said people employed in that task think they might still be paying attention, but they are being tricked by their own brains.

 

A crisis in a flight these days might be an equally rare occurrence, but the pilots always have cause to know the state of the plane to ensure normal operations, and are therefore usually equipped to respond to the unexpected to some degree. A backup human pilot who doesn't have any responsibility at all most of the time might be able to convince themselves they are maintaining situational awareness, but in reality they will inevitably zone out.

 

It would therefore be far better to spend effort in continued mprovement of the robot pilot, and/or making the backup pilot another computer that doesn't get bored or inattentive just because their efforts aren't needed most of the time.

Link to post
Share on other sites
4 hours ago, paralipsis said:

While computer controlled flight does clearly present far greater risk than operating a human-free production line, that isn't a fundamental argument against full automation. Rather, it means that it is a more complex automation process that has not been fully implemented yet.

 

Humans fail just like computers do. You don't need perfect computers any more than you need perfect human pilots. You only need the computers to be as good or better than their human equivalents. At that point, having human pilots makes no sense. It's also very unlikely that a human backup to a robot pilot that is proven to be more reliable than said human would make much sense either. If the human pilot did absolutely nothing at all to affect the flight 99.99% of the time, it would be very hard to go from idle watcher of instruments to prepared to do the unexpected for that 00.01% of the time. That's similar to the problem the TSA has. They employ people who are able to harass passengers who take "too much" liquid through checkpoints (a common occurrence) but are useless at detecting weapons when tested, because it's too rare for people to be able to keep paying attention for them. After looking at countless bags with no weapons in them, the human brain stops devoting attention to looking. Said people employed in that task think they might still be paying attention, but they are being tricked by their own brains.

 

A crisis in a flight these days might be an equally rare occurrence, but the pilots always have cause to know the state of the plane to ensure normal operations, and are therefore usually equipped to respond to the unexpected to some degree. A backup human pilot who doesn't have any responsibility at all most of the time might be able to convince themselves they are maintaining situational awareness, but in reality they will inevitably zone out.

 

It would therefore be far better to spend effort in continued mprovement of the robot pilot, and/or making the backup pilot another computer that doesn't get bored or inattentive just because their efforts aren't needed most of the time.

 

You make some good points. This whole discussion has become quite interesting. Full automation does have its benefits, such is with most production lines these days. 

 

Perhaps we will get to the point where it is viable that a computer should fly a plane instead of a human. I hope I never to get to see it. Perhaps my arguments against are emotional rather than logical. If technology is the answer to everything, why have humans at all?

 

Will we eventually make ourselves obsolete in all tasks? Humans are flawed, I have said it before in a previous post. But it makes me sad to think that we will destroy our own ability to create when technology is seen as the answer to everything. Creating and working together collaboratively is what humans do best. The Apollo program, Concorde, the Golden Gate Bridge, works of music, works of art, fighting disease etc. Just a few that immediately come to mind.  Are we to hand all of these tasks over to technology? When do we say enough? When do we cross that line?

Edited by Hobnobs
Tidy up
Link to post
Share on other sites

Of course it may happen that pilots remain ground based, and only go 'online' to the aircraft when intervention is required.

But I am not at all sure that I would fly with such an airline. We must not forget that the passengers are always in control of an airline's future.

Edited by mcdonar
Link to post
Share on other sites
2 hours ago, Hobnobs said:

If technology is the answer to everything, why have humans at all?

It's odd really. We so closely tie the worth of people to their wage paying jobs that this is a valid question, and yet it needn't be that way. There's a lot about the way the global economy works right now that doesn't really make a lot of sense in the world we are moving towards. It remains so because the people who are going to lose out from the eventual changes are those who hold all the resources now. Much as we see have seen within the music industry, change comes eventually, but it's messy and awkward, with those who were the gatekeepers in the past holding on for dear life. Change will come though. As the saying goes, idle hands are the devil's playthings. If vast numbers of people are shut out of the economy because they have nothing to do, there will be pushback.

 

From what I can see, the tipping point is likely to be self-driving trucks. Manufacturing has already dwindled to very little, and there's a limit to what the service sector can support. Few people would argue that being a truck driver is a dream job, so I doubt many will miss the hours spent staring at the road ahead, but it has been a massive employer globally, especially in proportion to the other, shrinking, employment sectors.

 

How messy the change will be, and what it will look like on the other side will only really be apparent when we get there. But it's something we've been moving towards for decades, and the rate of change is only increasing. Personally, I'm optimistic about the possibilities.

Link to post
Share on other sites
9 hours ago, paralipsis said:

While computer controlled flight does clearly present far greater risk than operating a human-free production line, that isn't a fundamental argument against full automation. Rather, it means that it is a more complex automation process that has not been fully implemented yet.

 

Humans fail just like computers do. You don't need perfect computers any more than you need perfect human pilots. You only need the computers to be as good or better than their human equivalents. At that point, having human pilots makes no sense. It's also very unlikely that a human backup to a robot pilot that is proven to be more reliable than said human would make much sense either. If the human pilot did absolutely nothing at all to affect the flight 99.99% of the time, it would be very hard to go from idle watcher of instruments to prepared to do the unexpected for that 00.01% of the time. That's similar to the problem the TSA has. They employ people who are able to harass passengers who take "too much" liquid through checkpoints (a common occurrence) but are useless at detecting weapons when tested, because it's too rare for people to be able to keep paying attention for them. After looking at countless bags with no weapons in them, the human brain stops devoting attention to looking. Said people employed in that task think they might still be paying attention, but they are being tricked by their own brains.

 

A crisis in a flight these days might be an equally rare occurrence, but the pilots always have cause to know the state of the plane to ensure normal operations, and are therefore usually equipped to respond to the unexpected to some degree. A backup human pilot who doesn't have any responsibility at all most of the time might be able to convince themselves they are maintaining situational awareness, but in reality they will inevitably zone out.

 

It would therefore be far better to spend effort in continued improvement of the robot pilot, and/or making the backup pilot another computer that doesn't get bored or inattentive just because their efforts aren't needed most of the time.

 

 

I agree with your philosophical  argument.  But it isn't scientifically valid.  Yes, humans fail just like computers do.  But computers fail WORSE than humans do in a lot of cases when THEY fail.  And when that happens, you need the LESS likely to "fail worse" option available to correct the computer's failure, especially if the consequences of the computer's failure might be catastrophic.  It's correct that you don't need perfect computers any more than you need perfect pilots...and that's a good thing because Humans are NEVER going to be able to MAKE a perfect computer, any more than we will be able to make a perfect Human.  When you say, "You only need the computers to be as good or better than their human equivalents", that is an UN-ACHIEVABLE GOAL in the specific qualities a Human brain has...intuition, judgement, and many others.  Think about it.  How would a computer programmer ever write a program that would be "equal to" or "better than" a Human brain?  NOBODY today...even the highest educated people in neurological sciences, biology, psychiatry, psychology, or even physics...completely understands how our OWN Human brains work.  How is anybody going to write a computer PROGRAM that could be "equal to" or "better than" it?

 

That's the problem facing the development of Artificial Intelligence, which is what these computer systems we are talking about really are.  If you don't know how to write a program that doesn't produce ONLY "machine logic", how do you write a program that would include things like Human Intuition, Human Judgement, or "intelligence" that isn't just based on "machine logic and nothing else? All the things that a computer CAN'T mimic that a Human brain can do?

 

The idea of building a computer that is "equal to"...let alone "better than"...a Human brain is a great story line for a Science Fiction book or Science Fiction movie.  But there is NOTHING in our current knowledge of HOW to DO it.  On the contrary, the past several decades of trying to do it have made most researchers in Artificial Intelligence come to the conclusion that it likely will NEVER happen. We might be able to make something like a Star Trek "Data" android computer, but it would lack the fundamental qualities of Human intuition and judgement (among many others).  We don't even know all of the things that make our own Human brains work the way they do.  Probably never will, either.  That's the major reason we will also probably never make a computer that is "equal to" or "better than" a Human Being. 

Edited by FalconAF
Link to post
Share on other sites
8 minutes ago, FalconAF said:

When you say, "You only need the computers to be as good or better than their human equivalents", that is an UN-ACHIEVABLE GOAL.  Think about it.  How would a computer programmer ever write a program that would be "equal to" or "better than" a Human brain?  NOBODY today...even the highest educated people in neurological sciences, biology, psychiatry, psychology, or even physics...completely understands how our OWN Human brains work.  How is anybody going to write a computer PROGRAM that could be "equal to" or "better than" it?

 

That's the problem facing the development of Artificial Intelligence.  If you don't know how to write a program that doesn't produce ONLY "machine logic", how do you write a program that would include things like Human Intuition, Human Judgement, etc...all the things that a computer CAN'T mimic that a Human brain can do?

That feels like you are asking an AI pilot to think like a human pilot in order to be safer than a human pilot. I don't believe that to be the case. I think it is entirely possible that the current AI design paradigms are more than adequate to the task. I doth think you need an AI to react in the same way as a human would. All the AI pilot has to do is fly the plane safely and efficiently. It need not solve the hard problems of AI and have self awareness or crack a joke.

 

It's not a fundamental science problem, but rather an engineering one that we've been steadily working on for decades. Given one or two more (perhaps less given the accelerating rate of developments in this area) it seems quite likely that computers will surpass human pilots, and at a very low cost to boot.

Link to post
Share on other sites

It really all boils down to money.  Especially labor costs which is any industry's major overhead.

 

If we can make more money by producing our widgets with robots instead of paying wages and healthcare costs to human beings we will do it.  Nothing personal, just business.

 

I suppose the big decision society will have to make when we will get to parlipsis' tipping point is re-evaluating the priorities.  What's more important, people or money?  Looking at our environment and the world today I'm not as optimistic as paralipsis is.

 

Noel 

Link to post
Share on other sites

 

How many millions did NASA invest in a zero-gravity pen that astronauts could use to write while in space?  The Soviets space program found a solution that was more practical -- they sent them up with pencils.

 

I think there comes a point where advance technologies reaches the absurd.  And certainly I will agree the root cause of that is money and the cost of producing/servicing/moving x, y and z, no question.  I suppose I am all for 'incremental advances' but I just don't see where we will remove humans altogether.  Robots can make, build, process and do, but they cannot plan, think outside the box, or sell to humans. They can aid in all those activities, but humans need to work with other humans (particularly in sales).

 

First get me a good hazelnut latte, made by a robot, the way I like it (and in a manner in which I am willing to pay $4 for it), and then maybe we can talk about fully automated drone operated cars, taxis, trains and eventually passenger flights.  Won't see it my lifetime, but no doubt we'll see folks working on it.

 

Edited by Ripcord
  • Upvote 1
Link to post
Share on other sites
1 hour ago, paralipsis said:

That feels like you are asking an AI pilot to think like a human pilot in order to be safer than a human pilot. I don't believe that to be the case.

 

Why not?  Let's consider two scenarios, specifically the two different types of "computers" we are discussing.

 

1.  The "machine" computer.  If it fails, would it RECOGNIZE it's own failure and be able to take the appropriate action to "save" the aircraft?  Yes, you could have "backup" computers that run cross-checks against each other to "recognize" when one of them may be malfunctioning (we actually do that today in some computer controlled systems)  But the possibility still exists that ALL the computers could fail at the same time, for a variety of reasons.

 

2.  The "Human Pilot" computer.  Yes, in this case I'm thinking of the Human pilot as a "computer" too.  That's what our brains basically allow us to do...analyze, make judgments, take actions, etc.  What happens when the Human computer fails...the pilot has a heart attack, the pressurization system fails at high altitudes and the pilot passes out (which has happened in the past resulting in the total loss of the aircraft and occupants).  Would that Human "computer" recognize it's own failure when it passed out or "died"?  Would the "machine" computer be FULLY capable of saving the aircraft and occupants then? For commercial airline flight (in most cases) we also provide multiple Human pilots ("computers") to back each other up, just in case the other Human computer is "failing" or totally fails too.

 

The crux of the question is WHICH "computer" do you want to be available when the OTHER computer fails.  If one of the computers is a machine, and the other is a Human pilot, I want the "Human computer" as the backup.  It will normally...not always, but normally...have more capability in reasoning and applying judgmental skills in a "non-typical" or "non-normal" scenario than the "machine" computer can be programmed for.  Yes, there are always exceptions, like the "human error" in something like the Control Room humans that ultimately caused the Chernobyl reactor tragedy.  But if you think the "machine" computers are that much better than the human computers, then the machine computers should have prevented the Humans from allowing it to happen in the first place (which, admittedly we DO allow some machine computers to do today, like the computers on an F-16 that will prevent the pilot's control yoke inputs from over-stressing the aircraft.  But in THAT type of configuration, the machine computer is acting as the BACKUP for the PILOT...not vice-versa).    

 

(As an aside for the Mods who may be following this thread.  I don't think there is any "personal attacking" going on in this thread.  It appears to be a rational discussion about a topic that may have varying points of view and opinions.  I hope nobody feels the need to lock the thread because of "disagreements" expressed in it.  That is what debate is about, as long as it stays respectful and nobody is "attacking" an individual personally, which I don't see happening).

Edited by FalconAF
Link to post
Share on other sites

Personally I'm enjoying this thread and I hope it continues. Rational debate makes us stronger and if we all had the same opinion the world would be a boring place. But I think I have reached the limit of my intellectual ability on this subject so I'll let you guys continue ;) 

Link to post
Share on other sites

Falcon, I'll grant you that a dumb AI is going to lack adaptability. Look at it this way though, if the system is robust enough that unexpected situations (i.e. those that the system isn't programmed to deal with adequately) are rarer than the frequency with which humans also fail to adapt to crises then that's a better pilot than a human. The situations where humans and AI systems fail may turn out to be quite different to one another. But if the net result is that fewer people die or other critical incidents occur then that's a win. It's my contention that this will be possible, not right now, but in the not too distant future.

 

I would be more worried about unscrupulous operators than with the general concept of computers flying planes. It's conceivable that an operator may, for example, prioritise their on-time record at the expense of taking greater risks. In such cases human pilots may push back and strike, but of course AI systems will do what they're told unless we have decent external oversight.

 

Oh, and I too have no problem with disagreeing. Ideas are always best tested by healthy debate.

Edited by paralipsis
Link to post
Share on other sites

I have heard today on tv that a private taxi service with a quad engine drone is actually in testing. It has an autonomy of about an hour, can reach a speed of 100 km/h. The company assure that's very safe, anyway i don't know if i could catch one.... maybe if it has also the ability to swtich in manual mode, if something goes wrong eh eh eh ;)

About cars that drives by itself it seems already a near future, there will be also a championship and technician says that those electric cars drives better than a real pilot, these cars are able to make better results in a circuit, this is not very strange because they are programmed to always perform the best trayectories at the right speed. It's clear anyway that all this is fruit of the human experience.

 

 

Link to post
Share on other sites
49 minutes ago, Maurizio Giorgi said:

About cars that drives by itself it seems already a near future, there will be also a championship and technician says that those electric cars drives better than a real pilot, these cars are able to make better results in a circuit, this is not very strange because they are programmed to always perform the best trayectories at the right speed. It's clear anyway that all this is fruit of the human experience

 

This is true, and it's the distinction I was trying to make about driverless cars vs pilotless aircraft.  If the "system" for the cars "glitches", one of the safety backups you CAN have the computers do is STOP ALL the cars and wait for a "fix" to the system, kind of like emergency braking systems on a elevator.  You could also program the computers to use their sensing devices to only do it in the immediate locality of where the malfunctioning car is...no need to stop cars 100 miles away.  But you can't do that with a flying system.  There is no stopping in mid-air in an airplane.

 

And interestingly, one of the main controversies still surrounding driverless cars is SHOULD there be a way for the human occupant to "take control" if the system malfunctions.  Should you allow the human to take over if they THINK the car is going to drive off a cliff, run over a pedetrian, or even hit another car?  Or even if it became obvious any of those things WAS going to happen?  It's a perplexing problem, because the human could potentially abuse that authority and make the OVERALL transportation system LESS SAFE for everyone if they did.  Is it safer to risk the very low probability that the car would actually drive off a cliff and kill the unfortunate occupants of that ONE malfunctioning car/computer, vs preventing ANY override capability of the human in ALL cars to ensure the NORMALLY safe operation of the overall system?  Do we allow a malfunctioning machine/computer to have the final "decision" in taking one human life to protect the rest of the humans using the system?  Do we let a machine/computer decide to run over and kill 5 adults instead of 1 child?  Currently...and nowhere in the foreseeable future...is there a way to program that kind of decision-making into a computer, unless you want to use the "pure logic" approach of the famous Mr Spock Star Trek quote, "The needs of the many outweigh the needs of the few, or the one."  But pure logic and morality always eventually come into conflict when you are taking about making decisions about the Human Condition. 

 

The questions and answers aren't always just about the science and capability of the computer system under normal OR adverse conditions.  There are moral issues that will be brought up too, and those can be even harder to answer.

 

Edited by FalconAF
  • Upvote 1
Link to post
Share on other sites

So what functions will humans have in this new world of uber-technology? 

 

An arm lifts us out of bed in the morning and takes to the shower already running.  It says, "Open your mouth," and brushes your teeth for you?  It feeds us a preprogrammed nutritional (tasteless) food for each meal?  It reads us the latest best selling novel written by artificial intelligence?  Sex is for recreation only and DNA programmed babies come out of test-tubes and incubators?

 

Everything done for us by robots?  Will we even need an education in atechno run future? 

 

Take these driverless car instance.  How much freedom and FUN are we sacrificing for safety and security?  Do we want to live in an absolutely safe and sterile world?  Or do we want to have some fun and enjoyment in our lives...like driving cars for instance...and accepting the risks that go along with life?  I choose the latter.

 

I and my generation have seen a remarkable change during our lives and I view it with suspicion and  sadness.

 

I played cowboys and Indians and cops and robbers when I was kid with cap guns out in the streets and empty lots.  We got hurt sometimes, but that went along with the territory. 

 

We played touch football and baseball and basketball by ourselves without adult supervision or organized leagues.  We chose up sides and didn't care some who weren't picked for the team were left out.  And when we won we gloated.  The concept of everyone being a winner by simply participating didn't exist then.  Winning and gloating added zest to your life and if you didn't win there was always next time.  When you lost you learned to live with disappointment.

 

Mom called is in from the outdoors for dinner; not from the computer room.  Our entire bodies got exercise, not just our fingers working keyboards.

 

Sue mentioned climbing trees a few posts back on some other thread.   Do kids still climb trees?  And fall out of them?  And break their arms?  That's hard to do with full immersion computer programs and virtual reality goggles.  We got dirty after school but every other kid in our class didn't need an inhaler for asthma.

 

The change between generations is accelerating at a frantic pace.  I, for one, can't keep up.  I struggle keeping my flight simulator up to date.

 

My great grandfather went from the horse and buggy to the Model T.  My father was six years old with Wilbur and Orville made their first flight.  He lived to see man walk on the moon...just barely.  I went from Studebakers and Hudsons and De Sotos and Packards to Hyundais and Toyotas and Nissans and Lexuses,  From starting a car with a foot switch to just pressing a button with a fob in my pocket.  Does anyone recall having a choke knob on the dash?  The Luscombe I learned to fly in had no electronics at all.  No radios...not even a starter.  You had to prop the engine.  That was fun. 

 

Flying was fun in DC-3s and Connies.  You could see the ground and houses and cars on the roads while enjoying a nice dinner of steak and potatoes on your tray.  Now they put you into a 600 mile an hour toothpaste tube, give you a bag of peanuts, and when they open the door you're somewhere else.

 

The wife of a friend of mine is an elementary school teacher and she said many kids entering school today have to be taught how to tell time on an analog clock.  How long are they going to last?  I carry an analog pocket watch and when I wear a wristwatch it is analog. 

 

Sorry for the long winded ramble, but I'm a simple analog man stuck in a complex digital world racked by the wake turbulence left behind by a society and technology that's moving much too fast.

 

Noel

 

 

  

  • Upvote 1
Link to post
Share on other sites

I really like to drive a car Noel, for this i've never bought a car with the automatic changing gears, anyway driving in the trafic, like nowadays we're all used to do, it's not funny. Making 1 km in 40 min, it's not funny and it's not drive. Under this aspect i would like an automatic car, but always with the ability to be driven also manually, if you like.

I think that robots never will substitute a man, i mean that a world of robots where a man is only a contour, i think it's not very desirable it happens. But technology will help us more and more, i think that technology would be helpful for this.

 

  • Upvote 1
Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...