Slashdot Mirror
Planes Without Pilots
HughPickens.com writes: John Markoff writes in the NY Times that in the aftermath of the co-pilot crashing a Germanwings plane into a mountain, aviation experts are beginning to wonder if human pilots are really necessary aboard commercial planes. Advances in sensor technology, computing and artificial intelligence are making human pilots less necessary than ever in the cockpit and government agencies are already experimenting with replacing the co-pilot, perhaps even both pilots on cargo planes, with robots or remote operators. NASA is exploring a related possibility: moving the co-pilot out of the cockpit on commercial flights, and instead using a single remote operator to serve as co-pilot for multiple aircraft. 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. "Could we have a single-pilot aircraft with the ability to remotely control the aircraft from the ground that is safer than today's systems?" asks Cummings. "The answer is yes."
Automating that job may save money. But will passengers ever set foot on plane piloted by robots, or humans thousands of miles from the cockpit? In written testimony submitted to the Senate last month, the Air Line Pilots Association warned, "It is vitally important that the pressure to capitalize on the technology not lead to an incomplete safety analysis of the aircraft and operations." The association defended 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 (PDF) and begin to formulate a course of action before even sophisticated sensors and indicators provide positive indications of trouble." Not all of the scientists and engineers believe that increasingly sophisticated planes will always be safer planes. "Technology can have costs of its own," says Amy Pritchett. "If you put more technology in the cockpit, you have more technology that can fail.""
Automating that job may save money. But will passengers ever set foot on plane piloted by robots, or humans thousands of miles from the cockpit? In written testimony submitted to the Senate last month, the Air Line Pilots Association warned, "It is vitally important that the pressure to capitalize on the technology not lead to an incomplete safety analysis of the aircraft and operations." The association defended 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 (PDF) and begin to formulate a course of action before even sophisticated sensors and indicators provide positive indications of trouble." Not all of the scientists and engineers believe that increasingly sophisticated planes will always be safer planes. "Technology can have costs of its own," says Amy Pritchett. "If you put more technology in the cockpit, you have more technology that can fail.""
'Nuff said
Many airline disasters are caused when sensors go wrong or the output is confusing. Air France 447 was one of those incidents. It would be even more confusing to a remote pilot.
And get rid of the large amount of useless bureaucratic sludge slowing everything down? Oh no, automation is only good when *you* lose your job and *they* save money.
Mostly random stuff.
Just connect the players to remote control the real planes. Same could be done with shooting games and robot soldiers in the middle east. Or people would grow the vegetables through farming simulators.The possibilities are endless.
...then it's not safe enough to have passengers.
TBH I'm not sure whether Poe's law is in effect and I'm witnessing a parody of what happens when people impose technical solutions to social problems (one in a million pilots homicidal? replace pilots with computers!).
There are good reasons to increase automation, but not to remove the possibility of dedicated human intelligence provided with the full gamut of sensory clues for complex (esp. exceptional) situations.
I'm still blown away that there is no active ground monitoring of all flight wherever they may be in 2015. We still have planes "disappearing" because nobody knows where they are once they leave the ground.
So instead of having two pilots, why not have a computer monitoring system that actively monitors airplanes with only 1 pilot in it. Any weird actions by the pilot would trigger a warning allowing ground operators to override it. Boom, no more missing planes, or suicidal pilots.
did you forget to take your meds?
Sadly, planes without real pilots worked well on 9/11.
Political correctness is really just herd psychology pushed by insecure people who desperately seek social conformity.
and we will need basic income to cover the people automated out of a job or we can kind get the same thing with people who will get locked up just for room / board / doctors at a much higher cost.
can we automate please the idiots who think the human is always a problem, so that we finally have an ENDLOESUNG for the 'telephone cleaners'.
Such as the pilot's control hardware, indicators, windscreens, oxygen supply....
And I know some people will just go, "Well, what if the pilot or copilot on the ground goes rogue and takes over?" The obvious response is that they'd be operating in a secure facility with dozens of other people and extensive supervision; nobody is ever going to just let any random person secretly take over a plane without anyone else knowing. And no, ATC systems are not net connected. I work in an ATC center, when I need to look something up online while trying to debug a problem I have to use my cell phone or go back to my office.
Probably the simplest solution to all of this would just be an additional entry to the CPDLC standard, and the hardware changes necessary to support it: one that locks out the pilot from all control and switches the autopilot on, set to the last flight plan agreed to by both ground and the pilot. One would of course have to make sure that there's no way for the pilot, while he's still in control, to sneakily break the datalink communication stream fast enough that ground wouldn't have time from the onset of suspicious activity to send the command.
Having highly reliable communications would be critical for any pilot-override or remote piloting system. In each case, any cutoff in communications should force on the autopilot as per above until communications could be reestablished.
Trump's plan to get rid of Mueller appears to be 'be so guilty of so many things that Mueller works himself to death.'
A pilot saves all passengers by emergeny landing in the Hudson river, and everybody prÃzises the virtues of human pilots. A pilot Mills himself and 149 passengers and crew, and promptly we should geht rid of human pilots.
“We say, ‘Well, I’m going to cover the 98 percent of situations I can predict, and the pilots will have to cover the 2 percent I can’t predict."
http://www.vanityfair.com/news/business/2014/10/air-france-flight-447-crash
So the flight control software will be able to handle 100% of the situations? Or will this be a case that the flight control software just has to outrun a pilot and not the bear?
Fully automated aeroplanes have been feasible for a while. But are they something we ought to desire? I think not.
I don't even need to trot out the "corrupted software" or "cyber attack" or such scenario, even if they are far less outlandish than you might believe (but not as easy as hollywood would have it). No, there's a simpler and much more tractable thing to think about.
Even fully automated drones of war that in theory can do the whole thing from start to landing on their own, can be and regularly are piloted from afar, as normal part of most any mission. It's going to be unlikely in the extreme that fully automated passenger aeroplanes will not have such a facility. The trouble with it is that such remote pilots don't have quite the same stake in the game as an on-board pilot.
For that reason alone I want a human pilot on board. Even if a very few of the thousands of pilots in the world turn out to be suicidal.
And in fact, shooting for no pilot on board is the same sort of error that anti-terrorism doors in aeroplanes turn out to have been. Good idea, but not without exploitable flaws. Given that these days passengers have wisened up and will tackle and sit on a would-be terrorist to the point that passengers have done so more often than on-board air marshalls have helped, it would be wiser to re-think those doors (and the marshalls, and the pervy scanners, and all those other measures, most of which have turned out to be damaging, expensive in every sense of the word, and benefit-free) than to try and shove the pilot off the aeroplane.
So, yes we can. No we shouldn't.
Let me get this straight...the thought is to propose a drone-based airline? While it may be mostly auto-pilot, there's a human remote co-pilot to watch over things, just in case. So, how many planes does this one co-pilot watch? What's the lag time between aircraft and their remote-cockpit? How fast do the backup remote-comm systems kick in when the primary is getting wonky? And what keeps the remote co-pilot session from being hacked/co-opted and then turned into a remote drone-weapon?
I ..just don't think it'll happen in my lifetime. I could be wrong, but ..I'm not putting any money in that stock today.
Awk! Pieces of eight. Pieces of eight. Pieces of seven... ERROR: General Protection Fault. [Paroty Error.]
So instead of a single aircraft that can be taken down by a single person, we have a disgruntled ground operator who can take down a dozen within minutes.
I think one important part of the trust we have in pilot is that they get to die with the rest of us if something goes wrong. Where it is obviously not a foolproof system, it is much more inherently trustworthy than a remote pilot. Also what happens when two autopilots have concurrent problems and they have the same remote co-pilot?
Let the military do several thousand cargo plan trips first. Then a few thousand military transport trips. THEN let civilians do some cargo trips.
When you have a pilot in the plane it is hard to make him fly a second or third. When the pilot is on the ground monitoring a couple planes it is easy to add a third, fourth, twentieth etc. Pressure to lower HR costs will make that happen until failure rates are shown.
Let the military find where those rates are at. BTW there was an article on the burn out rate of the contractors flying drones.
I have no desire to fly on one of these. I believe that human confidence in technology is a little too exuberant. I assure you that government and business will only take this risk with people who are expendable, meaning you will not see Air Force One with one or no pilot anytime soon.
...being the poor prick in charge of multiple airliners when the shit hits the fan.
How would the vast amounts of data required to fly the plane remotely be ensured? Now multiply that by all available commercial flights. And while you're at it how would you cover planes flying over the oceans?
We don't even have automated trains and they run on freaking tracks, honestly it's not a huge technical issue to get rid of the pilot, it's way more of a policy and social problem - at any moment there are many UAVs flying around autonomously
Whatever people invent some other people are gonna try to exploit it, and in some cases they will succeed. .. whatever.
To successfully prevent people murdering other people you need to treat the cause of that intent. Or at least start thinking about it instead of just forcing another dumb law or trying to put a camera in the cockpit, remote control
In 20 years there will be so many laws and regulations it will be impossible to live.
And when the first plane crashes due to a bug in the pilot software, we all start wondering again if removing the pilot was a wise decision.
This whole Germanwings plane crash shows, again, one important thing: people suck at dealing with risks. Several hundred thousands of flights went well. The last incident with a pilot causing a plane to crash was back in 1995. The Germanwings plane crash was an incident. We must learn to treat it that way, as an incident. No reason to panic and start changing policies, rules and procedures. With every change, new risks and new ways of things to go wrong will be introduced. When that happens and you again make changes, you end up in a loop of changing things. The result: the changes will cost a lot of time, energy and money while the risks are not reduced.
We need to start accepting that risks are part of our life. Unacceptable risks need to be dealt with, but more important: acceptable risks should be accepted, even when they occur!!!!
It doesn't have to be like this. All we need to do is make sure we keep talking.
and we will need basic income to cover the people automated out of a job
Airlines are already complaining about a shortage of pilots today. This will take years and people growing up will have the opportunity to pursue something else, like, I don't know, the technology for automating aircraft. And those military drone pilots will have an almost direct path into civilian employment, assuming that they can get over Macho Grande.
Just put two reasonably competent people in the cockpit at all times and stop trying to f**k an extra penny out of every dime, you cheap chiselling b*st*rds.
Right now many feeder airlines are barely paying a living wage for their junior cockpit staff, so stop pretending that the personnel costs are going to put you out of business. You're certainly not passing along the recent fuel cost savings to us sardines.
(I haven't had my coffee yet, so that's my excuse for the "negative tone" in this post.)
Sometimes the "writing on the wall" is blood spatter...
If they did, they would already have automated cargo planes.
SJW's don't eliminate discrimination. They just expropriate it for themselves.
They plan on taking the pilot out of the plane, but can't get an economically viable solution for recording all flight information remotely?
Anyone want to guarantee 100% perfect security for ANY wireless communication? Because if we have remotely piloted airliners (either because there's no pilot, or the pilot is suicidal) someone WILL hack into it.
I can't imagine any danger at all in having the ability to remotely fly a commercial plane. There is no way that would ever be abused.
Now I have to check if my insurance cover death by division by zero.
Remote operation imply data transfer, usually by radio, and this is the weakest part of the system. To ensure that the aircraft stay in safe operation without remote link, it fist must be able to sustain stand alone operation, including landing before running out of fuel.
Concentrating control to a single point will increase the risk, concentrating control to a remote point will add an another layer of risk. The only solution is to allow distributed control, and this imply that each aircraft is able to operate safely by itself without remote control.
If you are not convinced, just imagine that a remote control point is unable to operate for some reason: you have now dozen of flight without co-pilot and there need to all land as soon as possible, raising a another wave of problems.
descended to 100 feet into a mountain, right? So the autopilot has no failsafe, it just follows whatever flight commands are made?
I guess we can just rely on the sanity of the flight path programmer, the autopilot, various sensors, and the easily spoofed GPS network to keep us safe.
Now that is a worry. A suicidal engine is better than no engine. Believe. You. Me. Even a German suicidal engine is better than no engine, sacht der unter gleben glauben globen Meinsel Menschen!
in the aftermath of the co-pilot crashing a Germanwings plane into a mountain, aviation experts are beginning to wonder if human pilots are really necessary aboard commercial planes. ...a ground controller might operate as a dispatcher, managing a dozen or more flights simultaneously.
Or, a depressed, mentally-unstable ground controller might crash a dozen or more flights simultaneously into mountains.
Or, Skynet might crash all the flights in the world simultaneously.
Software without developers? Self replicating software attending to its own bitrot, the next phase of Systemd internal version by evil Pottering team. It's heard for us to conceive of AI? Conceive it this way: AI is software that only tends to itself, not humans. malware gains seem to scale with the market in complexity but it always linked back to the gain to its author. Systemd, consuming all for itself. True AI. Systemd tending to its own bitrot, global takeover.
It reminds me of the complaining that happens in high tech companies. They bitch and moan about there being no qualified employees, when what they really mean is... there are no qualified employees that are willing to work for the crappy salaries and dismal working conditions on offer.
Typical bureaucratic thinking. They created the problem by removing the crew's responsibility to protect their aircraft and replacing it with an impenetrable vault door. MANY failure modes were easily foreseeable 13 years ago. From suicidal pilots, to simple medical emergency, and a dozen others this was a stupid idea from people that didn't know what the hell they were talking about.
Now rather than remove the problem, they want to double down on Central planning and control. Given the current bleeding edge state of the art technology and adding best case advancements over the next 10 years, I think I'll drive - and possibly buy an armored car to do that in as well because it is going to be raining airplanes. The one certainty is whatever their next plan is, it will be worse than the last and make us look back at 2 suicides in a decade as "the good old days".
How about the Gimli Glider? If you run out of fuel, you suddenly have no power. Hard to run on automation without it. There have been a number of successful landings after a plane runs out of fuel, that could never have happened under automation. One of the success factors in USAirways flight 1549 was Capt. Sullenberger turning on the APU as soon as he realized he had no engines. Hard to do remotely when you are already disconnected.
I have wondered already, why the reckless ideas didn't poped up early and where invalided by sane arguments:
* remote control means == remote attack vector/system error, could crash a thousand planes at once
* humans can react indvidually on unknown problems (wrong and right, but at least they can)
* never ever a sane person will give his life to a computer (did anyone trust windows? linux? macos?)
* planes are not cars, if a computer on a car "fails" it will probably crash, if it fails on a plane *it will crash*
And here an good example to trust pilots in case of emergency (german source):
http://www.sueddeutsche.de/pan...
Short: ...
1. wrong sensor data
2. computer believes sensors
3. computer starts fast descent
4. experienced pilot decides to turn of flight-computer
5. 109 passenger saved
5. airbus releases updates for flight-computer and new manual for pilots
This story was released only four days before the tragic incident with Germanwings. Flight number "LH 1829" means Lufthansa, same basically same airline.
This technology was already demonstrated on 9/11
Why is every story submitted by this guy such bullshit? Why does it always get approved?
It makes complete sense we can't agree nor have widespread self driving cars, fuck it! Let's do planes instead.
Sometimes do you read headlines and think what the fuck?
I am Bennett Haselton! I am Bennett Haselton!
What about a bit of actually professional risk management? That one says this is an event that is exceptionally rare and that hence does not need countermeasures. The very extent of the press-coverage shows how exceptionally rare it was.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
Thank you!
You're right by logic and feelings at once.
e.g., ILS/VOR at the airport inoperative, or not all of the redundant components fully functional.
Now with a human pilot, this rarely is an issue. No single issue brings down a plane.
Fully automated flight with a faulty sensor and ILS inop? Well...
To switch to fully automatic, much tighter margins on faults are needed. Causing more canceled flights. Or more expensive aircraft (even more redundancy) and airports (even with today's regulations some don't get completed).
someone WILL hack into it.
It's worse than that - all they need to do is jam it which would be trivially easy to do. For example if you put powerful transmitters into a van, parked it somewhere on the approach path to a busy airport and turned it on you would suddenly have craft who were on approach lose all control and by the time authorities tracked down the van and shut it off who knows how many planes would have crashed.
Remote control planes with passengers on are a stupendously bad idea. There is no way I'm flying on a plane which is not under the control of someone onboard whose life also depends on the plane landing safely. Even with such a strong motivation as that we have seen disaster happen - how much more likely will it be if the pilots are sitting remotely and have even less at stake? Suddenly things like disgruntled employees crashing planes becomes imaginable.
It's a very old joke by now but...
The cockpit of the future will have two seats - for a pilot and a dog. The dog's job is to bite the pilot if he touches anything.
The pilot's job is to feed the dog.
This is what tech has done for us in the bad sense, it appears to me people no longer want to work with people, but rather will look to find a way without them, to use a machine rather than a human.
In some odd way our technological modern lives are causing to dislike humanity.
"If any question why we died, Tell them because our fathers lied."
If we want to prevent suicidal pilots, removing the second pilot, the only one capable of stopping him, is definitely not going to help with this. This is strictly a money saving venture, that will remove a layer of safety and redundancy.
Troll is not a replacement for I disagree.
Having supported NASA Global Hawk missions over the poles, I can tell you that high bandwidth satellite communications becomes dicey over the poles. We would fallback to Iridium... which was less than ideal and totally unsuitable for command and control. This was not a huge issue, should the plane have lost C2 comms it would simply return to base.
I am very small, utmostly microscopic.
So you want to give the bad guys a way to take over control of a plane remotely and crash land it into a big city, without even needing to die in the accident themselves ?
Sounds like a terrorists wet-dream...
And I am not used to do "think of the children" kind of spech, but allowing anyone outside the plane to take over control is the dumbest idea I ever heard.
And you know it is dumb because at commercial scale such a system would be full of explitable holes, and a determined person *will* get into the system and take control of a plane.
I guess we'd need to see the statistics for the failure-rate of drone flights, not just for 'mission terminating'-level incidents, but for every crack-up on landing etc that would cause passengers problems. Compare this to the failure-rates of flights that are human-piloted, including to some degree the number of 'cockpit events' where a pilot's human reflexes and situational awareness avoided an incident entirely.
Most likely, it's not even close. Planes are still staggeringly safer in the hands of human pilots - even recognizing their very-human fallibilities - than computer piloted.
In the same sense that people have greater fear of flying than driving (despite ample evidence as to which is more dangerous), while people will cheerfully get into driverless cars, I suspect pilotless airplanes will never...ahem... 'take off'.
-Styopa
So instead of having two pilots, why not have a computer monitoring system that actively monitors airplanes with only 1 pilot in it. Any weird actions by the pilot would trigger a warning allowing ground operators to override it. Boom, no more missing planes, or suicidal pilots.
Ok, Define "weird actions". What specific circumstances does the computer take over? How do you plan to have a computer program with adequate situational awareness to never make a mistake when overriding the controls? Remember, you are by definition working with weird corner cases here.
Just saying "let the computer override" is a nice notion that isn't really reasonable when you actually think about the details of what is occurring. Computers cannot generally deal with situations that were not considered in advance nor are they very good at certain types of situational awareness.
90%+ of comments here have been regarding lack of onboard pilots with commercial passenger flights.
Naturally, the first offboard pilot flights would be with cargo only. And that is way more relevant and less sexy discussion.
-- I was raised on the command line, bitch
Use freight flights to test the system. If it works OK for a couple of years flying freight, then try people.
Lure them on with cheap tickets to be the guinea pigs.
The number of people saved over the years by pilots doing things with airplanes that no computer system (or remote operator) could ever have done vastly outnumbers the number of people killed because of deliberate actions by pilots.
Examples of such heroic flights where lives may not have been saved without pilots in the cockpit include:
US Airways Flight 1549
British Airways Flight 9
Air Canada Flight 143
British Airways Flight 38
Northwest Airlines Flight 85
TACA Flight 110
United Airlines Flight 232
Reeve Aleutian Airways Flight 8
Ah, the race to the bottom. And at what point will the Minimum Wage TSA agents be remotely flying the planes? And then the airlines get the bright idea to outsource all that to India or maybe Vietnam?
If telephones are outlawed, then only outlaws will have telephones.
What the article in NYT seems to omit is that the German Wings mishap could have been prevented if the Captain was able to regain access to the cockpit. But in the politically-charged aftermath of 9-11, we mandated a cockpit-door system on aircraft that under certain circumstances has no back up option for the event that the captain is locked out of the cockpit. A more careful analysis of all the risk at the time might have produced a system with at least some means available for the Captain to open the door from the outside if necessary. Going pilotless falls into the same category of reactive thinking because it is creating another system without any backups. Computers may not be suicidal and they may not make the same kinds of mistakes that human pilots make, but they do make different ones. What's missing from the NYT article is information about the safety record for pilotless aircraft. They can and do fail. Good aviation safety requires taking into account the big picture and making educated tradeoffs to minimize the total risk. Quick fixes or knee-jerk reactions often end up creating new hazards and often increase the total level of risk.
I think a lot of the automation proponents are missing the point and fail to understand the role of a pilot in command of an aircraft. He is not just there to steer the aircraft. He is there to accept and take responsibility for the safety and well being of the aircraft and all those who are riding on it. If we go pilotless, who will check the maintenance records and accept that the aircraft is air worthy? Who is responsible for ensuring that the weather meets minimums? Who is responsible for ensuring that the aircraft is properly loaded and balanced and that the fuel load is adequate for the planed flight and any unplanned contingencies? I would rather trust a person who's ass is sitting in the same aircraft than some faceless manager who is sitting the ground who presses "GO". We have known for at least the past 1000 years the importance of having one person on board who is in charge who has the authority to make decisions when things go wrong. Technology will not change this.
what about satellite internet lag ON live failed and the lag there on land lines was bad for gaming.
A 4 seater prop plane (Cessna 172) has an hourly flight cost in the ballpark of $100 hour, not counting pilot cost. A nonliving wage for the pilot would be only a moderate fraction of the operating cost. I don't see how there will be major cost savings from getting rid of a pilot. Maybe, an autopilot will allow for a less experienced, and cheaper, copilot.
As a pilot, I would never fly an aircraft which has a remote capability to take control away from me.
Fair enough, why? What specific objective reason(s) causes you to oppose the idea completely? I'm not for or against the idea but I'd like to hear why it is a good or bad idea. I understand the issue of situational awareness by the remote operator could be an issue. What else?
To suddenly decide that we can't trust the crew despite the fantastic safety record aviation has is just ludicrous.
Agreed though that also doesn't mean we shouldn't analyze the situation thoroughly to see if anything can practically be done. I generally share your sentiment that the safety record of aviation is great but it got that way by carefully examining disasters to see if any improvements could be made. Maybe there is an opportunity of some sort here to introduce improvements.
the lone gunmen Episode 1.
With the X-files coming back maybe they can find out what really happened to MH370.
The whole point of having a human crew on a plane (or a ferry, or a train) is to be assured that someone with some authority in the company has made all the efforts to be sure the equipment is in fine conditions - otherwise he/she will die to.
Fringe cases when a pilot wants to die like the Germanwings case are not enough to invaidate this main theory.
>NASA is exploring a related possibility: moving the co-pilot out of the cockpit on commercial flights, and instead using a single remote operator to serve as co-pilot for multiple aircraft.
Could you imagine what would happen to the person flying an aircraft in an emergency that crashes? That'd be like the captain not going down with the ship, and if anyone remembers South Korea, that did not end well at all. They'd want his bloody head on a stick. The only reason people don't go after pilots when a plane kills their loved ones is because the pilots are already dead!
Latency...
Risk management is Risk=(event probability)*(event political/PR/monetary cost/legal)
Thus even a RARE even will have to be considered, if its cost is much higher than the company can afford, legally, monetary or politically.
We don't even have trains being run by computers yet and there is no driving involved, just acceleration and deceleration. Not only that, proper tailoring of acc/dec curves can save tremendous amounts of fuel. While successfully navigating a car in 2D or a plane in 3D may be much cooler than the mundane operation of a freight or passenger train, until we can have zero train accidents through the use of driverless trains, discussion of other forms of driverless transportation is premature.
Given how much time it takes for air safety organizations to approve a new technology aboard a plane cockpit, pilot-less civil airplanes are not for tomorrow. Keep this story in the freezer for another 20 years.
Slashdot, fix the reply notifications... You won't get away with it...
Automate the passengers. Problem solved. Forever.
Pilots are, in the overwhelming majority of cases, a good thing. They are good at dealing with the unexpected: while a computer might have better speed of reaction etc., I think it will be a long time before a computer could do something like the Hudson River landing - it just wouldn't have the other skills to formulate that plan.
To account for the very rare incidents where a pilot tries to suicide, planes could have a lock-out. New planes are generally fly-by-wire and have good autopilots. It's possible to create autopilots that can land a plane. So it would be possible to have a system located outside the flight deck that could lock out the pilot on the flight deck, engage the autopilot to fly directly to the nearest airport and land - and blat out a MAYDAY signal so that air traffic control get everything out of its way. The system would require activation by 2 members of the crew: one pilot / co-pilot and one member of the stewarding team. Once engaged there would be nothing that can be done to override it until the plane was landed, and there would be no other action it could take than "divert to nearest airport and land". This offers the possibility of defeating a lone rogue pilot who locks the other one out of the cockpit. It could not be used by other flight crew to crash the plane as there would only be one action it would allow and that would be automated. Two flight crew could conspire to cause a nuisance diversion, but that would be less fatal than instigating a crash. It could mitigate hijacking attempts as even if the hijackers took control of the flight deck they could be locked out.
Autopilots aren't a panacea and I don't think pilots will disappear any time soon, but for certain limited scenarios where you know there is clear and present danger to the aircraft, handing over control to the autopilot might be the safest option.
The pilot is one of the cheapest parts of the plane. A 747 costs $20,000 per hour in fuel alone. A well heeled pilot makes $80 an hour (and many make far less, $20 an hour is not uncommon for crap airlines moving cargo around). Saving $160 an hour seems decent, but I can't imagine an airline putting more than a minute's thought into it when they'd have to sink a pile of money into the aircraft and some remote piloting system to avoid it.
As a Sys Admin who works on many remote servers and systems, I will say that there is nothing like being on site. I think that holds true in this scenario as well.
"What the American public doesn't know is what makes them the American public." -Ray Zalinsky (Tommy Boy)
We don't even have automated trains and they run on freaking tracks
We absolutely do have automated trains in service today. Automated trains are actually fairly easy to do.
honestly it's not a huge technical issue to get rid of the pilot
Not even remotely true. It's achievable but hardly trivial.
at any moment there are many UAVs flying around autonomously
Most are not generally autonomous though there are some. Most are remotely piloted
First, a remote pilot can go amok just as easy than an on board pilot. Actually, he or she can do this even easier. So that would not help to solve the crazy people crashing planes in mountains problem (even though this is a very seldom cause). Therefore, only complete automation would be sufficient to heel this problem. Second, robot planes rely on instrumentation just like normal pilots. However, they only know those things about the world programmers have introduced into their programming. They cannot think outside the box. Therefore, they might come to the wrong conclusions for certain situations. We must therefore evaluate how often that is the case in comparison to human pilot error. Third, for people to use a completely automated plane, they must be convinced that such planes are safer than human controlled planes. This can be achieved by using automated planes for cargo first.
All in all, I assume that there is not much to be gained by automating flight. So this automation idea will most likely be dropped again as before.
Terrorists will use this remote manageability to kill more people by hacking into the aircrafts. Most of the these systems will be connected to the internet and it will only be a matter of time when crashing an aircraft will be possible by taking control of services/applications running on an IP address.
Solving the problem in the wrong way!!!!!!!!!!!!!!!!
There are two types of flying: instrument flight rules, and visual flight rules.
IFR is used at night, when the weather restricts visibility to under a certain amount (thick cloud cover can do this, no precipitation needed), on flights long enough that you can't guarantee VFR conditions at your destination, and just whenever you feel like it.
VFR is generally only used for beginning pilots or quick flights. It's sometimes seen as a relic of earlier times. Sure, you get taught how to fly this way, get taught some basic dead reckoning techniques, but nobody really flies this way, most of the time.
But instruments fail. Autopilots fail. Engines fail. When everything fails, you want someone in the cockpit who can look out the window, navigate by landmarks, and if necessary put the plane down on the straight sections of highway Eisenhower built to accommodate bombers returning from the Soviet Union.
When things go wrong, you want something smart and adaptable in control. There are procedures for damn near everything in aviation, but there's still things you can't pre-plan for. Until we get a general strong AI working, the only thing smart and adaptable enough is a human.
Now, that doesn't mean we can't have fully computer-controlled aircraft. It just means there shouldn't be people on board those computer-controlled planes. Drones are fine - even if it's a cargo-laden drone version of a 747, the loss of life it can accidentally cause is miniscule, compared to even a small passenger plane.
Not going to ride on such a thing.
Consider the comment at the end of the summary: "If you put more technology in the cockpit, you have more technology that can fail."
We just had a human brain fail -- in the computer sense -- in Europe. I sometimes wonder whether we have too much tech in the cockpit right now. A computer that stays focused on the act of flying as fewer mental "moving parts" than a human brain that has a life outside the cockpit.
usually.
It's the few corner cases that the computer can't handle that require the pilot.
Of course, security will be ironclad so hackers will never be able to seize control or randomize the operation, right? Just like e-voting.
Labor is about 35% of the operating expense of an airline. Fuel is a distant second. Pilots are the some of the pricier elements of that labor pool, and are often unionized and do strike on occasion. Bean counters just love the idea of taking them off the spreadsheet through the magic of modern technology.
The airlines would love to see these aircraft running on a subway model - no attendants or pilots, maybe one person on the ground monitoring that the airplane stays on track. Just put a soda machine on each plane, one that takes credit card and charges $10 for an RC Cola (Coca Cola is $15 for 1st class only). Send a cleaner through once a day. No carpeting or upholstery, need to be able to hose it out quickly. Complaints? No one on the plane to listen to them except the other equally disgruntled fellow passengers.
Left MS Windows for Linux Mint and never looked back!
Vote for Bernie in 2016!
Necessary? No. Planes have been technically capable of fully autonomous flight since the 90s. Pilots have remained in the cockpit for two reasons: unions, and consumer trust.
But the bigger question is this: Is there a technological solution to a social problem? The social problem is people wanting to harm others with as little effort as possible. Does remote or automated flight raise or lower that bar? From a physical security standpoint, not much has changed -- you *still* need to prevent physical access to the systems, because with physical access, all bets are off. But additionally, you also now need to provide cybersecurity for all aspects of the flight control, from the systems where automated software is coded, to (possibly) remote control locations, to the planes themselves. And imagine being able to control hundreds of planes at once instead of just one.
In my view, technology is not the answer to questions of air security. It may make sense for other reasons -- cost, optimal flight parameters, reliability -- and those reasons may well outweigh security, but it doesn't remove the need to trust. Whether you're trusting hardware engineers, software developers, technicians installing the hardware or firmware, or pilots. There may be ways to raise that level of trust, and that's worth looking at, but simply moving the goalpost doesn't remove trust from the equation.
https://www.eff.org/https-everywhere
It almost seems as if Hanlon's Razor is actively working to reassert itself in the commercial airline space.
Give me a break commercial Airline industry can't even be bothered to integrity protect basic routine communications with aircraft and now people are chattering about remote controls.
Airlines are already complaining about a shortage of pilots today.
Shortage of pilots, hah, that's a good one. Do you have any idea how many people are trying to become professional pilots? There are so many it drives the wages down to ridiculously low levels compared to the cost of training and certification.
Apparently Marshall Brain knows what he's talking about after all. http://marshallbrain.com/robot...
I think a more practical approach is to don't allow the pilots to crash the plane on a mountain. Humans are still needed in the cockpit, just remember the Hudson River Pilot. What would a computer do in that situation?
Get my e-mail after a captcha test in: http://tinymailt
Welcome ladies and gentlemen to the new age of flight - this is the first fully computer guided autonomous flight and this is your computer captain speaking! All systems have been fully checked out, we are taking off, and expect to land at our destination in 2 hours and 13 minutes. There is nothing that can possibly go wrong, go wrong, go wrong ...
In watching "Air Disasters" and in seeing post mortems of many accidents and near accidents, I have a hard time seeing that any solution will be better than what we have today.
In some cases weather radar is deceptive, with so much of a crap storm reflecting radar that you can't see the even bigger crap storm behind it and end up steering into an even worse storm than you think you are escapsing.
Sometimes you lose all hydraulics and have to feather the throttle to steer the plane.
Sometimes ducks destroy the engines and you have to land in the Hudson.
Sometimes you spend too much time fiddling with the autopilot that you respond badly when the sensors ice up or get jammed with a mud spider nest that you ride a stall all the way to the ground. I see bad sensor readings being a case where either an autopilot or a remote co-pilot will have even worse odds than what we have today.
It all boils down to us having an amazingly safe system today, and being careful not to throw the baby out with the bathwater.
Airlines without on-board co-pilots (or even pilots) may make economic sense but will do nothing to improve safety and may even hinder it significantly. If you want a perfect example of how automation can go wrong look at the loss of Air France Flight 296, the pilot applied throttle to try to climb out of a dangerous situation during a low pass of an airport but the onboard "safety" systems locked the throttle. If programmers couldn't foresee such a simple case where a pilot may want to throttle up quickly near ground level they will almost certainly miss the thousands of other situations where an autopilot/remote pilot system may respond improperly because of lack of sensors, feedback or fear of death.
like Captain Sully did, and the passengers would have all died.
There have been other inflight failures that were successfully resolved only by creative actions of the pilot that only a pilot in the cockpit could resolve, like the Sioux City crash.
And remember the Gimli Glider? Nobody died because the pilot remembered where a runway was that wasn't on the charts.
Possible. But crashing planes through remote piloting? I don't think so. Coming from a technical background, if I were to program a plane to fly itself, I'd restrict what remote pilots can do. As a small example, using artificial intelligence, if a plane was at a certain altitude, a sudden remote command to shut off the engine or accelerate while nose diving would just not work or be available. I would totally love to see planes fly themselves.
Why not have AI that doesn't fly the plane non-stop, but instead detects egregious attempts to down the plane, and overrides them? Or calls this fancy, distant co-pilot to take over?
I still think that well trained people are harder to hack than software, currently. The main issue is that if you manage to hack a person into wrecking a plane, that just works on the one person. If you hack the software for a plane you can hack all of the planes. That's really my only concern about self driving cars. If a car or two here and there freak out and crash it's still safer than people, but if you can somehow manage to freak them all out...
X
I think John Nance's book Skyhook sums up my fears...
http://amzn.com/051513712X
At the (relatively low) burdened wage of $50/hr, you've both increased you costs by 50% and reduced your flight capacity by 25%. Need a co-pilot? You've doubled your flight cost per hour and halved your available payload, Congratulations, you've just increased the cost to fly by a factor of 4.
Not that anyone is flying a 172 for commercial purposes, but $100/hr for four passengers ($25/hr/passenger) vs $200/hr for two passengers ($100/hr/passenger) is a pretty big difference.
Is it just my observation, or are there way too many stupid people in the world?
http://www.askthepilot.com/germanwings-crash/
after saving all those passengers, he had to RETIRE. This is like former Space Shuttle commander "hoot" Gibson, who went on to fly for Southwest Airlines but was then forced to retire at age 60. Under the crony-capitalism hand-in-glove relationship between the ALPA and the FAA, airline pilots in the US (no matter HOW good they were and no matter how much the passengers might prefer their experience and temperment) were forced to retire at 60 (this MAY have recently changed to 65 to align with international ailines) even if that meant their replacements would be inexperienced 26-year-olds, or 50 year old burn-outs who only pass minimum standards, have a history of ground taxi incidents, "iffy" landings, and getting drunk during layovers.
The FAA has a LONG history of mouthing the word "safety" to justify whatever is preferred choice is and then back-filling with an avalanche of so-called evidenciary documentation; for many years they rejected many concepts for digital avionics because their in-house "experts" (many of whom were WWII/Korean War era guys) did not trust computer chips, and even now some of their rules for software in avionics are silly if you really understand how microprocessors and code work down at the binary level ( I've argued with one of their "experts" about what would happen to a jump instruction with a corrupted bit in the destination address - the guy was completely delusional... ). Like all government people, they have their RULES, and they have their mountains of paperwork, and they have a mental list of what they want - and they will use the 1st two to demand and justify the 3rd whether it aligns with reality or not; it's apure illusion to think that they, or any other government bureaucracy "saves lives" - the industry itself saves those lives. It turns that it's bad (lost reputation and a huge financial hit) for airlines to kill their passengers and it's another huge financial hit for them to fly inefficient planes, so they want better planes and good pilots. This means it's good business for aircraft builders to make each generation of planes better (safer and more efficient). A bunch of idiots at desks in D.C. who claim credit for this naturally-occurring trendline just does not mean they ARE responsible.
Computers sometimes fail. So do humans. The best way to not be at the mercy of either is to have both. There's at least one incident on record where malfunctioning sensors told a plane computer that it was 4000 feet higher than it actually was, and it would've happily crashed into the ground during descent if the pilot hadn't looked out the window to say "wtf, that's the ground right there".
At HAL 2001, yes that was 14 years ago, there was a speech with the title "why my space ship will not run Linux" and that's as true today as it was back then: Our current software, from firmware and operating system to applications, is total crap, incredibly shoddy, and half of it is being held together by spit and duct tape. Fact is that while we make progress (and not a little, actually it's quite amazing), we still don't know how to write really good software. We know a bit about how to teach humans to write pretty good software, even though most companies use 10% of that knowledge in real production (mostly because next-quarter focussed managers don't understand the incredibly good ROI on high software quality).
But a lot of that knowledge is about software development processes.
But do we know how to make a non-trivial computer program that is guaranteed to behave correctly? How much software with an EAL5, EAL6 or EAL7 certification do you know? Wait, you can check here. Not very many.
Assorted stuff I do sometimes: Lemuria.org
I've been on planes landing in ridiculously high side-winds that I'd have a really hard time believing an autopilot could ever land safely, or for that matter, a human controlling the plane remotely - for the simple reason, that in these cases one needs to actually 'feel' how the plane reacts, and neither an autopilot, nor a remote joystick-operator can accomplish that. On a sidenote, there is no way in hell I'd trust hundreds on human lives to an autopilot built with technologies that we have today - what we call artificial intelligence, and what we have as machine learning are so far from such a thing, that it's not even funny. Unless we'll have R. Daneel Olivaw piloting that plane, I'm not boarding it :P
I am putting myself to the fullest possible use, which is all I can think that any conscious entity can ever hope to do.
Have they found MH370 yet? If not, maybe a good start would be to make all planes trackable all the time..
Read also: http://en.wikipedia.org/wiki/Qantas_Flight_32 One of most modern aircraft - the A380, with 440 people on board.
Suffered heavy damage thanks to engine blowing up and only just escaped disaster thanks mostly to a very competent crew.
and we will need basic income to cover the people automated out of a job
Airlines are already complaining about a shortage of pilots today. This will take years and people growing up will have the opportunity to pursue something else, like, I don't know, the technology for automating aircraft. And those military drone pilots will have an almost direct path into civilian employment, assuming that they can get over Macho Grande.
Pay me $30k to start on a regularish full time job and I'll ditch IT for piloting immediately for a couple years. Starting pilots work shit schedules for years, for shit regional airlines, and don't get paid to commute the 600 miles to/from work.
We have a pilot shortage the way we have an IT shortage. Skilled people, with tons of experience, wanted for shit conditions and pay are in short supply. How many pilots are making more than starting NBA power forwards? None you say? Then there's no shortage. The free market says if you can't find people for any given wage, then the wage is too low.
in the future will be two pilots. one human second a dog. human will feet the dog and dog will bite hand when human will try to navigate the plane.
We keep hearing on slashdot the men and women are "built differently". OK, so I'll go with that.
We have a suicidal pilot take out a plane every 5-10 years. Almost all pilots are males, and in western countries, white males.
We can drop that rate to once or twice a century by replacing all white male pilots with African-American females.
No pilots is probably going too far, but what about refusing to do stupid shit? A plane has GPS, it knows how high it is. Why would it even allow a pilot to crash into a mountain? Trying to land a plane in an area that is not an airport should require some sort of manual override by either two pilots or pilot plus ground crew. The computer could even give full control to pilot in case of emergencies (Sully).
Remote pilots probably don't feel the same level of anxiety as one aboard a craft that's in trouble...it might also mean that, because the remote pilot isn't in any danger, they might not put enough effort into trying to save the craft.
So long as, if a plane goes down and a "present" (on board) pilot would have died, the remote pilot is shot-on-the-spot, I think that'll work as providing enough incentive for them to make sure the craft doesn't crash.
Just my thoughts.
Give a hand, not a hand-out.
In order for automated pilots to replace humans, they don't need to be perfect, they only need to be better than human pilots, and judging by 90% of air accidents, that wouldn't be too hard. How many times have you heard of a plane crash because the Autopilot malfunctioned? Never. Its only a matter of time and if I was a human pilot I'd be thinking about my career prospects.
but arguably, if an autopilot had been given the hudson as a landing location it could have pulled it off just as well if not better than sully. don't forget, Sully "didn't know if they would make it back to the airport safely". a computer could have calculated exactly which airports were viable landing options given all the input (rather than a hunch). Sullys actions were only right because nobody died, but what if everyone died on board (which could have just as easily happened) and it was subsequently found he could have made it back to an airstrip? Nobody bothered to calculate for certain if he could have made it back, only because nobody died. If they had died then he would have been called a very poor pilot if he could have made it back. While Sullys cool and calm were admirable there was a considerable amount of luck involved in that incident. If every pilot on every plane was a Sully you might have a valid point, but they aren't all that good. An autopilot could consistently and accurately make the right call with the best chance of survival without any guesswork. It would know based on its altitude, winds and distance where it could make it back to.
consider this though, 99% of the technology in a cockpit is around the human interfaces to the 'brain'. dials, switches, screens, knobs, sticks, radios.... all that stuff only exists so that the pilot can control the plane. if you took all that stuff out you would remove a huge amount of technology, weight and cost. then you could replace it with technology that was only there to increase reliability and safety. a control stick does not make a plane safer, so get rid of it. a pilotless plane would have far less 'technology to go wrong' in it than a plane with a pilot. look at military drones, they very rarely fall out of the sky (and they are being shot at all the time) and the military often considers them as a disposable item. if we built a passenger aircraft with no pilot and made the focus of the technology onboard safe and reliable flight, there is no way that a piloted aircraft could ever come close to it in terms of reliability and safety.
In this scenario, a ground controller might operate as a dispatcher, managing a dozen or more flights simultaneously.
Air travel is pretty safe, although there is always room for improvement. But I'll take the flight where the human pilot is aboard the aircraft and subject the same same consequence of their actions as I, a passenger.
What about passenger-less planes? No pilot = no passengers ... or at least not me!
Remote piloting opens up another problem: murderous remote pilots. Now they can survive a crash to cause another!
Self-importance and self-indulgence is the root of ALL evil.
I mean, what could go wrong?
The shepherds did so well protecting the flock that the sheep no longer believed that wolves existed.
Here's what I don't get. You got an advanced plane like the A320 that has all sorts of fail-safes built in to it to prevent the dopey pilot from making a sudden move and crashing the damn thing. But nothing to say "hey - I'm flying at over 400 kts, the left seat is empty, the cockpit door is locked / overriding the keypad, and the dope in the right seat just told my autopilot to descend to 100 feet above sea level when what's in front of me are mountains that are much higher than that and he's increasing my speed too... maybe I should take over and resume level flight well above terrain - seeing how sensors indicate all is well with the aircraft, squawk 7700 on the transponder, start spitting out what's happening on ACARS, open the door and wait until the left seat is occupied and this gets sorted out".
As for flights like AF447 - when things like pitots and pressure sensors stop functioning, the plane should then revert to GPS to give the pilots a referential ground speed (yes I understand the difference between that and air speed... but when your pitots shit the bed... what else do you got?), direction and altitude... along with rate of climb/descent, and basic navigation like a waypoint to the airfield. Because in the middle of the night over the ocean you're flying on instruments. When your instruments go out and all you see is pitch black outside - you're fucked. A $99 Garmin GPS will do this. A free GPS app for your smartphone will do this.