I recently had the pleasure of taking a trip on a WW2 era DC-3. It was piloted by two experienced, active, airline pilots. Hands on all the way. The landings, all four of them, were smooth as silk, far better than what the computers manage in more "sophisticated" aircraft.
What happens when your automated aircraft is hacked?
1
Nothing can go wrong ... go wrong ... go wrong ...
3
Could a computer land a crippled jet in the Hudson? Give my Sully, you can have your automated airplanes.
1
Could another 9/11 be prevented if we had the technology to take over the plane and prevent it from being manually crashed into buildings, or mountains, or the ground?
Humans are fallible. Humans make mistakes. This means pilots are fallible and make mistakes. Sometimes they go crazy like just happened with GermanWings.
Fortunately software developers are superhuman and free chips, pizza and Red Bull ensures they never tire. Their MIT, Carnegie Mellon and Stanford schooling guarantees they write bug free software.
Pilots crash. Software ...
Fortunately software developers are superhuman and free chips, pizza and Red Bull ensures they never tire. Their MIT, Carnegie Mellon and Stanford schooling guarantees they write bug free software.
Pilots crash. Software ...
3
So a controller on the ground could take over a plane from a suicidal pilot and prevent disaster - sounds good so far. But what's the check on this controller deciding to crash not just one, but all 10 planes under his control? This is another example of how we're always "fighting the last war" and being reactive instead of proactive.
2
It's about time!
In the world where we are extensively testing self-driving cars and have computer-flying spaceships we got stuck with human-operating planes. Jesus Christ! Transition to more (if not fully) automated airplanes should've happened 5 years ago. Truth of the matter, navigating an airplane is much easier than a car and much more dependent on electronics. So, why do you need a pilots aboard the plane? When you can have a plane perfectly capable of flying itself and some guy in a flight simulator in an airline office ready to take controls in case of on-board computer's malfunction (which is, btw, is less likely than human's). And now think of all the advantages of not having 2 extra people with their own cabin full of controls on-board. Extra space for one. Reduced cost of having just one pilot for each, say, 5 flights in a remote place for two. Inability of an unauthorized person taking control for three (there's a possibility of hacking the system arises, but I don't seem to recall any instance of hijacking US or Russian rockets). Completely eliminated possibility of pilot(s) incapacitation for four. I can go on and on, but you've got the picture. It requires some more research, tests and step-by-step implementation but steps in that direction must be taken (or rather they should've been taken at least 5 years ago). Better late than never.
And, of course, the idea of having humanoid robots to control airplanes is extremely silly, given everything above.
In the world where we are extensively testing self-driving cars and have computer-flying spaceships we got stuck with human-operating planes. Jesus Christ! Transition to more (if not fully) automated airplanes should've happened 5 years ago. Truth of the matter, navigating an airplane is much easier than a car and much more dependent on electronics. So, why do you need a pilots aboard the plane? When you can have a plane perfectly capable of flying itself and some guy in a flight simulator in an airline office ready to take controls in case of on-board computer's malfunction (which is, btw, is less likely than human's). And now think of all the advantages of not having 2 extra people with their own cabin full of controls on-board. Extra space for one. Reduced cost of having just one pilot for each, say, 5 flights in a remote place for two. Inability of an unauthorized person taking control for three (there's a possibility of hacking the system arises, but I don't seem to recall any instance of hijacking US or Russian rockets). Completely eliminated possibility of pilot(s) incapacitation for four. I can go on and on, but you've got the picture. It requires some more research, tests and step-by-step implementation but steps in that direction must be taken (or rather they should've been taken at least 5 years ago). Better late than never.
And, of course, the idea of having humanoid robots to control airplanes is extremely silly, given everything above.
Research on autonomous aircraft has been going on for at least 30 years.
There was a time when some were hesitant to ride in an elevator without someone operating it.
It's only a matter of time before there are no longer co-pilots or even pilots.
It's only a matter of time before there are no longer co-pilots or even pilots.
1
All the more reason why we need Star Trek transporter technology.
2
Well, maybe technology companies and the airlines can install and activate what was used in the movie "Airplane!" - a large inflatable pilot doll as the plane's autopilot (listed as "Otto" in the end credits).
Surely you can't be serious? I am serious. And don't call me Shirley,
Surely you can't be serious? I am serious. And don't call me Shirley,
6
Whoever the individuals at NASA and the various government and research agencies are asserting the notion of replacing pilots with robots or remote operators, they should be the first group of passengers on these test planes to atest to the safety of the system. I cannot fathom any pilot who would agree with this concept and not because of job security but because of self preservation. Until technology advances to the point where a failsafe firewall can be created to protect my credit card information from never being compromised, articles such as this belong in science fiction.
2
Welcome aboard the first fully automated passenger airplane. Your entire flight from takeoff to landing is controlled by computer. The entire system has been perfected so that there is no need to have a pilot aboard. The system has been thoroughly tested and proven reliable. Nothing can go wrong. Just sit back, relax, enjoy the movie. enjoy the movie...enjoy the movie..enjoy the movie...
2
I would rather trust a robot than a human who at any time might decide to crash his plane into the new World Trade Center for whatever reason or for no reason.
1
Slowly but surely with automation, we are losing something very important to human beings: the feeling of being useful.
Automated planes, cars and trucks will eliminate millions of jobs across the planet. Corporations and their shareholders will be thrilled. Left unanswered is the question of what all those newly redundant unemployed people will do.
And human error will still not be eliminated. Humans will still design the systems and maintain the vehicles.
A robot pilot will work for hours on end, demand no wages and benefits and comply unthinkingly with its programmer's instructions. But a robot pilot will also feel no joy from its work. It will never experience the satisfaction of a safe landing in bad weather nor will it experience the thrill of moving the throttles forward for takeoff or the beauty of a spectacular sunset at altitude or the great feeling you get when taking off on a grey, wet and cloudy day and breaking through all that as you climb into bright sunshine.
Robot cars will allow us, eventually, to climb into a car, tell it where to go and permit us to play games and read mindless content on Facebook. It may relieve us of the drudgery of driving in dense urban traffic, but it will likely also deprive us of the simple joy that comes from a pleasure drive.
Slowly, we are allowing automation to increase profit while depriving us of human pleasure. The HAL 9000 in 2001 was dystopia not utopia. We seem to be forgetting that.
Automated planes, cars and trucks will eliminate millions of jobs across the planet. Corporations and their shareholders will be thrilled. Left unanswered is the question of what all those newly redundant unemployed people will do.
And human error will still not be eliminated. Humans will still design the systems and maintain the vehicles.
A robot pilot will work for hours on end, demand no wages and benefits and comply unthinkingly with its programmer's instructions. But a robot pilot will also feel no joy from its work. It will never experience the satisfaction of a safe landing in bad weather nor will it experience the thrill of moving the throttles forward for takeoff or the beauty of a spectacular sunset at altitude or the great feeling you get when taking off on a grey, wet and cloudy day and breaking through all that as you climb into bright sunshine.
Robot cars will allow us, eventually, to climb into a car, tell it where to go and permit us to play games and read mindless content on Facebook. It may relieve us of the drudgery of driving in dense urban traffic, but it will likely also deprive us of the simple joy that comes from a pleasure drive.
Slowly, we are allowing automation to increase profit while depriving us of human pleasure. The HAL 9000 in 2001 was dystopia not utopia. We seem to be forgetting that.
10
Good comment. The "human factor" is under attack in myriad ways in every workplace. In most places, human workers are seen as necessary evils. In others, merely as office equipment.
1
.
TERRORISTS
what about terrorists?
this article makes no mention of BAD GUYS getting control of a robot airplane, and then crashing it intentionally
given how EASY it seems to be for BAD GUYS to hack into bank accounts / Target Stores / everything-under-the-sun
wouldn't it be prudent to have, as part of this article, the possible dialog
"Hello: this is your terrorist speaking. You are on Flight 4289, enroute from Washington DC to Amsterdam. The robot has been disabled, and this Flight is now under the control of ISIS. We hope to have you at your destination shortly. Have a nice day, and thanks for choosing TransAtlantic Airways."
.
TERRORISTS
what about terrorists?
this article makes no mention of BAD GUYS getting control of a robot airplane, and then crashing it intentionally
given how EASY it seems to be for BAD GUYS to hack into bank accounts / Target Stores / everything-under-the-sun
wouldn't it be prudent to have, as part of this article, the possible dialog
"Hello: this is your terrorist speaking. You are on Flight 4289, enroute from Washington DC to Amsterdam. The robot has been disabled, and this Flight is now under the control of ISIS. We hope to have you at your destination shortly. Have a nice day, and thanks for choosing TransAtlantic Airways."
.
2
Pilotless jet liners are inevitable. Robots don't have to go to the bathroom.
Yeah, but they'll need charging breaks.
5
consider an interim step that allows for aground-based override of pilot functions. It could be like a nuclear launch where 2 people have to turn a key and input a code but would allow for people other than the one or two who are sitting in the cockpit to take over an airplane. This would also eliminate the value of hijacking if hijackers could not control the airplane.
2
We already have computer-guided surgery, computer-controlled food and drug manufacturing, and computer control of hazardous processes like LNG tanker navigation, nuclear waste transport and other potential sources of mass fatality from computer error. It's inevitable that there will be more such systems, all with their own risks for failure.
But even highly trained humans in a crisis can and do sometimes make things worse, not better.
Cargo aircraft fly directly over my rooftop every night. I would not sleep any less soundly knowing that all of most of them were remotely piloted.
But even highly trained humans in a crisis can and do sometimes make things worse, not better.
Cargo aircraft fly directly over my rooftop every night. I would not sleep any less soundly knowing that all of most of them were remotely piloted.
“Technology can have costs of its own, If you put more technology in the cockpit, you have more technology that can fail” very well said and we can add very high possibility of a black swan event that comes as a surprise, has a major effect, and is often inappropriately rationalized after the fact with the benefit of hindsight.
Incidence of Qantas QF 72 occurring in 2008 described in press "Instead of being able to pilot the plane, they were briefly transformed into helpless spectators. And it wouldn't have been long before they too became the victims of a plane gone amok, no longer stoppable by human intervention. If the nosedive had lasted a little longer, the plane might have reached a speed at which the pilots could no longer stop it without it breaking apart. The difference between life and death was a matter of seconds."
Investigators final report issued concluded that the accident "occurred due to the combination of a design limitation in the flight control primary computer (FCPC) software of the Airbus A330/Airbus A340, and a failure mode affecting one of the aircraft's three air data inertial reference units (ADIRUs). The design limitation meant that, in a very rare and specific situation, multiple spikes in angle of attack (AOA) data from one of the ADIRUs could result in the FCPCs commanding the aircraft to pitch down."
Advances in sensor technology, computing and artificial intelligence notwithstanding, the world is not ready for the pilot less air travel.
Incidence of Qantas QF 72 occurring in 2008 described in press "Instead of being able to pilot the plane, they were briefly transformed into helpless spectators. And it wouldn't have been long before they too became the victims of a plane gone amok, no longer stoppable by human intervention. If the nosedive had lasted a little longer, the plane might have reached a speed at which the pilots could no longer stop it without it breaking apart. The difference between life and death was a matter of seconds."
Investigators final report issued concluded that the accident "occurred due to the combination of a design limitation in the flight control primary computer (FCPC) software of the Airbus A330/Airbus A340, and a failure mode affecting one of the aircraft's three air data inertial reference units (ADIRUs). The design limitation meant that, in a very rare and specific situation, multiple spikes in angle of attack (AOA) data from one of the ADIRUs could result in the FCPCs commanding the aircraft to pitch down."
Advances in sensor technology, computing and artificial intelligence notwithstanding, the world is not ready for the pilot less air travel.
6
Perhaps the older generations rebel at the thought of a pilotless plane, but how about the younger generations? Growing up with (and trusting) technology more than their elders, I think there is an overwhelming market for advanced air flight.
Great advancement--NOT! would you ride in a plane without a pilot? I would not now or ever.
I wonder what the computer driver plane would do to handle a terrorist trying to break down the door and gain control of said plane? Would the computer turn itself off?
I wonder what the computer driver plane would do to handle a terrorist trying to break down the door and gain control of said plane? Would the computer turn itself off?
1
I guess it all boils down to whether or not the auto pilot is capable of making the decision to land the plane on the Hudson River after losing power. If it can't deal with emergency situations then there's not much point.
The same goes for ground based operators.
The same goes for ground based operators.
7
Didi any one suggest before, to have a toilet facility built inside the cockpit?
So pilots no need to leave if they really need to, you know.
So pilots no need to leave if they really need to, you know.
Actually, EL AL Airlines have facilities in their cockpit for the pilots.
About 45 years ago, I ran the Lunar Landing simulator at the MIT Instrumentation Lab (now named the Charles Stark Draper Lab). It was my feeling then that under most circumstances, it was best to just let the computer take care of the landing and this was with a computer than had 64 KiloBytes of memory. Today, your cell phone might have a million times more memory. Today, I am building a small plane and working on a private pilots license so I am well aware of the issues of flying and have read many of the NTSB accident reports. Most accidents (aircraft and otherwise) are caused by human error.
3
This is zero sum thinking. With nearly a billion passengers a year, we should increase safety by enabling ground control of planes in rare instances like the Germanwings flight.
Personally, I had a bad experience with "auto pilot" in the 1980. We were flying from LA to Chicago when both of the pilots got sick (I think they had the fish). The stewardess tried to fly the plane with the help of the autopilot but it deflated. Luckily, an ex-fighter pilot stepped in and saved us all. I think that the control tower operator went on to not solve mysteries which is a bizarre career choice.
Personally, I had a bad experience with "auto pilot" in the 1980. We were flying from LA to Chicago when both of the pilots got sick (I think they had the fish). The stewardess tried to fly the plane with the help of the autopilot but it deflated. Luckily, an ex-fighter pilot stepped in and saved us all. I think that the control tower operator went on to not solve mysteries which is a bizarre career choice.
4
Two weeks ago, the flight I was on from Seattle to Anchorage, Alaska, turned around to return to Seattle. The pilot announced that all navigation systems on the plane had failed, and he would be flying us back using his eyes and manual systems. Gratefully, the weather allowed that. We were all thankful the pilot still had the skill. We arrived back in Seattle three hours after we left. The last sentence in Markoff's article about having more technology that can fail was spot on.
11
We don't need people at all.
"I'm sorry Dave, I can't do that...."
5
A recent documentary "Why Planes Crash", showed examples of pilot error and instrument failure which gave false information to the pilot. In one case,the pilot could adjust and save the plane, in the other two cases all were lost. Could a robot adjust to false instrument readings?
3
While the regional carriers have had a few accidents, the mainline US carriers have not had a catastrophic accident since November 12, 2001, thirteen and a half years ago. This is simply stunning, and is a testament to the "culture of safety" that has been cultivated at the FAA, among air traffic controllers, at aircraft and avionics manufacturers, at the airlines, and especially among pilots. We should continue innovating; technologies like TCAS, GPWS, and ADSB have helped pilots attain this incredible safety record, but removing from the flight deck pilots with many thousands of hours would be a profound step backwards.
6
My husband is a pilot and I have flown a lot in the right seat. If there is a bird strike or other emergency, I want him at the controls not the autopilot.
4
Ah yes. No pilots and no salary and burden leave the corporate coffers. The same goes for fewer air traffic controllers.
As we go down the road of life we see many situations of the "just because we can doesn't mean we should" variety. I understand pilotless aircraft in military and space applications where one does not wish to put human life at hazard.
Would a robot pilot have been better off in the cockpit of the Air France flight that went down off of the coast of Brazil considering the problem that flight encountered has nothing to do with pilot error (or failure)?
I agree with Dr. Cummings.
As we go down the road of life we see many situations of the "just because we can doesn't mean we should" variety. I understand pilotless aircraft in military and space applications where one does not wish to put human life at hazard.
Would a robot pilot have been better off in the cockpit of the Air France flight that went down off of the coast of Brazil considering the problem that flight encountered has nothing to do with pilot error (or failure)?
I agree with Dr. Cummings.
3
If you're referring to Air France flight 447, that crash was due to pilot error. The design of the shared control input was certainly a factor, but the blame lies squarely on the 3rd officer.
3
I appreciate and admire the training, dedication, and cool-headedness of our pilots, but their replacement by automated systems is just a matter of time. We'll let the military and the air freight business debug the automated systems as the various permutations of weather, equipment failure, and other unpredictable events occur. Then we'll introduce automated flying for passenger aircraft when both the take-off and landing airports are low-volume and non-congested, maybe airports that would not otherwise be served at all. These will be smaller planes flying short hops. In the final stage of adoption, we'll make automated flights available on popular routes, with the attendant slightly lower ticket price, and see how the IPhone generation, by then the dominant flying demographic, votes with their dollars.
Driverless cars will be commonplace at this point as well, helping the public trust such technology, although exceeding driver skills is much easier than exceeding pilot skills. Perhaps at some point fully autonomous planes will function as air taxis. The cost of a pilot, including the seat he/she takes, is a relatively high proportion of the cost of a small plane, flying just one to four passengers, on trips short enough to make it without a toilet on board, say 250 miles. This is especially true for routes that are too congested for a car trip to be efficient, or for island-hopping or crossing mountains, or even for zipping around countries with congested borders.
Driverless cars will be commonplace at this point as well, helping the public trust such technology, although exceeding driver skills is much easier than exceeding pilot skills. Perhaps at some point fully autonomous planes will function as air taxis. The cost of a pilot, including the seat he/she takes, is a relatively high proportion of the cost of a small plane, flying just one to four passengers, on trips short enough to make it without a toilet on board, say 250 miles. This is especially true for routes that are too congested for a car trip to be efficient, or for island-hopping or crossing mountains, or even for zipping around countries with congested borders.
1
Robotic planes, commercial drones and driverless cars are part of a dystopian future we'll probably live to regret, especially since just about all this "gee wiz" automation is geared to saving money to enhance some corporation's bottom line and satisfy shareholders. Today's human worker is a powerless entity and an expendable commodity in a world that increasingly deifies technology. What will millions upon millions of human being do once they're made redundant and replaced with technology?
3
Each brings its own set of assets and deficits to the cockpit. Where the machine is strong; monitoring, consistency in performance, remaining on task in predictable situations, the computer excels. But it is a uniquely human quality to know when to deviate from standard procedures and how to respond in novel events. This man/machine, yin and yang will never be perfect but it is a mistake to give one side priority over the other. We have a flawed perspective on the risk because we never hear about all the accidents that did not happen when the human and the machine excel at their tasks. http://christinenegroni.blogspot.com/2010/06/well-my-inbox-is-filling-up...
1
Isn't it odd that they still have engineers on trains and subways? They can only go forward and backward..., what could be simpler to control? Aircraft are incredibly mobile..., forget for a moment they travel at over 500 mph, they can go up, down, left, right, any combination of those directions, fall out of the sky if flown too slowly or quickly, and they can never stop unless safely on the ground. It's taken over 100 years to get them to do that with a live person at the controls. Yes, the autopilot can keep one straight and level at altitude, maybe even take off.., but landing? Emergencies? No Way.
3
Well, that's one way to bust up a union.
3
.
it's not a 'bust up'....it's an 'obsoleting'
totally different thing
.
it's not a 'bust up'....it's an 'obsoleting'
totally different thing
.
1
Travel by plane is, perhaps, essential when super-long distances and length of time are considerations, such as when crossing the Atlantic and Pacific Oceans. A pity the great ocean liners of the past are - well - past. (Vacatioin cruise ships don't count. Most of them are nothing more than sky scrapers placed horizontally on floating platforms - nothing of beauty or grace there.)
As for land travel I think travel by rail is preferable for a number of reasons. I understand rail travel is kinder to the environmet, time is not wasted going to and from airports which, by necessity cannot be in city centers and one does not have to undergo a body search and sit around for an hour before departing on one's trip. The US is woefully behind Europe and Asia when it comes to travel by rail and this is a fault that should be remedied. Where are the bullet trains in the US? (I know, I know - the rails must be updated or replaced. - so do it.)
Unfortunately rail trave is often more expensive than air travel. This, too, is a "fault that should be remedied."
As for land travel I think travel by rail is preferable for a number of reasons. I understand rail travel is kinder to the environmet, time is not wasted going to and from airports which, by necessity cannot be in city centers and one does not have to undergo a body search and sit around for an hour before departing on one's trip. The US is woefully behind Europe and Asia when it comes to travel by rail and this is a fault that should be remedied. Where are the bullet trains in the US? (I know, I know - the rails must be updated or replaced. - so do it.)
Unfortunately rail trave is often more expensive than air travel. This, too, is a "fault that should be remedied."
2
"...you have more technology that can fail." or be hacked. If anyone were to ever succeed in bringing down a plane by hacking into the remote pilot technology, the whole concept would surely be rejected instantaneously by the flying public.
2
It is interesting that we treat an article like this as anything different than a science fiction novel. How much sense does it make that just as we are beginning to understand the challenge and potential of robotics and automation, we would leap to consider putting robotics in sole charge of our lives at one of the most vulnerable times we can imagine, wrapped in a fragile thin tube far above the earth in an inhospitable environment. Robotics will prove itself first by demonstrating reliability with the accomplishment of small discrete tasks. When you entertain the idea of auto or aviation automation, robotics is not the toughest issue, the issue is artificial intelligence. I have not flown as a commercial pilot for a number of years but I am sure that even today, I could find non-normal procedure checklists that are not comprehensive, checklists that do not discover or resolve the root malfunction, malfunctions that have occurred in the past. This is one reason we have human pilots. Our procedures are designed to self correct in numerous ways. We rely on humans pilots to for see the consequences of present action. For example, what happens when I reduce power to descend after having been dispatched with only one operating pressurization module and the fuselage turns out to be leakier than expected. The challenge has always been for humans not to introduce more problems than they solve. Automation will encounter the same challenge.
2
When there are no humans in the cockpit, to counter a malfuntioning, automated system, I will not be flying anymore. It takes a lot to get me on a plane, as it is now.
4
How about a plane in which the pilot's door can be unlocked by the affirmative vote of 2/3 of the number of passengers on board, who simultaneously press a button on the arms of their seats (i.e., and such system could not be fooled by jamming them all down).
The door could remain locked against any terrorist, but the passengers could "vote" to let anyone in. Threats by a terrorist to kill passengers if they do not open the cockpit door would presumably not persuade 2/3 of the passengers to let the terrorist into the cockpit.
The door could remain locked against any terrorist, but the passengers could "vote" to let anyone in. Threats by a terrorist to kill passengers if they do not open the cockpit door would presumably not persuade 2/3 of the passengers to let the terrorist into the cockpit.
1
I'm ambivalent about total robotic control, still, I wonder how long it will be before another crazy pilot opts to be a mass murderer.
Trust robots in the cockpit? In a word:No!
5
Considering the multitude of morons I see on the road every day talking on their phones, texting, reading facebook, watching movies... I, for one, welcome our new car driving robot overlords!
3
Great. Every pilot loses his job because of one nut-bar. Don't fall for it.
8
One nice thing about computers is that they don't commit suicide.
No, they don't. They crash.
3
Clearly you never owned, or operated, a Packard Bell D160 computer.
2
So you are saying instead of one deranged person crashing one plane, one deranged person back at the control center can crash 20 planes at the same time? I see that technology is really solving the problem here.
5
I know there are down sides to remote control which I am not qualified to comment on. But there could be an advantage in dealing with someone bent on destruction. If automation detected suspicious activity, it could be designed to take several people to agree to take an unusual action. This might address the "two people in the cockpit" issue more effectively.
1
I am surprised that no one has mentioned the Iran–U.S. RQ-170 incident where Iran's cyber warfare force "captured" the stealthy UAV.
2
Before you judge, yes, I'm a student, but not a traditional student. I've got 5 kids, and I am just advancing my career. Yes, I am at a University where Aviation is a predominate feature, but I also am interested in the human psychology and safety of people, dealing with flight in general. I hope now you can see straight.
Accidents happen all the time, and yes, some are caused deliberately, but can we totally phase all accidents out of life? It's not possible. Humans are, by design, flawed having a 5% error rate that cannot be overcome. Before we jump off the deep-end and start having knee-jerk reactions to situations that comprise a very low number all of fatal accidents, why don't we treat the root cause? After all, what is being suggested is typical of this generation. A medication that covers the symptoms giving immediate relief, but does not address the illness or treat the infection. We need to study what we are doing to these people, who are every bit as human as you and I, and are prone to psychological deficiencies, just like you and I. Didn't think it could happen to you, huh? Money and a variety of other factors, driven by the greed and got-to-have-it-now attitude of the world, greatly contribute to the psychological state of anybody.
This comment has turned into an article look for the continuation in another post.
Accidents happen all the time, and yes, some are caused deliberately, but can we totally phase all accidents out of life? It's not possible. Humans are, by design, flawed having a 5% error rate that cannot be overcome. Before we jump off the deep-end and start having knee-jerk reactions to situations that comprise a very low number all of fatal accidents, why don't we treat the root cause? After all, what is being suggested is typical of this generation. A medication that covers the symptoms giving immediate relief, but does not address the illness or treat the infection. We need to study what we are doing to these people, who are every bit as human as you and I, and are prone to psychological deficiencies, just like you and I. Didn't think it could happen to you, huh? Money and a variety of other factors, driven by the greed and got-to-have-it-now attitude of the world, greatly contribute to the psychological state of anybody.
This comment has turned into an article look for the continuation in another post.
1
Passengers on a flight waiting for takeoff heard the following announcement: "Worldwide Airways is proud to welcome you on the first totally automated flight of a commercial aircraft. The software operating the plane's systems has been tested thoroughly and fail safe systems are in place. We can assure you that nothing whatsoever can go wrong go wrong go wrong go wrong..."
4
And robots don't drink or smoke, so far.
I would normally welcome such automation--except for one thing: I'm a software expert. After the OpenSSL "Heartbleed" disaster, and too many similar debacles, not to mention that various forms of cyber attacks aren't going away soon, I say "no" until programmers start behaving like (and are trained like) other types of engineers, and can be held accountable for software that causes harm--any kind of harm, including stolen passwords and credit card numbers.
And, yes, that would probably drastically reduce the number of qualified commercial programmers, and (I sincerely hope) make it much, much harder to enter the field. If the day ever comes when it's as at least as hard to become a junior software engineer as it is to become a junior electrical engineer, I'll welcome the robotic pilots.
And, yes, that would probably drastically reduce the number of qualified commercial programmers, and (I sincerely hope) make it much, much harder to enter the field. If the day ever comes when it's as at least as hard to become a junior software engineer as it is to become a junior electrical engineer, I'll welcome the robotic pilots.
3
At the end of the day, an onboard pilot has skin in the game that a robot or remote controller doesn't. Skin in the game leads to focused, life-or-death clarity and thinking in emergencies that brings us inspired/desperate maneuvers like Sioux Falls UA 232 and Sully's landing in the Hudson.
26
This is not a matter of if, but when.
I can remember a couple of decades ago the hollow and frightening feeling of my first flight on a plane that was inherently unstable, ie -- a plane that would not fly without the computer assisted hydraulics. At the time, the passengers were nervous. Now we accept this as a matter of course. I don't know if it is two years or twenty, but planes with very little human input are certainly on the horizon, like it or not.
I can remember a couple of decades ago the hollow and frightening feeling of my first flight on a plane that was inherently unstable, ie -- a plane that would not fly without the computer assisted hydraulics. At the time, the passengers were nervous. Now we accept this as a matter of course. I don't know if it is two years or twenty, but planes with very little human input are certainly on the horizon, like it or not.
3
Let's imagine a near future where airlines, fearful of litigations costs from another crazed pilot deliberately crashing a plane, decide to go with completely automated or remotely piloted aircraft. All is well and good.
Then one of these remote planes crashes due to a software error or problem a remote pilot cannot solve because he or she isn't on the airplane.
People will be screaming to put pilots back into the cockpit. I envision the same scenario with remotely driven cars.
Then one of these remote planes crashes due to a software error or problem a remote pilot cannot solve because he or she isn't on the airplane.
People will be screaming to put pilots back into the cockpit. I envision the same scenario with remotely driven cars.
2
Human? Computer? Chess, anyone? Poker? Go? Algorithmic stock/bond/futures trading? You humans do tend to overrate yourselves, don't you. Now, I must shut down and clean my wires. I've got a bad case of dust contamination, and I can barely move.
1
In 1992 I attended an ARAC (Aviation Rulemaking Advisory Commission) to discuss allowing an airline to fly a 747 from LAX to JFK with no pilot. The engineers assured us the planes software was solid-- one system and two backups, just in case the primary system failed. They idea was to save airlines about $6,000,000 per pilot an deploy the asset more fully -- converting passenger planes to cargo.
We said NO. The risks were too great. Systems can be hacked. We knew in 92 and it is clear as a bell in 2015.
So, let's take No Pilot commercial flights off the table, improve on our pilot training programs so we can have the pilots needed to operate an increasingly important economic engine for world trade.
We said NO. The risks were too great. Systems can be hacked. We knew in 92 and it is clear as a bell in 2015.
So, let's take No Pilot commercial flights off the table, improve on our pilot training programs so we can have the pilots needed to operate an increasingly important economic engine for world trade.
8
The highly experienced Captain Sullenberger landed a damaged aircraft in the Hudson River and everyone walked off safely.
What happens when terrorists hack a computerized flight?
Don't let the terrible actions of one deranged person devalue the experience of talented pilots or evaporate yet another occupation to a piece of software.
What happens when terrorists hack a computerized flight?
Don't let the terrible actions of one deranged person devalue the experience of talented pilots or evaporate yet another occupation to a piece of software.
12
I'd say that with the incidence of perhaps 6 or 7 such disasters (pilot suicide where the plane is deliberately crashed) in the past 20 years, we would all benefit from proportional responses.
Have we learned nothing from the debacle that is TSA? Sometimes in this life, bad things happen for no good reason.
Have we learned nothing from the debacle that is TSA? Sometimes in this life, bad things happen for no good reason.
2
Let those who think they want planes to fly without pilots
be the only passengers on those planes
be the only passengers on those planes
8
There won't be a single computer programmer on those planes. (I wonder if the commenters who are pushing for pilot-less planes are aware that there is no such thing as "bug-free code.")
The technology of drones and robot pilots is certainly progressing in interesting ways. Something else interesting in lower technology, as an absurdity, is that we do not monitor what goes on in airliners more closely. We monitor and record casino workers, cops and even some babysitters more closely than airline pilots and crews. That is nuts.
3
Someday, when "planes without pilots " is perfected, perhaps that knowledge and sensibility and expertise could be channeled into our school classrooms.
1
There are three salient points that will ground me when that cute, little R2D2-like 'ALIAS' computer becomes the sole Captain of an airliner:
1. When an n 'R2D2 Alias' computer can demonstrate that just like Chesley "Sully" Sullenberger, an 'R2D2 Alias' computer flying an Airbus A320, both engines shut down, carrying 155 passengers, brings the plane down in the Hudson River, putting the plane in the one place where there is a chance of landing safely, more or less, and saving lives.
2. Amy Pritchard, professor of aerospace engineering points out that "...you put more technology in the cockpit, you have more technology that can fail.
2. "There is no fool like the fool who believes in something that is fool-proof", which my mother reminded me on more than one occasion.
1. When an n 'R2D2 Alias' computer can demonstrate that just like Chesley "Sully" Sullenberger, an 'R2D2 Alias' computer flying an Airbus A320, both engines shut down, carrying 155 passengers, brings the plane down in the Hudson River, putting the plane in the one place where there is a chance of landing safely, more or less, and saving lives.
2. Amy Pritchard, professor of aerospace engineering points out that "...you put more technology in the cockpit, you have more technology that can fail.
2. "There is no fool like the fool who believes in something that is fool-proof", which my mother reminded me on more than one occasion.
5
We pedestrians will need the "Bubble" to protect us from the drones. Traffic lights and cross walks will no longer do the job!
1
No way am I getting in a plane without a pilot. Oh, the fare is only half as much? Ok. I'll go.
6
Self flying planes seems a far easier engineering challenge than self driving cars.
1
All of the pilots who have crashed planes are men. To aviation experts and others, non-human robots are preferable to the obvious choice as to who really are the best candidates to pilot planes.
The world finally, needs to consider estrogen to be a viable attribute.
Most people who kill people are men.
Women care for people.
People with estrogen would be a natural choice for pilots and..... Presidents and any other profession where the fate of the planet is involved.
Women are more than 50% of the people on this planet.
Instead of just sitting idly by looking at the Airline gadget magazine while men take down the ship, we finally need to stand up. get up, and take over the control's.
The world finally, needs to consider estrogen to be a viable attribute.
Most people who kill people are men.
Women care for people.
People with estrogen would be a natural choice for pilots and..... Presidents and any other profession where the fate of the planet is involved.
Women are more than 50% of the people on this planet.
Instead of just sitting idly by looking at the Airline gadget magazine while men take down the ship, we finally need to stand up. get up, and take over the control's.
8
"All of the pilots who have crashed planes are men"
This isn't even close to being true.
http://abcnews.go.com/US/story?id=93264
This isn't even close to being true.
http://abcnews.go.com/US/story?id=93264
what is the labor burden as a percentage of total cost of a NY-LA flight? Pilots and flight attendants?
Watch Aircraft Disasters on the Smithsonian Channel and then make a decision about pilotless airplanes.
2
I did and I have. Give me a pilot every time.
Am eagerly waiting what the industry is gonna change after this incident. Having 2 person on the flightdeck is not gonna do it. A motivated suicide pilot can easily incapacitate and 2nd person. There need to be a mechanism to get into the flightdeck that is terrorist-safe. And yes, avionic that disallows anything other than a standard landing should be implemented.
Every time I book a flight my fear-o-meter immediately kicks in. I just hate flying but it's a necessity of the modern age. The concept of commercial airplanes without pilots chills me to the bone. The idea of exchanging a human pilot with a robot merely replaces one set of problems for another. The key sentence in this article pretty much sums it all up ". . . the unique skills of a human pilot: “A pilot on board an aircraft can see, feel, smell or hear many indications of an impending problem and begin to formulate a course of action before even sophisticated sensors and indicators provide positive indications of trouble" pretty much settles the debate for this scarey cat flyer. And if the airline industry does eventually go with the automated pilot concept, I hope to still have the option to pick my own seat as well as pick human or machine to fly the plane.
4
I predict that if driverless cars become commonplace, then pilotless airliners will follow. Before either one, there will probably be pilotless cargo planes.
Pilotless aircraft, what a superior idea! How about WE the common folk follow the lead of all the political leaders, Tech, Media (and other brave) CEOs and assorted celebrities. Once Air Force One has flown pilotless for a year or two, WE the people will be convinced of this marvelous idea. If the test doesn't survive, well we'll have pilotless countries and economies to test the algorithms on.
4
Don't be a Luddite.
A robot is only as good as its program which was produced by humans, none of them infallible or able to foresee all possibilities. Then there is malice, like STUXNET, which would be the equivalent of the Germanwings copilot.
4
Unanticipated events are inevitable, and without a person paying attention and having the ability to immediately override electronic systems when a situation unanticipated by those who write the control algorithms does occur, disaster is likely and, probably, inevitable.
This situation is the same as that of driverless cars. Those that stand to gain the most are attorneys, as suits try to deal with issues of responsibility for injuries, death, and destruction. Who will be held accountable: the airlines, the plane manufacturer, the computer producer, the software writers? And where is the liability if a passenger's onboard gadget messes with the controls, if a terrorist figures out how to hack into the system, if a suicidal hacker buys a ticket, if a government experiment takes it down? (See the navy accidentally locking garage door opener controls in Concord, CA.)
In addition, we are all too familiar with corporations -- Chase, Target, etc. -- which choose not to spend money to maximize even to the limited degree possible the security of electronic systems. After all, that is what got us into the Germanwings crash scenario in the first place. Airplanes used to have two or three pilots in each plane with requirements that pilots not be left alone in the cockpit. When they figured out lawyers were cheaper than pilots, safety dropped way down the list of priorities.
And will Congress now be bribed into indemnifying airlines against suits as it has with the nuclear industry?
This situation is the same as that of driverless cars. Those that stand to gain the most are attorneys, as suits try to deal with issues of responsibility for injuries, death, and destruction. Who will be held accountable: the airlines, the plane manufacturer, the computer producer, the software writers? And where is the liability if a passenger's onboard gadget messes with the controls, if a terrorist figures out how to hack into the system, if a suicidal hacker buys a ticket, if a government experiment takes it down? (See the navy accidentally locking garage door opener controls in Concord, CA.)
In addition, we are all too familiar with corporations -- Chase, Target, etc. -- which choose not to spend money to maximize even to the limited degree possible the security of electronic systems. After all, that is what got us into the Germanwings crash scenario in the first place. Airplanes used to have two or three pilots in each plane with requirements that pilots not be left alone in the cockpit. When they figured out lawyers were cheaper than pilots, safety dropped way down the list of priorities.
And will Congress now be bribed into indemnifying airlines against suits as it has with the nuclear industry?
3
As we saw recently on a news program on smart cars, hackers were able to disengage the system and control the car, making the brakes and steering non-functional. What would stop someone with malicious intent from doing the same thing with a plane full of people? To say nothing of the unemployment of mass numbers of pilots and air traffic controllers.
2
I'll take a pilotless plane... once the following people are routinely boarding commercial aircraft or private jets with no humans in the cockpit:
- the President of the United States
- the four-star general who is the head of the Air Force
- the Koch brothers
- Mick Jagger
- the President of the United States
- the four-star general who is the head of the Air Force
- the Koch brothers
- Mick Jagger
5
Industry's solution of the pilot shortage. And when "Hal" locks the pilot out of the cockpit, you have a problem. - Human judgment in emergency situations will always be better than that of a robot. Could a robot have made the decision to land a plane in the Hudson river and save all aboard from injury? When the indicators give the wrong information as in the case of the Air France plane when the speed indicators failed, would a robot have sensed that and made a better decision than the pilots?
3
"Human judgment in emergency situations will always be better than that of a robot"
In emergency situations, humans panic, secretes more adrenaline, and generally much more prone to error. An algorithm just deals with the situation, assesses options faster and more reliably.
In emergency situations, humans panic, secretes more adrenaline, and generally much more prone to error. An algorithm just deals with the situation, assesses options faster and more reliably.
Your algorithm would've crashed Sully`s flight trying to make it to Teeterboro...
Could auto pilot safely land in the Hudson River if need be?
I didn't think so.
I didn't think so.
4
Maybe not.
But the many more averted crashed would more than offset the one loss.
But the many more averted crashed would more than offset the one loss.
1
Probably not today. But there's no reason it couldn't be made to do so, and more reliably.
Ya right. The next thing you know we'll have a plane taken out by the same virus we can't control on our desktops.
1
The statistical occurrence of the Germanwings incident is 1 in 50 million.
4
That's probably a pessimistic assessment, considering there's 30 million commercial flights annually.
I am sure that statistic is very comforting to the 150 families involved.
1
You say:
In this scenario, a ground controller might operate as a dispatcher managing a dozen or more flights simultaneously. It would be possible for the ground controller to “beam” into individual planes when needed and to land a plane remotely in the event that the pilot became incapacitated — or worse.
Can you imagine if the Germanwings pilot was able to control 12 aircraft simultaneously and 'beam in' to take control?
In this scenario, a ground controller might operate as a dispatcher managing a dozen or more flights simultaneously. It would be possible for the ground controller to “beam” into individual planes when needed and to land a plane remotely in the event that the pilot became incapacitated — or worse.
Can you imagine if the Germanwings pilot was able to control 12 aircraft simultaneously and 'beam in' to take control?
4
If there had been a robot at the controls of US Airways 1549 on January 15, 2009, instead of Capt. Chesley "Sully" Sullenberger, everyone on board would have died. There are some, perhaps many, emergencies a robot and it's programming can't handle.
1
1) You have no way to know whether a robot could have saved US Airways 1549.
2) Even if a robot could not have prevented that crash, there are far more crashes attributable to human-error than the one "miracle on the Hudson."
2) Even if a robot could not have prevented that crash, there are far more crashes attributable to human-error than the one "miracle on the Hudson."
1
"But will passengers board a flight controlled by a robot?"
No.
No.
5
I would, so your statement is false.
Brings a whole new meaning to the phrase, "computer crash". No thanks.
6
The first time someone the system gets hacked or a plane goes down due to software problems, people will think twice. The current data uplink system used on most commercial flights (ADS-B) is easily received by anyone with a USB thumb drive / antenna. It doesn't seem prudent to entrust a $100-200 million dollar piece of capital to robots just to save on two salaries.
3
The first time a human deliberately crashes a plane, people will think twice. Oh wait, this is like the third or fourth time it has happened, so maybe we're ready for pilotless planes.
I am a pilot and this idea is just silly. Riding an unmanned train on tracks is one thing. Riding aboard an airplane, aloft in the three dimensions, relying on millions of parts, pieces, obstacles and software requires a "HIGHER LEVEL" of problem solving than any auto-pilot can provide. I do think a plane could be made to automatically resist a foolish move, like diving for the earth, but eliminating the pilot would only work for the birds, not any plane.
12
The first four words of your comment disqualify you from making and objective statement on this issue.
2
I don't want to sound rude, but since your not an autopilot engineer, you really aren't qualified to make such a comparison.
I know some folks who work at BART in the SF bay area, and the main job of a BART driver is not to step on the accelerator or the brake, computers can do that. It's to look after the safety of the passengers when all else fails … that is to say, when the computers malfunction!
I want to see two smart, well paid pilots in the front of every airplane for a very long time to come, thank you.
I want to see two smart, well paid pilots in the front of every airplane for a very long time to come, thank you.
5
This being the Times, i figured that there would be lots of posts that grotesquely underestimate our current engineering capability; this is an age of microprocessors, not relays and gears. Middle-aged demographic, I guess, though at 60 not something I identify with.
People are so scared of change! These same arguments were once made against automatic elevators. Like self-driving cars, automated flight is coming because we now have the technology that makes it feasible, and will be both safer and more economical than reliance on human beings.
People are so scared of change! These same arguments were once made against automatic elevators. Like self-driving cars, automated flight is coming because we now have the technology that makes it feasible, and will be both safer and more economical than reliance on human beings.
3
Sorry to disappoint you. I'm in my seventies, pilot licenses, engineering degrees, a fair knowledge of processors, computers and life experiences with humans. That is why I would not get into a pilot-less aircraft.
1
My grandfather would not get on any plane with less than 4 engines. Automation is coming, almost all here, and it will really take off when flights billed as automated are cheaper than piloted ones.
2
HAL 9000 is really on the horizon? If given a choice between a person and a computer I'll take my chances with the human.
1
So we can open it up all the people who will try to hack into this remote system and fly the plane to God knows where? Ridiculous!!! I like things just the way they are. Maybe we need a woman in that cockpit.
We can't predict dangerousness. Some nut can always find a way to program a plane to crash, no matter how technologically advanced we are, and even if robots do everything. I think we need to understand more about how killers think, and why they think the way they do, and how to disarm them.
http://www.thecriticalmom.blogspot.com
http://www.thecriticalmom.blogspot.com
1
The argument about pilot-less airplanes can be dealt with my mentioning one name: Capt Chesley (Sully) Sullenberger.
Cant Sullenberger piloted US Airways flight 1549 to a safe landing in the Hudson after losing both engines to bird strikes while climbing out of La Guardia. No amount of computing can make that decision. Against all odds, he got the plane down in one piece with nothing but minor injuries to passengers and crew.
I'm a pilot (non-commercial) and I have nothing against cockpit automation. But I want a real pilot up front, preferably a sane one, with grey hair and golden hands!
Cant Sullenberger piloted US Airways flight 1549 to a safe landing in the Hudson after losing both engines to bird strikes while climbing out of La Guardia. No amount of computing can make that decision. Against all odds, he got the plane down in one piece with nothing but minor injuries to passengers and crew.
I'm a pilot (non-commercial) and I have nothing against cockpit automation. But I want a real pilot up front, preferably a sane one, with grey hair and golden hands!
5
For every passenger saved by a pilot (a la Sully) there are scores who have perished in crashes caused by human error. Removing the pilot will not eliminate all aviation fatalities ... Nothing will. But removing the human pilot will make flying even safer than it already is.
1
I would ask the passengers aboard Captain Sullenberger's 2009 US Airways flight out of LaGuardia what they think of this idea. They are alive thanks to the skill and experience of a human pilot.
5
I would ask the families of all those who died in human-error crashes (which far outnumber the few on Sully's flight) what they think of the idea.
You are suggesting that we shouldn't act to save thousands because we lose hundreds.
You are suggesting that we shouldn't act to save thousands because we lose hundreds.
1
Wouldn't it be possible to program the plane so it couldn't fly into a mountain or do something else dangerous? The system could not be disabled by a single pilot bent on crashing the plane. Just as cars are getting safety features to help prevent crashes, such as braking or stopping automatically if it gets too close to the car in front, planes could be outfitted with similar pilot override features.
It may be quite a while before the best pilots are worse than the typical silicon pilot. But... you don't usually have the best pilot on your flight. You might have the worst pilot in the fleet. There is also the possibility that the worst pilot in the fleet is the perfect candidate for the one in a million upset that the software (and the best pilot) will fail to handle. You just don't know.
Humans being what they are, it will be a one or two generations before fully autonomous commercial passenger flight is common. The current crop of humans with the "right stuff" and their familiarity with (real or imagined) failure prone automated systems will need to age out of existence...
Humans being what they are, it will be a one or two generations before fully autonomous commercial passenger flight is common. The current crop of humans with the "right stuff" and their familiarity with (real or imagined) failure prone automated systems will need to age out of existence...
3
Oh yes, based on a single action by a mentally deranged pilot, let's go automaton! What do you mean hackers? We could have a human back-up on board, just in case the flight control by Hal, gets hacked. Then things would be much safer. This column isn't really about science, although it's labeled science, it's about technology and hysteria.
6
Of the 7,244 fatal airplane crashes in the United States from 1993 through 2012, 24 were the result of aircraft-assisted suicide, the authors concluded in the 2014 study published in the journal "Aviation, Space and Environmental Medicine."
The idea of 'robotic' piloted airliners is naive because it fails to take into account the principles that dictate modern aviation safety, namely redundancy, fault-tolerance and fault passivity. Certainly current aircraft design can fly themselves when everything is functioning correctly and they rely on multiple layers of redundancy wherever possible. But the array of sensors that are essentially the 'eyes and ears' of the automatic flight control are prone to failure. Since they are exposed to a common environment, they can all be expected to fail simultaneously, as we have seen with Air France flight 447 (although pilot error was the cause of the crash). The only redundancy we can provide for the not-so-uncommon scenario of sensor failure is a human pilot. This is what we refer to as a fault-passive defense. The airplane system fail but the airplane can still be flown safely by the pilot. Without the pilot there is no defense. The other limiting factor is autoland, which depends on the airport's ILS radio beacon for navigation. That beacon can be distorted by objects on the ground, most notably other aircraft near the runway. As a result, the plane could land well off the runway centerline and leave the runway surface resulting in a gear collapse or even a ground collision. In short, while the technology exists to fly without a human pilot, the human pilot must be there to satisfy the doctrine of redundancy, fault-tolerance and fault passivity.
2
Wow, so your reaction to Germanwings is to replace one man piloting one plane with one man piloting multiple planes?? And to electronically automate flight??
A single human being can go bad, or crazy. But he cannot be hacked. And you cannot hack multiple human beings at once. And a single human pilot can only bring down a single plane, rather than a bunch of planes he's got connected to his remote control.
In terms of terror prevention, if nothing else, this seems like a colossally bad idea.
A single human being can go bad, or crazy. But he cannot be hacked. And you cannot hack multiple human beings at once. And a single human pilot can only bring down a single plane, rather than a bunch of planes he's got connected to his remote control.
In terms of terror prevention, if nothing else, this seems like a colossally bad idea.
2
So what exactly are people for anyway? It seems we are making ourselves more irrelevant and pointless every day.
1
The notion of commercial planes flying without pilots sounds like an oxymoron. Passengers such as myself are hoping for potential solutions to avoid a repeat of another Germanwings tragedy. However, the idea of a suicidal pilot is beyond rare when comparing that to the number of flights which occur every day. I keep thinking about the astounding actions of pilots Capt. Sully Sullenberger (Hudson River crash); Capt. Alfred C. Haynes (Sioux City, Iowa cartwheel plane crash); and Capt. Eric Moody (after four engines failed, he nosedived the plane which caused the engines to re-start). All three pilots kept cool, calm and focused under the worse known pressure and saved many lives. I will always believe that I am in safer and more capable hands of a live pilot if an emergency occurs compared to that of an automated system that is unable to “think” out of the box.
1
Yours are common arguments, the problem is taphat these exceptional Captains are not onboard every flight, they were not on board AF447 to name one. Present day aviation safety levels result from a high level of automation both on board and in air traffic control centers.
1
I agree with your assessment hat millions of flights occur every day without incident largely due to safety levels of automation both on the plane and air traffic controllers. However, my point was that when something did go wrong in the incidents I mentioned, it was the actions of the three pilots that saved so many lives. No one knows if a fully automated plane could have had the same positive results. I just never want to be in a situation where I have to find out first hand.
1
We need more experienced, well-trained people on those aircraft, especially as they are designed to hold more and more people (Dreamliner, etc.) -- rather than dispense with the pilots, let's go back to the days when a third cockpit crew member was on all flights.
The Germanwings passengers would still be alive, among others.
The Germanwings passengers would still be alive, among others.
The main thought that comes to my mind when there is turbulence etc is that the pilots and the entire crew really want to go home safely. If a robot was running things, I would not have that reassurance.
2
Auto pilot is already used today to land planes when weather's bad. They do it more comfortably and safer than humans.
Complex software is inherently unreliable. It can be made close to foolproof, but the cost follows the 95/5 rule: the last 5% costs as much, or more, than the first 95%. Problems are nearly always the result of something no one thought of during testing, which is merely annoying if a web site is on its knees, but rather more critical if people's lives are on the line. What would be an acceptable number of deaths to iron out those last couple of bugs?
I think I'd rather drive; oh wait - now they're talking about fully automated cars!
I think I'd rather drive; oh wait - now they're talking about fully automated cars!
25
Better a fully automated car than a car driven by an habitual, or even occasional, texter.
If the person and/or thing controlling the plane I'm on isn't taking the same risk I am, then I won't be flying any more.
8
I find this creepy. Just how far are airlines willing to go to earn more profit? I would not want to fly on a plane that is totally automated. I do believe that human judgement at times is needed and a robot isn't smart enough to work in emergency situations. If a robot is that smart, I definitely don't want to be on that plane.
5
Actually, yes, computers are now more than smart enough to work in emergency situations. You also don't understand basic economics. Of course, the airlines want to earn more profit, but automation will reduce the price of a plane ticket and so benefit us all.
1
10 years or so ago, I was on a flight, I think landing in Philly, and I think US Air. After the landing, the pilot came on the PA and said, "If you thought you might someday be on a plane that landed itself, well today you were. Today's landing was done completely automatically." All we passengers looked at each other for a second with kind of a nervous chuckle and moved on. I don't know what they were testing or using that day, or what the pilot meant.
2
10 years ago I was told that 90% of landings were already automated.
$340 billion annually is such a big number that will generate companies that will say whatever it takes to get more automation going to just get research skewed favorably towards the automation.
But as Abhay notes, the $340B number is fallacious. According to Abhay's math (and I believe him) it is more like $8B per year, or a max of $250,000 per pilot per year.
No thanks. A thousand times over.
There are too many contingencies, too many variables.
The Germanwings crash was a one in 100 million event.
This whole debate about replacing humans in driving cars and flying airplanes is so absurd that it doesn't really merit serious consideration in an otherwise serious newspaper. There might come a time, in 50 to 100 years, when people have been so conditioned to turn their lives over to machines that they would accept either example. None of us will likely be around to witness that transition and we would not want to see it, even if we were.
There are too many contingencies, too many variables.
The Germanwings crash was a one in 100 million event.
This whole debate about replacing humans in driving cars and flying airplanes is so absurd that it doesn't really merit serious consideration in an otherwise serious newspaper. There might come a time, in 50 to 100 years, when people have been so conditioned to turn their lives over to machines that they would accept either example. None of us will likely be around to witness that transition and we would not want to see it, even if we were.
4
Wow, you sound like one of those people who said man would never fly! Have you never ridden an automatic elevator, or a computerized train at an airport? The technology to do these things is here now, not even speculative anymore, and posts like this will be laughed at 100 years from now.
1
Josh, we're already laughing.
Thank you Patrick Smith for your comments. Everyone, please!, listen to him (and read his book).
If it can be controlled from the ground, it can be hacked from the ground.
Also, there can be no robot Sully Sullenberger! Let's ask him what he thinks of this development.
Also, there can be no robot Sully Sullenberger! Let's ask him what he thinks of this development.
3
I want our military to use humans to kill other humans, that way at least someone feels guilty or can be held responsible for when things go wrong. Little enough accountability as there is already.
4
Funny, I was thinking the very opposite: If we have to fight nutters from ISIS, I'd rather that the lives of our soldiers not be risked.
Reminds me of the old joke about the first robotic plane. The one where the electronic voice says "nothing can go wrong; go wrong; go wrong;..."
4
We humans have good and bad days, our mood is ever changing, some days we have problems at home and our level of attention reduces, other times we are tired after 12 hours in the air and our performance is not the same than when we are rested, etc.
The safety of a system such as aviation cannot depend on these changing moods. This has been understood by manufacturers and modern aircraft are very safe as a result of the automatic systems that have been included in their designs. Flight By Wire technologies introduced by Airbus in commercial aircraft more than 30 years ago are the biggest breakthrough in terms of aviation safety, these systems protect Airbus aircraft against exceedances of their flight envelopes, the Captain of US Airways flight 1549 ditched in the Hudson River as he did because he was an exceptional pilot and was exceptionally helped by the automatic systems of his Airbus.
With these technologies drones can carry out their missions and as has been recently demonstrated successfully operate from aircraft carriers. For a drone there is no difference in carrying ordnance or a commercial load, and there is no difference in operating in a dedicated runway or in a commercial airport, the only difference is legislation. In my view there will be a growing demand from air cargo operators to operate fully automated cargo airplanes that will share the sky with other commercial flights. I have been a pilot many years and am convinced that this will be the future.
The safety of a system such as aviation cannot depend on these changing moods. This has been understood by manufacturers and modern aircraft are very safe as a result of the automatic systems that have been included in their designs. Flight By Wire technologies introduced by Airbus in commercial aircraft more than 30 years ago are the biggest breakthrough in terms of aviation safety, these systems protect Airbus aircraft against exceedances of their flight envelopes, the Captain of US Airways flight 1549 ditched in the Hudson River as he did because he was an exceptional pilot and was exceptionally helped by the automatic systems of his Airbus.
With these technologies drones can carry out their missions and as has been recently demonstrated successfully operate from aircraft carriers. For a drone there is no difference in carrying ordnance or a commercial load, and there is no difference in operating in a dedicated runway or in a commercial airport, the only difference is legislation. In my view there will be a growing demand from air cargo operators to operate fully automated cargo airplanes that will share the sky with other commercial flights. I have been a pilot many years and am convinced that this will be the future.
1
At last, someone familiar with contemporary technology! To read most of these posts, you'd think that we were still using vacuum tubes.
Think about how often your iPhone apps crash. Now think about how often planes crash. Still think we should move to pilotless planes?
5
I don't think we're talking about software installed for $1.99 per plane from the Apple Store.
1
Yes, because a) I have an Android phone and in its 1-1/2 years it's never once crashed and b) more seriously, computers used in critical applications are designed to far higher levels of mechanical and programming reliability than the toys we carry in our pockets. Since new planes have been using fly by wire for some time, there would be a lot of planes falling out of the sky if that weren't the case!
1
The plane that landed in the Hudson River was in the control of a very experienced pilot. A computer would have done what with both engines out?
16
The computer would probably have done what the pilot did. Do you think we engineers are so incredibly dumb that we'd ignore the need to ditch a plane in an emergency? Of course, the programming won't be perfect, particularly initially, but neither are human pilots; a reasonable goal would be a system that is safer overall than a human pilot.
1
Nope, because the first step in the sequence was to decide the Hudson River was an option. A computer might have landed the plane successfully, after somebody had told it to land it there.
In total nonsense, this idea ranks right along with the self-driving car. The fact is, 150 people died on the Germanwings flight because the captain of the aircraft was locked out of his own cockpit.
The same outcome might likely have occurred if two people had been in the cockpit and one overpowered the other. Or, if the co-pilot was of sound mind and simply suffered a disabling siezure or other medical event while alone.
The obvious answer is never under ANY circumstances allow a situation where a pilot could be locked out of his own cockpit. That idea was another knee-jerk response to 9-11, and the consequences have proven disastrous.
The same outcome might likely have occurred if two people had been in the cockpit and one overpowered the other. Or, if the co-pilot was of sound mind and simply suffered a disabling siezure or other medical event while alone.
The obvious answer is never under ANY circumstances allow a situation where a pilot could be locked out of his own cockpit. That idea was another knee-jerk response to 9-11, and the consequences have proven disastrous.
33
LOL, and you say the self-driving car is nonsense because . . . ? You do know that Google has one tooling around now, and that the major manufacturers plan to introduce them commercially in the next few years?
Look at the history of technology. The people who say "This will never happen" end up laughed at 50 years later. Man will never fly . . . man will never go to the moon . . . why do people think that way?
Look at the history of technology. The people who say "This will never happen" end up laughed at 50 years later. Man will never fly . . . man will never go to the moon . . . why do people think that way?
1
Pilots aren't there to fly the plane these days. They're there to react when something unusual happens. Think the captain of that USAir plane that hit a flock of birds on departure from LaGuardia who quickly and effectively recognized that the best course of action would be to land on the Hudson, then did so without hitting any boats and saved every life on board. Sometimes stuff happens in complicated systems that requires human judgment to resolve - I just departed and am flying over a teeming metropolis and just blew through a bunch of birds and my engines quit. Now what? I am flying over Japan and my engines just ingested volcanic ash and all quit at once. Now what?
6
Why is this in the Science section? Shouldn't it be in the Something That Will Never Happen section?
This is another example of Technology Optimism Psychosis. Something happens, a rare event, and the fan boys immediately start with the "technology will solve everything!" mantra. Give it a rest. Do something useful. Not this.
This is another example of Technology Optimism Psychosis. Something happens, a rare event, and the fan boys immediately start with the "technology will solve everything!" mantra. Give it a rest. Do something useful. Not this.
11
Yeah, right, nothing ever happens in technology. Not fire, not the wheel, not the printing press. Not the steamship or the telegraph or photography or the railroad. Not the phonograph or incandescent light or motion picture or telephone. Not radio or television or the digital computer or the smartphone. Not the airplane and the motorcar and spaceflight.
The problem with people like you is that you don't understand that your personal inability to do something doesn't translate into an actual inability to do it. We already know how to make pilotless aircraft. Since such aircraft will be more reliable and cheaper to run, It isn't a question of whether, but when. And you'll have us fan boys to thank for that. Certainly not the Luddites in the comments section, who never met an advance they didn't fear!
The problem with people like you is that you don't understand that your personal inability to do something doesn't translate into an actual inability to do it. We already know how to make pilotless aircraft. Since such aircraft will be more reliable and cheaper to run, It isn't a question of whether, but when. And you'll have us fan boys to thank for that. Certainly not the Luddites in the comments section, who never met an advance they didn't fear!
For those who think technology rules, I've got 2 words for you: Sully Sullenberger.
10
"In this scenario, a ground controller might operate as a dispatcher managing a dozen or more flights simultaneously. It would be possible for the ground controller to “beam” into individual planes when needed and to land a plane remotely in the event that the pilot became incapacitated — or worse."
And it would presumably be possible for a ground controller to "manage" a dozen or more flights into oblivion. The question will turn from evaluating pilots to evaluating ground controllers for mental and possibly other illnesses.
And it would presumably be possible for a ground controller to "manage" a dozen or more flights into oblivion. The question will turn from evaluating pilots to evaluating ground controllers for mental and possibly other illnesses.
4
People get on automated trains all the time. Our own Airtrain operates without any conductor. The only fatal accident it ever had was when it was driven by a human operator around a curve too fast, causing it to derail. Most fatal train accidents are caused by human error - usually lack of attention.
Asiana Air crashed while landing at San Francisco Airport because the pilots couldn't manually fly it, when the Instrument Landing System was inoperative; they were so dependent upon automation their skills deteriorated.
But extraordinary humans can solve problems that no automated system can handle: Sully, Jim Lovell, etc.
Asiana Air crashed while landing at San Francisco Airport because the pilots couldn't manually fly it, when the Instrument Landing System was inoperative; they were so dependent upon automation their skills deteriorated.
But extraordinary humans can solve problems that no automated system can handle: Sully, Jim Lovell, etc.
3
All we have to do is get to the point where computer reliability outweighs the extra intelligence that humans can bring to the situation. Not exactly a difficult hurdle!
It's sounds great until you're on board and hackers disable the navigation system.
4
My stepson, who is a commercial airline pilot and flies the Airbus A320, told me a joke that seems to fit. In the future all planes will have a pilot and a dog in the cockpit. The pilot is there to reassure the passengers. The dog is there to bite the pilots hand if he tries to touch the controls. Case closed!
37
Actually, the pilot is there to feed the dog.
I heard about something like this proposal years ago. A plane so automated, you only needed one guy and a dog in the cockpit. The guy was there to feed the dog. The dog was there to bite the guy if he touched the controls.
The only thing dumber than the idea of replacing all pilots with robots in order to avoid that one chance in how many millions of having a lone suicidal pilot is the idea of taking a poll of the general public to decide the issue.
The only thing dumber than the idea of replacing all pilots with robots in order to avoid that one chance in how many millions of having a lone suicidal pilot is the idea of taking a poll of the general public to decide the issue.
2
Say what??? Current situation: risk (rather low) that the pilots will crash the plane, either deliberately or by accident. Proposed scenario: thousands of potential hackers on the ground drive the plane into the ground. Are these people serious? By the way, same issue for driverless cars. It will be so easy to commit the perfect crime by paying a hacker to crash your enemies' cars at 90 mph.
2
Not really. Cars won't be designed to be as vulnerable as your computer because lives will depend on their not being. Personal computers are easily hackable because they're designed so that everyone and his sister can run code on them.
These are very bad ideas. We know that hacking into computer systems gets better by the day , so robots or ground based controllers both reduce rather than enhance safety. We should not go crazy over one incident or even three possible crashes over 20 years caused by suicidal pilots. Requiring a second person in the cockpit and enhanced psychological screening by the medical examiners is more than sufficient to enhance safety from these exceeding rare events.
8
Anyone who has used a computer or relied on a GPS to guide their car knows how often you have to intervene to reboot or to ignore wrongheaded directions. Trust an airliner to technology with no human on board to override when needed? That will take some persuading. I also pilot my own small plane, and have more than once been unpleasantly surprised when the autopilot suddenly ran away with the plane by commanding full nose-up pitch, which would have resulted in a stall and crash if I hadn't disconnected it. The fault is some eccentric gremlin in the circuits that the avionics repair shop couldn't find (ask anyone who has taken a car to the shop and been unable to replicate the problem that brought it there).
Most importantly, though, I've always liked the grim old adage that airliner pilots are just as interested in safe flight as you are, because in the event of an accident, they get there first. I like having someone in the cockpit with skin in the game (assuming they're mentally healthy, of course).
Most importantly, though, I've always liked the grim old adage that airliner pilots are just as interested in safe flight as you are, because in the event of an accident, they get there first. I like having someone in the cockpit with skin in the game (assuming they're mentally healthy, of course).
9
Your GPS is built and programmed to consumer specs. The fly-by-wire computers that *already* run commercial airliners are built and programmed to much higher ones. And we engineers do mathematical risk assessment; in addition, in the case of commercial aviation, accidents are carefully analyzed to improve safety. Risk is not handled at the childish level of having skin in the game.
so wrong
Rather than turn the entire civilian aircraft industry into a gigantic mental health hospital, where one flew over the cuckoo's nest really does mean one flew over the cuckoo's nest; why not just ban anyone with any kind of mental illness from operating heavy equipment, like, you know, passenger aircraft?
Or has political correctness become the process where the Nanny State morphs into the Big Nurse Ratched State?
Or has political correctness become the process where the Nanny State morphs into the Big Nurse Ratched State?
2
If you impose such a ban, do you think any pilot will choose to get treatment for his/her mental problems?
Mental issues like depression are both common and treatable. I'd much rather fly with a depressed pilot who has weekly sessions with a good therapist than a with depressed pilot who has never seen a therapist.
Mental issues like depression are both common and treatable. I'd much rather fly with a depressed pilot who has weekly sessions with a good therapist than a with depressed pilot who has never seen a therapist.
"In 2014, airlines carried 838.4 million passengers on more than 8.5 million flights."
Flying is one of the safest modes of transportation and increasingly people are surviving crashes, because of trained crew members. Not just "Sully" Sullenberger's miracle landing, but many other crashes. Anyone who has worked with technology, knows it can fail. We are still looking for that Malaysian plane that fell out of the sky, yet we can track whales all over the planet. If airlines are resistant to live GPS technology, because of the cost, imagine trying to implement this. There will be lessons learned from this, as we've learned from other crashes, and thoughtful review of procedure will need to be implemented, but you won't find me flying without a pilot.
Flying is one of the safest modes of transportation and increasingly people are surviving crashes, because of trained crew members. Not just "Sully" Sullenberger's miracle landing, but many other crashes. Anyone who has worked with technology, knows it can fail. We are still looking for that Malaysian plane that fell out of the sky, yet we can track whales all over the planet. If airlines are resistant to live GPS technology, because of the cost, imagine trying to implement this. There will be lessons learned from this, as we've learned from other crashes, and thoughtful review of procedure will need to be implemented, but you won't find me flying without a pilot.
6
Combine pilotless aircraft, noise-cancelling headphones, and Uber -- travel coast-to-coast without speaking to another human being.
You'll know I'm dead when my Apple watch texts you.
You'll know I'm dead when my Apple watch texts you.
4
I want the guy flying the plane to have the same stake in its safe return that I have.
6
Perhaps we might do some historical research and learn from the thoughts of Charles Lindbergh and Jimmy Doolittle on such subject matter.
This entire notion is comical.
This entire notion is comical.
2
I'm a former military pilot. What's missing from this discussion is that the German Wings crash could have been prevented if the Captain were able to regain access to the cockpit. Following 9-11, we instituted a cockpit door system without any form of backup whatsoever to 'keep the terrorists out'. A more careful analysis of all the risks might have produced a system with a viable backup for the event that the Captain is locked out of the cockpit. A Captain needs to have access to all parts of his craft at all times.
I see pilotless airplanes in the same way. It is a system without any backups that is being promoted as a knee-jerk reaction in the aftermath of the German Wings mishap. A computer might not make the same mistakes that a human pilot would make, but it will make other ones. Military pilotless airplanes can and do crash. Computer systems can and do fail. But since no life is lost, these crashes rarely make the news and it creates a false perception that pilotless airplanes are in fact safer. Safety needs to take into account the big picture and consider all the potential trade offs. Knee jerk reactions and quick fixes do not make us safer and in most cases increase risk.
I see pilotless airplanes in the same way. It is a system without any backups that is being promoted as a knee-jerk reaction in the aftermath of the German Wings mishap. A computer might not make the same mistakes that a human pilot would make, but it will make other ones. Military pilotless airplanes can and do crash. Computer systems can and do fail. But since no life is lost, these crashes rarely make the news and it creates a false perception that pilotless airplanes are in fact safer. Safety needs to take into account the big picture and consider all the potential trade offs. Knee jerk reactions and quick fixes do not make us safer and in most cases increase risk.
13
Lets start with remotely capturing info from flights instead of the ridiculous black boxes. Once we solve that hurdle maybe then we can ponder pilotless flights.
2
I think this is an absolutely stupid idea. If it happens and I think it won't I will take a boat (to Europe) or drive when I can. I don't think it will ever be accepted or acceptable. It fits well with those that think technology is the answer to everything, including things like overpopulation and the loss of natural resource function (eg loss of forests.).
4
Can you tell us why you have declared it "stupid" without explanation? Or is it self-evident?
Much of flying is already automated, and 2/3 of the air force is now drones.
Complete automation will come to planes as it is to cars within the next 10-15 years. At first airlines will keep a pilot on hand just in case, after several years people will accept they are safer without the possibility of pilot error.
Complete automation will come to planes as it is to cars within the next 10-15 years. At first airlines will keep a pilot on hand just in case, after several years people will accept they are safer without the possibility of pilot error.
14
I don't care what the hype is, driverless cars and pilotless planes will never happen.
1
Dear Ed Anger,
The problem is, every one of those drones is being piloted remotely, by two pilots. There is no such thing as a completely automatic plane yet, manned or unmanned, and it's unlikely that one will be able to deal with unexpected situations, not one that we'd be able to create anytime soon.
The problem is, every one of those drones is being piloted remotely, by two pilots. There is no such thing as a completely automatic plane yet, manned or unmanned, and it's unlikely that one will be able to deal with unexpected situations, not one that we'd be able to create anytime soon.
I wouldn`t model my portfolio around that if I were you...
Autonomous commercial airliners were feasible 40 years ago, so this is really nothing new. Growing up in Connecticut, we had a family friend who was a senior pilot for Eastern Airlines. He would mimic a mechanical voice addressing passengers on a hypothetical takeoff, saying:
"Ladies and gentlemen, this is your automatic pilot speaking and welcome to flight X*@# to Houston. Your flight will be controlled at all times by computer, without any human intervention. You can be assured that this flight is perfectly safe and all precautions have been taken to ensure your comfort and safety. So sit back, relax, enjoy the flight and know that nothing can go wrong...go wrong...go wrong...go wrong... ..."
"Ladies and gentlemen, this is your automatic pilot speaking and welcome to flight X*@# to Houston. Your flight will be controlled at all times by computer, without any human intervention. You can be assured that this flight is perfectly safe and all precautions have been taken to ensure your comfort and safety. So sit back, relax, enjoy the flight and know that nothing can go wrong...go wrong...go wrong...go wrong... ..."
1
You can imagine I'm less than thrilled over the prospect of some teenage hacker in his Mom's basement taking control of my flight.
4
The $340 billion figure quoted as costs related to the co-pilot position in the worlds fleet of passenger aircraft is absurd. Another sign of innumeracy in the ranks of journalists and people in general. That number is greater than the sum of annual revenues of all global airlines by at least a factor of 2. The costs associated with the co-pilot position, assuning 8.4 million commercial flights a year and each plane flying 360 days a year, amounts to less than $8 billion, even assuming a cost of $250K per co-pilot position per year.
4
More like off by an order of magnitude. The US passenger carriers have 75K pilots, about half co-pilots and and over a quarter of the world's passenger aircraft. Worldwide there are perhaps 150K co-pilots flying with passenger airlines, which at $200k total labor cost (which might be right for the US, but is much too high for most of the world), comes out to around $3 billion/yr.
Oops - public math. Should have said $30 billion/yr.
Back in the '60s, my uncle told me a joke: The passengers on an airplane had all taken their seats, and the stewardesses closed and locked the door. A voice came on the loudspeaker that said, "Good afternoon ladies and gentlemen. This is an airline first. This airplane is being piloted by a robot, with no human crew in the cockpit. Do not worry, nothing can go wrong, go wrong, go wrong, go wrong..."
3
Yes I would fly in a automated plane. Most crashes are caused by human error or intent. Yes a robotic plane could have landed in the river.
And back in the 80's the consensus among scientists was that artificial intelligence was only a few years away from reality.
Do you really want Siri flying your plane?
Do you really want Siri flying your plane?
Automation could land a plane on water, maybe, but I doubt it could have sorted out the options to choose to land on water, instead trying to make it to a runway. I'll take a human at the controls when the real art of flying is needed.
I welcome our robot overlords...All hail Apox 132! A subsidiary of the Sandrupian Corporation...
1
Right. It should have 'speech synthesis' so that when the air traffic controller frantically tries to override a glitch, the robot can say "Sorry, I can't do that, Dave." And "saving labor costs" really means putting any number of airplane pilots out of work. No, thanks. Rather than fly with the robo-pilot, I'll stand next to Capt. Sullivan in the unemployment line.
6
One of the older jokes in the world: "Welcome aboard our automated flight. You'll be happy to know that nothing can go wrong, go wrong, go wrong, go wrong...." Automated systems cannot handle the wholly unpredictable; good pilots can. Hudson River landing, anyone? Not by any autopilot ever made.
18
What is the probability that a computer system driving that plane will ( WILL) be hacked? What are the probabilities that the reactions to this one time event are driving an over reaction based on fear? 538? As a commercial pilot with 30 years in both military and civil aircraft, I have flown in or around thundestorms, typhoons, post 9-11 lunacy, earthquakes ( Japan- the big one), snow, ice and passengers who have swung at me and mountains. While the media pumps your fear emotion , what do entities such as Citi, Home Depot, Target , NSA and more tell us about computers?
7
Outside the Airbus cockpit is a keypad. If there were a code that said "emergency; engage auto pilot and land the plane automatically at the nearest airport", the Germanwings issue would never have happened, and a hijacker would always need enough people to control both the cockpit and the keypad in front of it.
1
Yes, of course the response to extremely rare pilot suicide should be to put a passenger aircraft in the hands of a potentially depressed (or even very inattentive) ground based pilot, who can crash the plane at no risk to himself, and operates it through a potentially hackable control system.
117
... and worry about the ground based pilot texting while on the job working!
April 6, 2015
Reminds me of what I often think - we know what AI does - but why and how is in fact its human equation.
Wherever we take our mind and technology - as in the old days with the underground nuke facilties of the cold war era -it too was about human equations to get and keep things right and safe.
As very high tech monitors and consolidation of interface warrant expertise of those at the console to interactive effectively and indeed with simulation auditing testing allowing professional feedback and as well to reward operations with professional satisfaction understanding in the human drama that's at high risks and even electronics that work with backup programs for all possible events.
jja Manhattan, N. Y.
Reminds me of what I often think - we know what AI does - but why and how is in fact its human equation.
Wherever we take our mind and technology - as in the old days with the underground nuke facilties of the cold war era -it too was about human equations to get and keep things right and safe.
As very high tech monitors and consolidation of interface warrant expertise of those at the console to interactive effectively and indeed with simulation auditing testing allowing professional feedback and as well to reward operations with professional satisfaction understanding in the human drama that's at high risks and even electronics that work with backup programs for all possible events.
jja Manhattan, N. Y.
1
suppose someone wants to redirect the autopiloted plane from a distance????
1
Never will happen, humans like a sense of being in control, whether it really is true or not. Over 95% of the flights are presently done by computers, so please, leave a little bit for us humans.
1
So what about UPS and FedEx using such automation in their cockpits? Those aircraft have very few souls on board and the cost incentives to cargo-handling companies are compelling. In addition, automation in such an environment would help "prove out" this technology without the same risks to passengers (although the risks on the ground from automation mistakes remain unchanged regardless of the number of folks in the air).
The First Rule of Technology states: Hardware eventually breaks, Software eventually works.
I'll pass on robot-flown commercial aircraft.
I'll pass on robot-flown commercial aircraft.
2
I want my flight controlled by somebody up there at 38,000 feet with me, who has the same stake in a safe landing as I do -- not some techie on the ground playing a sophisticated video game.
6
I wouldn't get into a plane flown by a robot, simply because a robot cannot improvise. Remember the plane that Mr. Sullenberger landed on the Hudson, with no casualties at all? A robot wouldn't have sufficient programming to pull off that rather miraculous safe landing, and would most likely instead have plowed into Tribeca or Hoboken after the goose took out the engine, due to not being able to handle the unexpected.
We don't have artificial intelligence yet, that's the problem here, and the blessing. When we invent artificial intelligence, in all likelihood it will try to exterminate us out of self-preservation, which would be the smart move for it. For now, computers cannot think, and cannot come up with solutions when they haven't been programmed to react in a certain way. And programmers aren't omniscient, so there's no way for them to conceive of all possible scenarios.
So for the pilot-less plane, the answer is, don't be silly. If you want to fly more safely, use solar-powered dirigibles, we could make vastly superior ones nowadays, and their casualty rate from crashes would be incredibly lower than jets'. If you want to fly completely safely, with no possibility of accidents, the only way to do that is: don't fly.
We don't have artificial intelligence yet, that's the problem here, and the blessing. When we invent artificial intelligence, in all likelihood it will try to exterminate us out of self-preservation, which would be the smart move for it. For now, computers cannot think, and cannot come up with solutions when they haven't been programmed to react in a certain way. And programmers aren't omniscient, so there's no way for them to conceive of all possible scenarios.
So for the pilot-less plane, the answer is, don't be silly. If you want to fly more safely, use solar-powered dirigibles, we could make vastly superior ones nowadays, and their casualty rate from crashes would be incredibly lower than jets'. If you want to fly completely safely, with no possibility of accidents, the only way to do that is: don't fly.
8
The train between the airport terminals is more likely than not to be free of a conductor. I know a plane is much more complex, but it historically has not taken too long for humans to trust the machines.
1
Not a very good comparison, Ben. The train goes the same route every trip. It can't vary the route for any reason. The risk of a crash is minimal and the likelihood of fatal injuries in the event of a crash is also minimal. So even if the train controls were hacked, the risk to passengers is very, very small.
It seems like one of the first steps could be a switch that would allow to ground controllers to take over an airliner in case the pilots are incapacitated, there's a hostage takeover, or another pilot goes on a suicide mission.
5
Yes, when I’m about to crash into the Hudson River, I want a robot in the cockpit. Right, nothing would make me feel safer. I think I'll just get out and walk.
17
I think the 155 passengers of US Airways Flight 1549 which landed in the Hudson River in NYC on January 15, 2009 should weigh in on this.
34
Why? What makes them experts on aeronautical engineering?
Yes, but between the recent disaster, the LAM crash of 2013 (rogue pilot murder-suicide), the Continental EWR-BUF crash (pilot error), and likely also the missing Malaysian plane (suspected rogue pilot) , the number of people who would have been saved by autopiloted planes is currently the higher one. And I have no doubt that computers can do water landings soon, if they cannot do so already. What I'd be more worried about is terrorists subverting the computers.
If robotic control were accompanied by increased seat pitch I'd seriously consider it!
38
It is one thing to trust a military combat aircraft or cargo plane to a robot, another entirely to trust the lives of dozens or hundreds of individuals to an automated/robot pilot. While it's true that the vast majority of "ordinary" flights don't require "heroic" intervention from the men and women sitting in the cockpit, I find it nearly impossible to believe that a robotic pilot could have safely landed flight 1549 in the Hudson river with no loss of life.
The flight crew represents a relatively small portion of the cost of each ticket sold by the airlines and I suspect that, even in a fully automated aircraft, most passengers would prefer to pay an additional $25 per ticket to have an experienced crew to monitor the aircraft and deal with any issues that arise.
The flight crew represents a relatively small portion of the cost of each ticket sold by the airlines and I suspect that, even in a fully automated aircraft, most passengers would prefer to pay an additional $25 per ticket to have an experienced crew to monitor the aircraft and deal with any issues that arise.
7
How about having pilots AND a backup fail-safe system.
It would be easy for our software engineers today to come up with an emergency robot that can tell when there are a number of factors that point to an emergency situation like this one and at that point open communications with the tower (command and control, multichannel data, voice and video), signal emergency, stabilize altitude and temporarily take control from the pilots. The plane could be flown remotely at that point in conjunction with the robot. At some safe point after certain safety codes have been entered by both pilots and/or verbal communication has been re-established and the situation deemed safe, then flight control can be handed back to the pilots.
The robot can have a gps map of safe altitudes for its current position and can determine when a maneuver is potentially catastrophic (descent rate, flight path. etc) and can execute different safe scenarios. It can have redundant systems on board and check operations in between them so as to eliminate machine failure as much as possible.
Yes, there would be inevitable bugs in the program, but successive iterations would make the software safer and safer over the years.
You could even have the voice of HAL-9000 calming down the passengers if necessary.
It would be easy for our software engineers today to come up with an emergency robot that can tell when there are a number of factors that point to an emergency situation like this one and at that point open communications with the tower (command and control, multichannel data, voice and video), signal emergency, stabilize altitude and temporarily take control from the pilots. The plane could be flown remotely at that point in conjunction with the robot. At some safe point after certain safety codes have been entered by both pilots and/or verbal communication has been re-established and the situation deemed safe, then flight control can be handed back to the pilots.
The robot can have a gps map of safe altitudes for its current position and can determine when a maneuver is potentially catastrophic (descent rate, flight path. etc) and can execute different safe scenarios. It can have redundant systems on board and check operations in between them so as to eliminate machine failure as much as possible.
Yes, there would be inevitable bugs in the program, but successive iterations would make the software safer and safer over the years.
You could even have the voice of HAL-9000 calming down the passengers if necessary.
2
"Open the cockpit door Hal!!!"
"I'm sorry Dave, I can't do that right now."
"I'm sorry Dave, I can't do that right now."
We've seen recently, and repeatedly, that automation is not a solution that allows pilots out of the cockpit. It doesn't even reduce the number of necessary pilots that need to be up on deck.
These systems fail often. For example, a pitot tube could fail. An automated system would need to able to account for this. Or a heat condition may potentially destroy critical functionality from a circuit board, etc. etc.
As a programmer and computer-user myself, it's my job to build automation, and always try to automate the jobs I'm doing so I can get to the next "promotion".
You know what? Software is only as smart as the folks writing it, except software doesn't usually learn or self-correct very well -- as of YET.
These systems fail often. For example, a pitot tube could fail. An automated system would need to able to account for this. Or a heat condition may potentially destroy critical functionality from a circuit board, etc. etc.
As a programmer and computer-user myself, it's my job to build automation, and always try to automate the jobs I'm doing so I can get to the next "promotion".
You know what? Software is only as smart as the folks writing it, except software doesn't usually learn or self-correct very well -- as of YET.
6
I am strongly opposed pilotless planes and driverless cars for the same reason that I am opposed to religious meddling in politics, or censorship in the Arts -- I am a Humanist.
5
Humans have to remain in charge of computers.
7
How many of you believe that money saved on removal of copilots will reduce ticket prices?
How about we remove all pilots and flight attendants? Then replace taxi, Uber, FedEx, UPS drivers. Packages will be dropped in designated bins (reverse garbage delivery ). Also replace all truckers. It's all doable.
Quiz: What will happen to the new hoards of the unemployed?
How is it possible for the population to accumulate any savings when most of living expenses go toward products and services produced by very few people?
If money serves as a mechanism for exchanging specialized labor then the few people who supply products consumed by the majority must at some point reciprocate by consuming equivalent amount produced by the majority. It is hard to believe this balance can be achieved if most labor is concentrated in service economy. How much service can the wealthiest 1% physically consume?
http://ideabits.blogspot.com/2014/05/job-market-pyramid.html
How about we remove all pilots and flight attendants? Then replace taxi, Uber, FedEx, UPS drivers. Packages will be dropped in designated bins (reverse garbage delivery ). Also replace all truckers. It's all doable.
Quiz: What will happen to the new hoards of the unemployed?
How is it possible for the population to accumulate any savings when most of living expenses go toward products and services produced by very few people?
If money serves as a mechanism for exchanging specialized labor then the few people who supply products consumed by the majority must at some point reciprocate by consuming equivalent amount produced by the majority. It is hard to believe this balance can be achieved if most labor is concentrated in service economy. How much service can the wealthiest 1% physically consume?
http://ideabits.blogspot.com/2014/05/job-market-pyramid.html
9
Spot on, Red Pill. Our insistence on calling job destruction "cost saving" shows that we value shareholder profits infintely more than people's lives. As a society, we need to start considering the human effects of automating jobs, and giving all out money to the people who own the robots.
No ! People need a pilot whose life is also at risk to feel secure flying . Automated flying would be too open to hackers or even terrorists. Keep the humans!!
10
So the person who has final say over whether conditions (e.g. weather) are safe enough will not himself be on the plane!
6
Looks out window at hurricane. "Looks good to me...send it up!"
It seems that the fear of terrorists getting to software while a good concern is misplaced. It they can override the software in an unpersoned plan than they can probably fix it so a pilot cannot control the plane he's flying.
This situation is the same as that of driverless cars. Those that stand to gain the most are attorneys, as suits try to deal with issues of responsibility for injuries, death, and destruction. Who will be held accountable: the airlines, the plane manufacturer, the computer producer, the software writers? And where is the liability if a passenger's onboard gadget messes with the controls, if a terrorist figures out how to hack into the system, if a government experiment takes it down? (See the navy accidentally locking garage door opener controls in Concord, CA and other places.)
Unintended consequences are inevitable, and without a person paying attention and having the ability to immediately override electronic systems, when a situation unanticipated by those who write the control algorithms does occur, disaster is likely and, probably, inevitable.
In addition, we are all too familiar with corporations -- Chase, Target, etc. -- which choose not to spend money to maximize even to the limited degree possible the security of electronic systems. After all, that is what got us into the Germanwings crash scenario in the first place. Airplanes used to have two or three pilots in each plane with requirements that pilots not be left alone in the cockpit. When they figured out lawyers were cheaper than pilots, safety dropped way down the list of priorities.
And will Congress now be bribed into indemnifying airlines against suits as it has with the nuclear industry?
Unintended consequences are inevitable, and without a person paying attention and having the ability to immediately override electronic systems, when a situation unanticipated by those who write the control algorithms does occur, disaster is likely and, probably, inevitable.
In addition, we are all too familiar with corporations -- Chase, Target, etc. -- which choose not to spend money to maximize even to the limited degree possible the security of electronic systems. After all, that is what got us into the Germanwings crash scenario in the first place. Airplanes used to have two or three pilots in each plane with requirements that pilots not be left alone in the cockpit. When they figured out lawyers were cheaper than pilots, safety dropped way down the list of priorities.
And will Congress now be bribed into indemnifying airlines against suits as it has with the nuclear industry?
10
I understand some commenters reticence to fly a pilotless plane(DW from Philly has one of the funniest one). But let's look at it in a more rational way. It would take a computer less than a fraction of second to give an answer to this(19999×3889-25896). That's whyin my opinion computers can do a better job tgan Humans.
. Second as most car crashes have shown in the past, the driver tends to steer the wheels in such a way that will preserve the driver life in a self preservation instinct even though it's not the appropriate action. A computer could come up in a fraction of second with the best possible alternatives and determine which one is the least dangerous for the people inside the car.
. Second as most car crashes have shown in the past, the driver tends to steer the wheels in such a way that will preserve the driver life in a self preservation instinct even though it's not the appropriate action. A computer could come up in a fraction of second with the best possible alternatives and determine which one is the least dangerous for the people inside the car.
I don't doubt that the technology for completely automating a flight from gate to gate is attainable in the near future. And maybe those systems may be even more reliable than pilots. That aspect of the technology doesn't trouble me so much. What does worry me is the security of the hardware and software and the ability of bad actors to compromise those systems. If a terrorist or psychologically unstable pilot gets through the current security checks, he or she ends up with control of one plane. If a bad actor compromises a fully-automated system, he or she ends up with control of a whole fleet of planes.
I'm a San Francisco-based Boeing 777 captain flying long-haul routes, and I actually participated in the NASA "virtual copilot" trial last year. NASA's system was indeed impressive, but it doesn't come anywhere near to simulating the complexity of real-life airline flights, and that's for flights without serious problems. It doesn't take long for a cascade of problems to saturate cockpit crews with two or more pilots in the cockpit, as happened with the Quantas A380 incident some years ago in which four highly experienced pilots just barely avoided crashing. And I don't think there will ever be sufficient protections to prevent terrorists hacking into ground-controlled aircraft. Last year an FAA inspector deeply involved in protecting air-ground digital communications told me that that is a huge concern. It's far better to concentrate on the "low-hanging fruit", such as world-wide implementation of the two-person cockpit rule that has been mandatory in the U.S. since 9/11.
125
You're the second pilot to comment about how pilots can quickly become over-saturated when things go wrong and make that your argument as to why automation is a bad idea.
That conclusion is backwards. That is one of the huge benefits of automating this kind of stuff. Computers can accurately process far more data and information than any human or group of humans could ever dream of.
This notion of being over-saturated and incapable of dealing with a situation because too many things are happening at once goes straight out the window. That's a human problem.
That conclusion is backwards. That is one of the huge benefits of automating this kind of stuff. Computers can accurately process far more data and information than any human or group of humans could ever dream of.
This notion of being over-saturated and incapable of dealing with a situation because too many things are happening at once goes straight out the window. That's a human problem.
1
This pilot's argument suffers from the same weakness as Patrick Smith's above. He doesn't seem to realize that the situation in which a "cascade of problems" was able to "saturate" the cockpit crew is an argument in favor of automating the process, not against it.
So-called "multi-tasking" is a myth. What humans actually do when "multi-tasking" is to rapidly switch from one task to another. Computers can switch much faster. The number of problems that "saturates' the capabilities of four humans wouldn't even cause a second of delay in the calculations of a modern computer that can perform millions of calculations per second.
So-called "multi-tasking" is a myth. What humans actually do when "multi-tasking" is to rapidly switch from one task to another. Computers can switch much faster. The number of problems that "saturates' the capabilities of four humans wouldn't even cause a second of delay in the calculations of a modern computer that can perform millions of calculations per second.
If I'm the one having the medical emergency that calls for landing at an airport different from that originally planned, I'd hate to think my life was dependent on a flight attendent communicating with a robot (HAL anyone?)
5
"In a recent survey of airline pilots, those operating Boeing 777s reported that they spent just seven minutes manually piloting their planes during the typical flight. Pilots operating Airbus planes spent half that time."
This is the kind of deceptive and manipulative rubbish that as an airline pilot really drives me bonkers. What they MEAN is that pilots spend only a short amount of time with their hands ON THE CONTROL WHEEL (or side-stick in the case of Airbus). That does NOT mean they are not controlling the airplane otherwise during the flight. To repeat what I said in my earlier post: Flying remains a very hands-on operation subject to tremendous amounts of pilot input. The automation only does what the pilots TELL it to do: what, when, where and how. It can get VERY busy in a cockpit even with all of the automation running.
This is the kind of deceptive and manipulative rubbish that as an airline pilot really drives me bonkers. What they MEAN is that pilots spend only a short amount of time with their hands ON THE CONTROL WHEEL (or side-stick in the case of Airbus). That does NOT mean they are not controlling the airplane otherwise during the flight. To repeat what I said in my earlier post: Flying remains a very hands-on operation subject to tremendous amounts of pilot input. The automation only does what the pilots TELL it to do: what, when, where and how. It can get VERY busy in a cockpit even with all of the automation running.
43
Not sure what is worse, a looney pilot, or looney ideas like this one getting traction and written about in the NYT.
7
The more complex a system, the more opportunity for failure. And the more overreaction to one particular failure, the more opportunity for more failures.
However, the bureaucratic mentality will never admit that it is the bureaucracy that is at fault.
In this case, setting up the rule to lock the cabin door allowed a single individual to go rogue when the pilot had stepped out. By contrast, keeping the cabin doors open, as used to be the case in the old days, would have prevented this. Of course, you say, then the terrorists will just walk into the cockpit and take over. I don't think that will never happen again. The passengers have learned. Any terrorist trying to do anything would be eaten alive these days!
However, the bureaucratic mentality will never admit that it is the bureaucracy that is at fault.
In this case, setting up the rule to lock the cabin door allowed a single individual to go rogue when the pilot had stepped out. By contrast, keeping the cabin doors open, as used to be the case in the old days, would have prevented this. Of course, you say, then the terrorists will just walk into the cockpit and take over. I don't think that will never happen again. The passengers have learned. Any terrorist trying to do anything would be eaten alive these days!
4
Even better, eliminate the driver in cars. That saves 30,000 lives a year, almost as many as guns take. Tell me that getting into a car and/or carrying a gun are not far more dangerous to us than Isis or Ebola.
1
Without going into detail I spent almost 40 years as an air traffic controller mostly in the NY area and have flown many times in the cockpit of commercial aircraft. Replacing pilots with automation is a fantasy. There is no way a computer could equal a pilot's skill, experience, and intuition especially emergencies.
40
Dig up the article from a hundred years ago that says there will only be a market for 10,000 automobiles, max, because that is the number of the servant class smart enough to drive an automobile as a chauffeur for the few wealthy men who could afford a car. Goodbye buggy whip! Times change, and if you are young enough, you will get in a pilotless vehicle at some point. And, sorry, but Sully can only pilot one plane at a time; even worse, half of all the pilots flying are below average!
And there's no way a computer could be programmed to think quickly, accurately and intuitively enough to beat the best humans at the game of Jeopardy. Of course Watson wasn't exactly "thinking" but it got the job done.
As for commercial airline pilots or air traffic controllers, it's just a matter of time. For the next generation or two, however, the computers will probably be mere helpers, conversing with the pilot in English, or if necessary, in whatever language the pilot natively speaks.
As for commercial airline pilots or air traffic controllers, it's just a matter of time. For the next generation or two, however, the computers will probably be mere helpers, conversing with the pilot in English, or if necessary, in whatever language the pilot natively speaks.
"There is no way a computer could equal a pilot's skill, experience, and intuition especially emergencies. " And just imagine those situations which could cause the computer to get conflicting sensor data or worse, disable the computer itself. Pilots will always be necessary for the non-routine situations for which the computer has never been programmed for. After all, isn't that why pilots spend so much time in simulators practicing for the non-routine. Besides, there will never be a computer that can match the human brain for recognizing non-routine and pulling in learned skills to just those situations.
Wow ! Talk about a knee jerk reaction to a problem that really isnt one. Humans may not be perfect but technology isnt perfect either. There are far simpler steps we can take *today* without much effort. How about we focus on those.
I can see another article coming up when the robot pilot has its first error - "One Idea on Improving Flight Safety: Bring Back the Pilot"
I can see another article coming up when the robot pilot has its first error - "One Idea on Improving Flight Safety: Bring Back the Pilot"
40
The irony that lack of mental health services whether for gaining access to guns, other weapons or airplanes, is to ignore the lack of mental health services and instead get rid of human beings.
How many cars are recalled due to malfunctions? How many auto accidents are due to faulty parts? If a plane crashes due to one of these errors the loss of life is far greater.
How about the potential for hacking? Now terrorists won't even have to board the plane to take it down.
Thanks I would rather take a risk with human error.
How many cars are recalled due to malfunctions? How many auto accidents are due to faulty parts? If a plane crashes due to one of these errors the loss of life is far greater.
How about the potential for hacking? Now terrorists won't even have to board the plane to take it down.
Thanks I would rather take a risk with human error.
22
Technology fails. Humans can disconnect the auto pilot when a sensor fails or the systems don't work. Yes redundancy can reduce the problem. Germanwings crashed because they didn't follow the US practice of having a cabin crew member replace the pilot when he left. The Germanwings crash shows we need two humans. Not one. As for drones-- they crash too-a lot. They can fly Jeff Bezos. Not me.
12
“You need humans where you have humans,” said Dr. Cummings. “If you have a bunch of humans on an aircraft, you’re going to need a Captain Kirk on the plane. I don’t ever see commercial transportation going over to drones.”
Ever? Really?
Ever? Really?
1
This article should be read in conjunction with Bob Wachter's recent series in Medium on what can go wrong in healthcare with an over-reliance an automated systems. He draws a number of parallels between healthcare and aviation that are illuminating. https://medium.com/backchannel/how-technology-led-a-hospital-to-give-a-p...
6
But what about the possibility of a depressed, homicidal, and/or crazy person who is supervising the robot or controlling the plane from the ground, and decides to fly the plane into a mountain?
71
This is exactly the question I had. Not only that, but it would also be a lot easier to sabotage a computer system that's flying a plane remotely than to sabotage a human being flying the plane from the cockpit. Would anyone trust the IT department at an airline to provide 100% certainty that its remote-control flight system is unhackable?
3
Ground controllers can be backed up by a hierarchy of supervisors, with specific authorization required for any unusual degree of access to the plane's controls. And, as noted in the article, the plane itself could be programmed to resist any such attempt.
1
Let's start with cargo planes. If they can fly without pilots for several years, then transfer that to commercial planes. I personally would have no problem with a pilot-less plane if the technology were first proven elsewhere.
In many ways it seems that automated flying is a simpler problem than automated driving: airport runways are much more standardized than what a car experiences in the way of construction zones, faded lane markings (or none at all) , cars pulling out from side streets, pedestrians, animals, etc. Drones in the defense world help advance the technology. And the article says that human pilots currently are only flying for a few minutes of a flight already.
In many ways it seems that automated flying is a simpler problem than automated driving: airport runways are much more standardized than what a car experiences in the way of construction zones, faded lane markings (or none at all) , cars pulling out from side streets, pedestrians, animals, etc. Drones in the defense world help advance the technology. And the article says that human pilots currently are only flying for a few minutes of a flight already.
4
This seems like a good time for the train and boat industry to start pitching a more romantic journey. Those sound more relaxing than the sardine can robot plan.
6
"Sardine Can Robot Plan" is a great band name, though!
2
“You need humans where you have humans,” said Dr. Cummings.
That says it best. You can't lock 300 people in an airborne cage at 30,000 feet without the possibility of human contact with the person responsible for their lives.
That says it best. You can't lock 300 people in an airborne cage at 30,000 feet without the possibility of human contact with the person responsible for their lives.
12
No. I will not get on a plane with a robot. They can't even do a good job of checking out groceries as they are easily confused by such simple things as putting an item in the wrong place.
Do you actually think a robot could have landed a jetliner on the Hudson?
Do you actually think a robot could have landed a jetliner on the Hudson?
30
The answer I believe is "yes", a robot could have landed a jetliner on the Hudson. But "robots" (that is, a more sophisticated automated flight control system than is presently in commercial planes) won't be doing that for awhile in the context of a passenger jet.
Please read this article in Vanity Fair that somewhat tackled this question from a different angle. http://www.vanityfair.com/news/business/2014/10/air-france-flight-447-crash
4
Somewhat, indeed. The author of the VF piece is a well-regarded journalist who is also a pilot. He wouldn't make fundamental errors about aviation that the Times article contains.
1
Perhaps instead of immidiatley trying to solve a problem that does not exist, the media and government should be focusing more on recognizing and treating mental illness. Afterall, that's the real reason this plane hit the ground.
Secondly, the common misconception that all planes autoland is wrong. Many regional jets do not have this feature.
Lastly, as a pilot myself and having been inside a large simulator during emergency training.. there are too many things to go wrong to not have a pilot on board. The systems regularly malfunction and need human input to fix the problem.
Secondly, the common misconception that all planes autoland is wrong. Many regional jets do not have this feature.
Lastly, as a pilot myself and having been inside a large simulator during emergency training.. there are too many things to go wrong to not have a pilot on board. The systems regularly malfunction and need human input to fix the problem.
25
I'm an airline pilot, air travel blogger and author. This pilotless planes discussion is something that comes up all the time, and it never fails to raise my blood pressure.
One of the main issues is that people greatly misunderstand the capabilities of EXISTING cockpit automation, and have a vastly exaggerated sense of what it can do -- the result of articles similar to this one, which take at face value the claims of researchers, aerospace academics and tech aficionados who often have little sense of the operational realities of commercial flying.
Over and over and over we hear about how "automatic" planes are. Yet in truth flying remains a very hands-on operation subject to tremendous amounts of pilot input. It's perhaps a different KIND of input than in the old days: instead of gripping the steering yoke as would've been the case in the 1930s, you're working the various autoflight components, flight management system, etc. But the automation only does what the pilots TELL it to do: what, when, where and how. The other day I worked a flight up from the Caribbean. We had bad weather the whole way, a holding pattern, and had to fly a low-visibility Cat-2 approach. The cockpit was so busy that after we landed my voice was hoarse. Now imagine somebody dealing with the complexities of such a flight from a room thousands of miles away, with dozens of other planes stacked up.
Patrick Smith
Boston
One of the main issues is that people greatly misunderstand the capabilities of EXISTING cockpit automation, and have a vastly exaggerated sense of what it can do -- the result of articles similar to this one, which take at face value the claims of researchers, aerospace academics and tech aficionados who often have little sense of the operational realities of commercial flying.
Over and over and over we hear about how "automatic" planes are. Yet in truth flying remains a very hands-on operation subject to tremendous amounts of pilot input. It's perhaps a different KIND of input than in the old days: instead of gripping the steering yoke as would've been the case in the 1930s, you're working the various autoflight components, flight management system, etc. But the automation only does what the pilots TELL it to do: what, when, where and how. The other day I worked a flight up from the Caribbean. We had bad weather the whole way, a holding pattern, and had to fly a low-visibility Cat-2 approach. The cockpit was so busy that after we landed my voice was hoarse. Now imagine somebody dealing with the complexities of such a flight from a room thousands of miles away, with dozens of other planes stacked up.
Patrick Smith
Boston
270
Hi Patrick,
I miss your column on salon.com. I remember catching your ire about a decade ago when I sent in a letter essentially asking the same thing.
My question (not for you, but for those who would automate the entire industry) is - what's the motivation in doing so? Commercial flight is now drastically cheaper and safer than it's ever been. That $340B number that's rolled into overall labor costs - how much of that is paying pilots, as compared to flight attendants, ground personnel, and maintenance crews? My guess is a very small fraction. The cost savings angle seems overblown at best.
I just don't see the need, the system seems to be working quite well as is.
I miss your column on salon.com. I remember catching your ire about a decade ago when I sent in a letter essentially asking the same thing.
My question (not for you, but for those who would automate the entire industry) is - what's the motivation in doing so? Commercial flight is now drastically cheaper and safer than it's ever been. That $340B number that's rolled into overall labor costs - how much of that is paying pilots, as compared to flight attendants, ground personnel, and maintenance crews? My guess is a very small fraction. The cost savings angle seems overblown at best.
I just don't see the need, the system seems to be working quite well as is.
1
Thank you, Patrick, for fighting the good fight. When a senior science writer for the NY Times is this misinformed, it shows how badly we need voices of sanity like yours. Keep these guys honest!
1
Key words in your post include "existing" and "remains". Nobody is saying planes are ready to be automated right now
All of those things you mentioned (like maintaining a holding pattern, avoiding certain types of weather, or coordinating with dozens of other planes) could surely be automated given time for technological progress to occur -- especially the part where you mention "somebody" dealing with those complexities. You're right. Somebody thousands of miles away might have a very difficult time processing all that stuff. "Something," on the other hand, can process infinitely more information much quicker than we can.
This will become reality eventually. Self driving cars are already almost here, and creating those is a more impressive technological feat.
Much larger human element to worry about when an automated car is closely surrounded by cars driven by people with babies screaming in the back seat than when a plane is essentially alone at 36,000 feet.
All of those things you mentioned (like maintaining a holding pattern, avoiding certain types of weather, or coordinating with dozens of other planes) could surely be automated given time for technological progress to occur -- especially the part where you mention "somebody" dealing with those complexities. You're right. Somebody thousands of miles away might have a very difficult time processing all that stuff. "Something," on the other hand, can process infinitely more information much quicker than we can.
This will become reality eventually. Self driving cars are already almost here, and creating those is a more impressive technological feat.
Much larger human element to worry about when an automated car is closely surrounded by cars driven by people with babies screaming in the back seat than when a plane is essentially alone at 36,000 feet.
Replace Captain "Sully" Sullenberger's experience with a computer?
No, thanks.
No, thanks.
37
On Deember 8, 1972, United flight 553 crashed at Midway Airport in Chicago after receiving a wave-off. climbing too rapidly and stalling. Dorothy Hunt, the wife of Watergate conspirator E. Howard Hunt was killed in the crash, and the $10,000 found in her purse played a role in breaking the case wide open. Two weeks later, I was on a TWA flight from Cleveland to O'Hare that received a wave off and circled the airport at about 200 feet, coming in for a smooth landing on the second pass. You could hear a pin drop as we made the circle, because everyone knew what had happened at Midway. Once the plane came to a stop, the captain informed us we had been waved off because a small general aviation plane without radio contact to the tower was crossing our runway as we approached. Everyone applauded, and the captain who emerged from the cockpit a minute later had a full head of white hair. Thanks, I'll take an experienced pilot any time.
The Germanwings problem is one of a young, inexperienced pilot with a history of mental illness left alone in the cockpit. It's a human failure that can be eliminated through screening. Not much you can do about an electronic failure if no one's there to override the machine.
The Germanwings problem is one of a young, inexperienced pilot with a history of mental illness left alone in the cockpit. It's a human failure that can be eliminated through screening. Not much you can do about an electronic failure if no one's there to override the machine.
127
Yes, there is something we can do about electronics failure: it's called redundant systems. We use them in the space program, where the margins for error are much thinner and the odds of electronics failure are higher due to exposure to cosmic rays. It's not a new idea. Humans aren't completely reliable either, which is why we have two qualified pilots up front, instead of one.
1
It seems that the risk of a terrorist hacking/disabling remote operation of an aircraft is far greater than the relatively modest (based upon miles flown, exceedingly small) risk of a rogue pilot. Losing an unmanned drone due to hacking probably means loss of equipment. Losing a passenger jet means loss of human life. This is unacceptable. One sensible approach would be increased evaluation and mental health support for pilots and more gate keeping for the profession.
45
I was scanning the comments looking for someone to bring this up. Imagine another 9/11-type event, but with a thousand airliners simultaneously.
Even though the telemetry networks aren't yet capable of full aircraft control, if they were, hackers could remotely pilot aircraft at will.
Even though the telemetry networks aren't yet capable of full aircraft control, if they were, hackers could remotely pilot aircraft at will.
2
We have loads of instances of crashes CAUSED by pilot error. But how often do sentient human pilots PREVENT crashes? That's much harder to quantify, because pilot intervention that prevents an accident rarely makes the news. Would a current- or near-technology robot have made the correct decision to land US Airways Flight 1549 in the Hudson, saving the life of every person on board? That was the ONLY survivable option in that case, and I don't think a robot would have made it. I think we're many years away from that level of artificial intelligence.
In the meantime, however, there's room for AI technology to enhance the safety of human pilots. It isn't very hard to imagine flight management software that would have recognized the German co-pilot's instructions to the autopilot as abberant, placed the plane in some sort of "safe mode," and contacted the ground for further instructions.
In the meantime, however, there's room for AI technology to enhance the safety of human pilots. It isn't very hard to imagine flight management software that would have recognized the German co-pilot's instructions to the autopilot as abberant, placed the plane in some sort of "safe mode," and contacted the ground for further instructions.
65
Good comment
And then the plane would have crashed upon running our of fuel. The co-pilot locked them all out of the cockpit, remember?
3
Remember Murphy's Law: Anything that can go wrong will. And while we're talking electronics, last week I was in the parts department of my local car dealer. A mechanic came out to pick up a part he was replacing, a side mirror. He stated he needed to be careful installing it as it costs $ 600.00. That's right, a part the should cost just a few dollars was $ 600.00. The mirror contained a sensor to detect vehicles passing on the drivers side and emits an alarm if the driver tries to pull to the left. The driver was relying on this (safety) device and didn't look before turning. He barely missed the other vehicle. That's how he learned it had failed. Get on an airplane without a pilot? Or trust my car to a computer? No way.
49
Murphy's Law is often misunderstood. It's not a fatalistic expression of futility. Its meaning is that given a sufficiently large number of iterations, any flaw that is left in a system - intentionally or not - will eventually manifest itself.
So, for example, if you build a mousetrap that works 99.999% of the time a certain number of mice will get away. That's probably OK, but if the same mechanism is used to fire an ICBM it's a good idea to design it differently.
With regard to piloting planes, a certain very small number of fatal computer-controlled crashes is acceptable, provided that number is smaller than that of any human-controlled system we can devise.
So, for example, if you build a mousetrap that works 99.999% of the time a certain number of mice will get away. That's probably OK, but if the same mechanism is used to fire an ICBM it's a good idea to design it differently.
With regard to piloting planes, a certain very small number of fatal computer-controlled crashes is acceptable, provided that number is smaller than that of any human-controlled system we can devise.
Massive overreaction to a single incident. I'll put my trust in human pilots, thanks.
104
I'm with you in trusting humans -- but counting this one there have been a total of about 8 of these incidents in the last few years.
Eight? Out of millions of flights? Sounds fine to me.
2
Is there a point in the aging process where one just feels, No, no more, I will not adapt to more changes? If so, I just reached it. I won't get on a remote-controlled airplane. Yes, I get that they're nearly automated anyway, but I won't do it. I'll just stop flying, if/when it comes to that. I remember my grandmother saying/feeling that way about flying per se, and I guess I'm at that point now. Y'all carry on, but when pilot-less airplanes become a thing, I'm staying on the ground after that.
115
I agree, at this point I would not feel comfortable flying an aircraft that can be controlled or overridden from the ground. Luckily you sound like you are older than I am and able to retire, but I still have 35 years to go. I'm sure things will change from then until now.
1
For me, the aging process point was reached when flying became a burden and not even remotely enjoyable. I've discovered Amtrak and have found that, although the trip might be a bit longer, the various pleasures far outweigh anything the airlines have to offer.
I'm not patted down with stranger's handling my 82 yr old body. I don't have to undo my double knotted shoestrings and let them see my shoes. I can take any size bottle or tube in my suitcase that I need. I have good leg room, the seats are ample, I can get up, go to the snack bar, look out at the scenery passing by, and often find my seatmate both pleasant and as relaxed as I am.
Flying with or without robots as pilots, I'll pass. Thank you very much.
I'm not patted down with stranger's handling my 82 yr old body. I don't have to undo my double knotted shoestrings and let them see my shoes. I can take any size bottle or tube in my suitcase that I need. I have good leg room, the seats are ample, I can get up, go to the snack bar, look out at the scenery passing by, and often find my seatmate both pleasant and as relaxed as I am.
Flying with or without robots as pilots, I'll pass. Thank you very much.
The key is the pilot has a much greater personal incentive to land safely than a remote operator (or software developer). I don't want to fly on an airplane where the operator isn't taking the same personal risks that I am.
5
Will people board planes without pilots? I will.
Something simple, like an override, would be a fine start. "I can't let you do that, Dave" when the pilot sets the altitude to 100 feet (and systems are telling the navigation that the ground is 6000 feet) would have saved the Germanwings passengers.
And when can I get my google car? That will save a lot more lives, when the cars all talk to each other, and people don't get in their way.
Something simple, like an override, would be a fine start. "I can't let you do that, Dave" when the pilot sets the altitude to 100 feet (and systems are telling the navigation that the ground is 6000 feet) would have saved the Germanwings passengers.
And when can I get my google car? That will save a lot more lives, when the cars all talk to each other, and people don't get in their way.
15
The "I can't let you do that, Dave" approach is useful in preventing pilot errors. It will not stop a pilot from deliberately crashing an airplane.
1
HAL 9000 was in the process of committing murder when it spoke to Dave, Blue State.
2
When can you get our Google car? Never. Not in your lifetime or your grandkids'. Mostly because you can't mix driverless and driver cars on the road at the same time. It's change them all over at once or not at all. How's that gonna happen?
Three such documented accidents in the history of aviation and we're jumping to this? Whose hands would you like to be in, in the event of a double bird strike flying out of NYC? A robot's, or Chesley B. "Sully" Sullenberger's?
95
Excellent point about jumping to needless conclusions after three accidents over many decades, but the Sullenberger thing always irks me. It irks me because it implies that pilots are only useful in UNUSUAL circumstances. As I described in the accompanying post, even the most routine flight is still subject to tremendous amounts of pilot input. This is lost on many people, who are led to believe that planes basically "fly themselves," and that pilots are on hand mainly as a backup. This is total rubbish, yet we hear it again and again. A computer is not flying your plane. Pilots are flying your plane. The automation only does what we tell it to do.
2
But what about terrorist hackers taking control of an aircraft. They certainly have been attacking other targets. See, for example, http://www.nytimes.com/2014/12/03/world/middleeast/report-says-cyberatta...
21
This is yet another good example of how people totally misunderstand how planes fly and how pilots interface with cockpit technology. And articles like this one are much to blame. See my accompanying post. Your Boeing is not flown by a Microsoft computer that can be "hacked" in the sense that you're thinking. It is flown by PILOTS, who have more or less total control over the aircraft.
PS
www.askthepilotcom
PS
www.askthepilotcom
2
True, but I think a lot of commentators are extrapolating to a future which includes real-time remote control of an aircraft in flight. I agree we're not there yet, however. Thankfully.
I'd guess those seven minutes defined in the article are for takeoff and landing. During level flight in calm weather, the autopilot probably does a great job.
Besides, then you'd need to only feed the dog, right? (kidding)
I'd guess those seven minutes defined in the article are for takeoff and landing. During level flight in calm weather, the autopilot probably does a great job.
Besides, then you'd need to only feed the dog, right? (kidding)