'This Plane Flies Itself. We Went for a Ride' (arstechnica.com) 36
Xwing envisions a fleet of retrofitted Cessna cargo planes flying without a human onboard, performing automated takeoffs and landing while being controlled by pilots on the ground (who might even be juggling a few aircraft at a time). They've logged "dozens" of hours of flight time — though because of FAA guidelines the future "ground" pilot has always been esconced on-board the plane. Long-time Slashdot reader PuceBaboon shares this report from Wired (alternate URL here):
The conditions are not ideal for our landing. A hard wind is blowing over the low hills east of San Francisco, and at just the wrong angle — straight across the runway where we're set to touch down. But as we ease into our final approach, our two-winged shadow clipping the suburban homes below, the veteran pilot sitting beside me makes a gentle suggestion. "I like to do it hands up. Like a roller coaster," he says.
He removes his hands from the wheel of our aircraft, a 27-year-old Cessna Caravan that once shuttled United Nations dignitaries in southern Africa. It's nothing especially fancy, with aspects that feel more go-kart than airliner. The cockpit is filled with manual toggles and analog dials; pulleys connect the pedal directly to the rudder at the tail. But recently, this plane underwent some modifications. As we descend past 500 feet, the 15-knot gusts hitting our side and the pilot's hands still hovering, the wheel and pedals begin to jostle, compensating for the wind with inhuman precision. The descent remains smooth — serene, even, as we touch down.
"It will be very uneventful, almost boring," Maxime Gariel, the chief technology officer of Xwing, had assured me shortly before our fully autonomous takeoff, flight, and landing. "That's what we're aiming for."
He removes his hands from the wheel of our aircraft, a 27-year-old Cessna Caravan that once shuttled United Nations dignitaries in southern Africa. It's nothing especially fancy, with aspects that feel more go-kart than airliner. The cockpit is filled with manual toggles and analog dials; pulleys connect the pedal directly to the rudder at the tail. But recently, this plane underwent some modifications. As we descend past 500 feet, the 15-knot gusts hitting our side and the pilot's hands still hovering, the wheel and pedals begin to jostle, compensating for the wind with inhuman precision. The descent remains smooth — serene, even, as we touch down.
"It will be very uneventful, almost boring," Maxime Gariel, the chief technology officer of Xwing, had assured me shortly before our fully autonomous takeoff, flight, and landing. "That's what we're aiming for."
This won't fly with the FAA (Score:4, Insightful)
Ground to air communication is bad enough. If you think satellites are the answer, UAVs piloted by the military have no lives on board so if the sat link is latent or lossy there's no loss of human life.
Lots of areas lack ground-station coverage. Watch a SpaceX or Ariane flight and see how often there's no ground contact.
Sorry, wishful people who want to drump up investment, this is not how it's done... and the proof of concept doesn't involve "reporters" in the cabin of an obsolete jump aircraft.
Ground control of passenger aircraft will NEVER be a thing in the US (FAA), Europe (EASA), Japan (JAXA), the UK, New Zealand, Australia, etc.
Ehud Gavron
FAA Commercial Helicopter Pilot (which means I know what a Pilot In Command is... and without one, no aircraft is certificated airworthy nor is allowed to fly.)
Re: (Score:3)
Ground control of passenger aircraft will NEVER be a thing in the US (FAA), Europe (EASA), Japan (JAXA), the UK, New Zealand, Australia, etc.
Well we already have drones that do exactly that. I believe border patrol uses them all the time. One can expect that they will be used more frequently and in other situations as time goes on.
The idea of a ground handler assisting an AI in piloting is a good one. This idea could even be applied to current planes where a ground pilot replaces a co-pilot for commercial flights. This would offer a better solution to air security then confiscating screwdrivers and toothpaste from airline passengers. Ins
Re: (Score:2)
passenger...
I guess I read over that part a little too quickly. But the rest of the comment stands.
Re: (Score:2)
I did say "passenger"* and explained why UAVs don't count - the CBP and the military can use those and if they drop it in the desert like in Iraq https://www.nbcnews.com/news/w... [nbcnews.com] or Nevada no lives were lost. Good thing it didn't hit a house or a car or a wedding party of 37 now-dead people https://blog.witness.org/2014/... [witness.org].
The point I was making isn't just that this is technically unsound (no redundancy in the systems) but that regulations require SOMEONE to certify the aircraft not just before the flight
Re: (Score:3)
The US government flies drones over cities, and this story is about *cargo* aircraft, which is stated right in the summary.
You also seem to be unaware that the FAA, as well as any other government's regulatory authority, can change their regulations.
Re: (Score:2)
no pilot.
no problem.
i am leaving the plane.
the rule i am applying is.
every plane i walk away from is a safe landing
Re: (Score:2)
Try reading the first line of the summary before posting: This won't affect you unless you're in the habit of mailing yourself to your destinations.
Re: (Score:2)
I have no problem with robots out and about as long as the owner, the engineers and the coders, all die when the robot kills. You might think that is harsh but they assure us that they as safe and as such they of course have nothing to worry about, they won't be executed when their robots kill because they are just as safe as they claim. This would include hacking because they claim the software is secure.
So only claim you robot is safe is public, if you are willing to accept all penalties for any act the
Re: (Score:2)
Insisting that robotic aircraft be perfect is silly.
The proper criterion is that they be at least as safe as human-piloted aircraft.
Since 80% of fatal aircraft accidents are caused by human error, that should not be hard to achieve.
Yep, easy to do - but not worth the risk (Score:3)
Ardupilot has been flying planes with no pilot since 2009. There is nothing about Ardupilot, iNav, PX4, Cleanflight, etc that limits the size of plane it can control. They all run quite well on sub-$100 hardware. There is just no damn reason to take that risk with people aboard, or with a $200,000 plane flying over people, for that matter. A pilot doesn't cost that much, considering the risks.
Given that you HAVE a pilot, as in this case, there is simply no point in putting an unreliable long-distance radi
Re: Yep, easy to do - but not worth the risk (Score:3)
It's more efficient and likely safer to have pilots rotating out on shift jobs in one location. Instead of the pilot who didn't get much sleep because of a delay and some hotel mixup, you get the fresh guy who switches in every four hours and who comes in every day at 7am.
Re: (Score:2)
There is just no damn reason to take that risk with people aboard.
The 9th word in the summary is 'cargo'...
Re: (Score:2)
You can count the words until you get to "or with a $200,000 plane flying over people".
Re:This won't fly with the FAA (Score:5, Insightful)
Ground control of passenger aircraft will NEVER be a thing
From TFS:
Xwing envisions a fleet of retrofitted Cessna cargo planes flying without a human onboard
So, no humans on board. Check. Now if they can operate out of remote cargo airstrips dedicated to autonomous aircraft, even better.
Re: (Score:2)
So, no humans on board.
But plenty of humans on the ground beneath the flight path, some of which will die if the plane crashes.
Re: This won't fly with the FAA (Score:3)
Given the number of low earth orbit low latency satellite internet systems being launched, Iâ(TM)d say bollocks to your entire point.
Re: (Score:2)
Ground to air communication is bad enough. If you think satellites are the answer, UAVs piloted by the military have no lives on board so if the sat link is latent or lossy there's no loss of human life.
Lots of areas lack ground-station coverage. Watch a SpaceX or Ariane flight and see how often there's no ground contact.
Sorry, wishful people who want to drump up investment, this is not how it's done... and the proof of concept doesn't involve "reporters" in the cabin of an obsolete jump aircraft.
Ground control of passenger aircraft will NEVER be a thing in the US (FAA), Europe (EASA), Japan (JAXA), the UK, New Zealand, Australia, etc.
Ehud Gavron FAA Commercial Helicopter Pilot (which means I know what a Pilot In Command is... and without one, no aircraft is certificated airworthy nor is allowed to fly.)
And what if the end goal here is to simply save on payroll by removing the (now obsolete) co-pilot?
Let's say everything else stays the same to alleviate your concerns. Primary navigation can be provided autonomously, while backup navigation can be accomplished by one trained aviator instead of two. You still retain redundancy, while saving a lot of money on payroll.
Autonomous solutions are the same no matter what industry it's applied to. The very act and goal of applying it is to remove human error, whi
Re: (Score:3)
Just downloaded and read FAA FAR 91.
It does state there *MUST* be a "Pilot in Command", but I can't find any section that demands the PIC *MUST* be physically on board the aircraft...
I smell "wriggle room" for clever lawyers in that wording...
Re: (Score:1)
See https://www.skybrary.aero/inde... [skybrary.aero] for RPIC (Remote Pilot in Command) vs PIC (Pilot in Command).
It makes reference to ICAO Circulars and FARs part 107.
107.3 says if 107.3 conflicts with 1.1 then 107.3 controls. 1.1 defines PIC. 107.3 defines RPIC
107.15 says prior to flight the RPIC has to physically check the UAS to ensure its airworthiness. That works well for a military launch from a military base (or CBP). Not such more a "remote ARTCC" with "operators" (RPICs in all but name) controlling aircraft
Re: (Score:2)
Ground to air communication is bad enough. If you think satellites are the answer, UAVs piloted by the military have no lives on board so if the sat link is latent or lossy there's no loss of human life.
Lots of areas lack ground-station coverage. Watch a SpaceX or Ariane flight and see how often there's no ground contact.
Sorry, wishful people who want to drump up investment, this is not how it's done... and the proof of concept doesn't involve "reporters" in the cabin of an obsolete jump aircraft.
Ground control of passenger aircraft will NEVER be a thing in the US (FAA), Europe (EASA), Japan (JAXA), the UK, New Zealand, Australia, etc.
Ehud Gavron FAA Commercial Helicopter Pilot (which means I know what a Pilot In Command is... and without one, no aircraft is certificated airworthy nor is allowed to fly.)
What the hell are you talking about? It's right there in the very first sentence of TFS:
Xwing envisions a fleet of retrofitted Cessna cargo planes flying without a human onboard...
The only reason the reports went for a ride-along was because there was a human pilot ready to take over if something went wrong during testing.
Re: (Score:2)
what they use for positioning though? the ground pilots assisting is fine and dandy as long as you have your own network with failsafe connections AND you have as many ground pilots ready to go as you have planes in the air, which makes it as expensive as having pilots.
because say, what if usa decides they want to disable gps for a while for a military operation or whatever and all of the planes need help coming down at the same time.
Re: (Score:1)
> what they use for positioning though?
ADS-B out where required and configured, GPS where it isn't. [ADS gets its data from GPS but adds altitude and direction]
> the ground pilots assisting is fine and dandy as long as you have your own network with failsafe connections AND you have as many ground pilots ready to go as you have planes in the air, which makes it as expensive as having pilots.
Very true. If you've ever flown a "drone" (UAV) you know what any latency will cause - a problem. So having t
Re:This won't fly with the FAA (Score:4, Insightful)
Following that... let me contradict you... or maybe challenge you. I won't know for sure until I've finished writing.
I think for computer control of an aircraft, we'll have more than enough processing ability on the plane in the future to be able to handle as many cases as possible. We don't need super computers for this and frankly the need for a human pilot is temporary. Through many test flights (and I mean tens or hundreds of thousands of hours) with a certified pilot on board, the updated flight control systems will evolve to have an equal or better flight record than human pilots in all conditions over time. This is not a guess... this is a matter of effort and time.
You're thinking very short term... like 5-10 years and you're making blanket statements about how "NEVER" is a reality. Even today, human pilots are far less than perfect. And frankly, I believe passenger aircraft pilots will be far easier to replace with computers than bus drivers. And each day I pass by buses that lack a driver but have an emergency operator in place. I suspect we're still a minimum of 5 years away from one of these buses operating without a human in place. In Japan, trains often already operate without a human.
We start with a disposable Cessna as a proof of concept. We will see investments... eventually people will be offered plane tickets at a discount (or maybe at an additional cost for the novelty) to fly on passenger aircraft where pilots are on board, but only working as emergency backup.
The benefits of all this is that even if we never get rid of the pilots altogether, the technology to assist the pilots will improve by multiple orders of magnitude. The sensors will improve, feedback mechanisms will improve, etc...
Then consider that eventually when we get rid of the pilots, there will still be cabin crew on the plane. So rather than training a person to be nothing more than a full time driver. If the sensors are so much improved and the computers are pretty much able to fully assist and walk a human pilot through, then cabin crew will be trained on simulators to assist the auto-pilot in case of emergency. There will be no need for a fully certified pilot on board. Just one that may or may not be accessible by radio.
You make a truly valid point about communications... but you're also suggesting the plane would be mostly fly-by-wire. So far as I know, the most dangerous part of flying is takeoff and landing and these would be the most important times to have contact. If the sat-link (or high powered local radio from the airport) is negatively effected, a plane should be able to safely navigate itself to a higher altitude and circle until communications can be reestablished or even redirect itself to another airport if needed. There should never be any possibly scenario where a plane could not alter its course to reach a radio signal eventually without risking passengers.
UAVs piloted by the military have no humans on board because it would just be stupid. A military UAV can operate without a human which makes it far smaller, lacking life support or even cabin pressure. What would be the point of a military UAV with a human on board?
Additionally, military aircraft with humans on board are operated by other humans because the military loves finding work for people. During peace-time you have hundreds of thousands of people who you need to keep moving, so even if a C130 can operate with one person on board, they'll send 5 as it's something to do.
As for SpaceX... sure... they lose contact... and the pilots don't really do much when it does. The rocket still flies on a predesignated pattern. Additionally, when it loses contact, there are times where human intervention is useless... such as exit and reentry. When in space, it runs entirely on autopilot. Of course the astronauts could in theory control things, but you don't exactly have to
Re: (Score:1)
Ok, I marked this day on the calendar. I didn't buy MSFT when it was low, and I didn't buy APPL when it was low, and I'll make other mistakes -- for sure.
However, there are differences in the examples listed vs current commercial flight. I say current -- because regulations may change -- but not in my lifetime. Regulators are too averse to even more the terrorism "threat color indicator" to green, because "What if, tomorrow... terrorists..." and they won't allow aircraft in the NAS without a pilot.
Japane
What's not to like (Score:3)
A pilot who doesn't care if the plane crashes, because he flies it from the airport bar, is just what the doctor ordered.
Re: What's not to like (Score:2)
Also ... juggling *several* planes? ... Are they even listening to themselves when they say these things?
Re: What's not to like (Score:4, Informative)
Re: (Score:1)
Re: (Score:2)
So you want to sell this to the aviophobia people by saying:
This way we can save thousands of pilots and we can let fly the planes even from top-pilots in Egypt and Saudi-Arabia.
Isn't it great?
Is there some shortage of qualified pilots? (Score:1)
Re: Is there some shortage of qualified pilots? (Score:2)
Of course it's about profit, dummy. ... *bites down on large fatcat cigar* ;)
Making a number grow bigger in some database, that is absolutely meaningless for the advancement of humanity or even anyone's happiness, is absolutely worth as many deaths as necessary
Re: (Score:1)
Even more insane than self-driving cars and smartp (Score:1)
For the types who do not only think it is "stupid" to think for yourself, and make your own choices, so you're an actual person with an actual own life, ... but want it "Now with even more risk of death by the slightest tech glitch." ;)
We already got autopilots and landing computers, yes. But the key difference there is, that you are always in control, and can always go as close to the bare metal as you want, and it is merely a very
high-level tool.
Instead of being basically the massive projectile version of
just don't fly on windows update day! (Score:2)
just don't fly on windows update day!
Rules can be changed (Score:2)
Naysayers are always amusing. "This is impossible, the FAA won't allow it."
The FAA is a rule-based organization. Change the rules and the FAA will be all "sure, go for it."
Re: (Score:2)
It's also just a slightly higher lift capacity version of military drones which are already accepted, and without the risk of exploding in the middle of a wedding.