Moderators: jsumali2, richierich, ua900, PanAm_DC10, hOMSaR

 
Raventech
Posts: 208
Joined: Thu Aug 30, 2012 10:25 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 4:35 am

While I completely agree that we have the technology today to create an aircraft that could fly from gate to gate without human intervention on board, here are a few points to consider

1 - Reliability
Can you design it so that the system will have reliability equal or more likely more strict than the highest ETOPS rated aircraft. I highly doubt that such a system is viable today with that level of reliability at a price anyone is willing to pay.

2 - Situational Awareness
As long as nothing that hasn't happened before then it can be accounted for in the system design and programming easily. The issue resides in how will it handle a situation that it has never seen before. We would need an AI type program that could do analysis and problem solving on the fly in order to get that aircraft on the ground safely. While it is possible to that such AI could be in the near future I have to bring up is it at a cost that any airline or aircraft manufacturer will pay.

3 - Maintenance
Again, in order for regulators and the public to sign off on such a aircraft there will have to be on an even more stringent ETOPS like program in place for the aircraft and such program will apply to ALL aircraft from the ULH aircraft that are currently already on such a system to a 19-seater that would otherwise would not need it. Again is this program still going to be cheaper than paying pilots to sit up front

and probably the biggest hurdle needed
4 - SECURITY
As of now and probably for the next couple decades the only way that someone could guarantee a computer could not be hacked is if it is air gaped (not connected to the internet or any kind of external network). That is going to be impossible for an aircraft that would need to communicate with ATC and Airline Ops. The only way you may have a shot of locking it down is if you set up the infrastructure for the aircraft to have their own "net" that is isolated at every single possible point. I really don't see anyone interested in doing this just for the pure PITA that it create along with again cost of having separate computers isolated from the airlines network that are aircraft facing (as opposed to net facing). Even then it still will be extremely hard to create locked down software on a locked down network and keep it so.


What I do see happening though would be that systems become reliable enough that we could have single pilot ops where the aircraft can land in the event of captain becoming incapacitated but a human to take over in the event of malfunction. Once we are at that point then we have reached a reliability point where we can then work on autonomous flight and actually have it in reachable . That said though it is probably a few decades off though before single pilot ops and all the above would still need to be answered.
 
User avatar
Boeing778X
Posts: 3268
Joined: Sun Nov 17, 2013 7:55 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 5:59 am

Quoting scbriml (Reply 27):
And of course, human pilots never fail or malfunction? Oh no, never happened before.

For what it's worth, I would rather have a "malfunctioning" pilot over a malfunctioning computer.
 
bueb0g
Posts: 656
Joined: Fri Jul 02, 2010 5:57 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 7:56 am

Interesting how the pro-pilotless airliner lobby are resolutely not responding to the fundamental arguments, or to posters like Pihero, rather ignoring them and hoping that means their arguments don't have to be considered. Again, I'm sure it will happen, but some of the timescales posted here are utterly ridiculous. Planemaker has even previously gone on record to say that the A320NEO and the 737MAX will be single-pilot certified between 2020 and 2025! Well, I'm sure quite a few of us will be happy to make a bet on that not being the case...

Quoting lightsaber (Reply 99):
Most of the anti-automation posts just do not understand where the technology is today. I understand the passion for piloting, but this is the Luddite argument.

No it really isn't, and you'd to well not to wilfully misrepresent the other side's argument.

Quoting lightsaber (Reply 99):
But I know of prototypes where the must have a pilot due to regulations, but every flying issue they've had was the pilot, not the autonomous system. They finally had to tell they pilot if they touched the controls, he had to file an incident report and defend it to a committee. The committee is 100% test pilots or ex-test pilots. So far, every pilot gets told 'let the computer fly.'

Which prototype?

Quoting lightsaber (Reply 99):
Again, I am in awe of the new low power i7 CPU from Intel. It is an enabling technology.

No it isn't. Computing power isn't the problem. It's the fundamental limits of computers, i.e they are dumb as rocks.
 
tomlee
Posts: 610
Joined: Sat Aug 21, 2010 9:01 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 9:38 am

Quoting Raventech (Reply 100):

4 - SECURITY
As of now and probably for the next couple decades the only way that someone could guarantee a computer could not be hacked is if it is air gaped (not connected to the internet or any kind of external network). That is going to be impossible for an aircraft that would need to communicate with ATC and Airline Ops. The only way you may have a shot of locking it down is if you set up the infrastructure for the aircraft to have their own "net" that is isolated at every single possible point. I really don't see anyone interested in doing this just for the pure PITA that it create along with again cost of having separate computers isolated from the airlines network that are aircraft facing (as opposed to net facing). Even then it still will be extremely hard to create locked down software on a locked down network and keep it so.

An automated plane could easily have a set of three or more computerized pilots one being a self contained internally redundant controller with a distributed voting logic to determine the course of action talking to the similarly distributed and isolated flight control system which already has a group pool logic. The only interfaces with ATC and airline ops are the standard unencrypted or encrypted radio data links that provide the air plane with various non critical information of which it can be ignored by declaring a mayday/emergency/other situation and saying the plane must land here or do this action and cannot/won't listen to the commands ("I'm sorry Dave, I'm afraid I can't do that" basically probably even with voice as you have to integrate with existing radio links and other piloted planes). A stand alone computer system as you said can be made secure as long as nothing has remote control authority. (no backdoor remote piloting function for example)

Technically speaking the flight deck is already "connected" it just has no practical means of hacking the fly by wire system because the attack surface is basically non-existent which maintains the airgap.

I think there is one your missing however which is going to be an issue even if the technology is cheap/commonly available,

5 - Legal

The biggest hurtle I think to having automated flying is that will the mfg's (and their insurers) accept the liability risks of being completely at fault for any "automated pilot error" indicated crashes (computer bug, hacking, reliability issue, training problems, ...)? If you say had all major airlines flying automated planes that is a lot of risk to take on which is currently taken up by individual companies that trained/employees the pilots and the people themselves. One way to get around that would obviously be a single human pilot who can take over if the computers give up (for Airbus I guess it would be sit and watch law, and then normal law and so on, many automated factories have many staff to sit and watch it technically requires more skills/qualifications than previously just less people get employed in a roundabout way it might be a good way to relieve human pilot stress and increase pay)


I think the most feasible way instead of direct automation is just make flying a plane so totally assisted that to land a plane you merely have to point the plane in the general direction (you could even have no one on the plane and it would still do it) and crashing is made virtually impossible by the authoritative control of the fly by wire system taking into account things like taxing, take off, landing, buildings, fuel load, damage to the plane, terrain, other planes. (Airbus normal law already provides many aerodynamic protections so integrating things like forward terrain avoidance and range based limits wouldn't be too onerous)(And it is basically the same as a totally automated plane just with the human in the loop as a fall back) Imagine in a hijacking situation planes got near buildings/mountain just dodged them and land at the nearest suitable airport. (Not sure if the tech is actually at that point)

To cover the legal bit you still would want a fully trained and qualified pilot. Training and evaluation costs would sky-rocket though as you have to make sure instead of knowing how to fly a plane normally they also have regular tests in real planes and simulators to test various assist failure cases. Key bit is managing system failures would also have to be assisted/automatic so all those switches are suggestions instead of actual commands. (Button by wire basically)
 
Raventech
Posts: 208
Joined: Thu Aug 30, 2012 10:25 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 10:35 am

Quoting tomlee (Reply 103):
Technically speaking the flight deck is already "connected" it just has no practical means of hacking the fly by wire system because the attack surface is basically non-existent which maintains the airgap.

I'm pretty sure they have said similar
http://www.cbsnews.com/news/car-hacked-on-60-minutes/
an aircraft equivalent would be if a hacker was able to spoof a message from Ops.

http://twit.tv/show/security-now/497
The researcher were interviewed and able to go into a lot more detail in the show above on how and why they were able to get from the On-Star like system in the car to the power brakes and other systems.

Is this apples and oranges, probably, but it shows how any connection can lead to a hole for hackers to crawl into. Point I'm trying to make is making something with any kind of public face is very hard to secure and as of now probably cost way to much to even attempt outside of R+D experiments.

Quoting tomlee (Reply 103):
5 - Legal

The biggest hurtle I think to having automated flying is that will the mfg's (and their insurers) accept the liability risks of being completely at fault for any "automated pilot error" indicated crashes (computer bug, hacking, reliability issue, training problems, ...)? If you say had all major airlines flying automated planes that is a lot of risk to take on which is currently taken up by individual companies that trained/employees the pilots and the people themselves. One way to get around that would obviously be a single human pilot who can take over if the computers give up (for Airbus I guess it would be sit and watch law, and then normal law and so on, many automated factories have many staff to sit and watch it technically requires more skills/qualifications than previously just less people get employed in a roundabout way it might be a good way to relieve human pilot stress and increase pay)


I think the most feasible way instead of direct automation is just make flying a plane so totally assisted that to land a plane you merely have to point the plane in the general direction (you could even have no one on the plane and it would still do it) and crashing is made virtually impossible by the authoritative control of the fly by wire system taking into account things like taxing, take off, landing, buildings, fuel load, damage to the plane, terrain, other planes. (Airbus normal law already provides many aerodynamic protections so integrating things like forward terrain avoidance and range based limits wouldn't be too onerous)(And it is basically the same as a totally automated plane just with the human in the loop as a fall back) Imagine in a hijacking situation planes got near buildings/mountain just dodged them and land at the nearest suitable airport. (Not sure if the tech is actually at that point)

To cover the legal bit you still would want a fully trained and qualified pilot. Training and evaluation costs would sky-rocket though as you have to make sure instead of knowing how to fly a plane normally they also have regular tests in real planes and simulators to test various assist failure cases. Key bit is managing system failures would also have to be assisted/automatic so all those switches are suggestions instead of actual commands. (Button by wire basically)

  

That is more or less what I was getting at. My post was mainly the bare minimum that would probably be needed to appease regulators and lawyers.
 
Pihero
Posts: 4318
Joined: Mon Jan 31, 2005 5:11 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 11:38 am

Quoting bueb0g (Reply 102):
No it isn't. Computing power isn't the problem. It's the fundamental limits of computers, i.e they are dumb as rocks.

If intelligence is only referred in terms of computing power, the most intelligent entity on Earth would be a super comp in Asia, and we're , all of us just 8 billions drooling idiots...
...That includes, of course the super-drooling idiots trying to improve computing power even more.

Quoting bueb0g (Reply 102):
Interesting how the pro-pilotless airliner lobby are resolutely not responding to the fundamental arguments,

You'll be ignored, too, I'm afraid.

Ah ! Those arrogants in their ivory tower !   
 
User avatar
lightsaber
Moderator
Posts: 23362
Joined: Wed Jan 19, 2005 10:55 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 11:48 am

Most of the technology we're discussing is at a TRL (technology rediness level) of 6 or better. We're getting there.   

Quoting bueb0g (Reply 102):

Interesting how the pro-pilotless airliner lobby are resolutely not responding to the fundamental arguments, or to posters like Pihero, rather ignoring them and hoping that means their arguments don't have to be considered. Again, I'm sure it will happen, but some of the timescales posted here are utterly ridiculous.

What about the timescales are ridiculous? Look at what has been developed in the last few years. Take UCAS. Who here believed a plane could operate off an aircraft carrier?

And what am I to respond to Pihero? When a discussion is 'arrogance of engineers?' I've heard that a thousand times in my career, always from those who want technology to stop. But technology doesn't stop. The sub-routines are being developed. The current limit is the computing power that can be airborne. Note: Because of the conservatism aircraft must have, a CPU has only a quarter of its processing power usable as there is margin built upon margin for aircraft.



Over the years as this topic was debated on a.net, UAVs have gone to waypoint navigation, logic based avoidance loops, new sensors (e.g., multi-spectral cameras).

Quoting tomlee (Reply 103):
The biggest hurtle I think to having automated flying is that will the mfg's (and their insurers) accept the liability risks of being completely at fault for any "automated pilot error" indicated crashes (computer bug, hacking, reliability issue, training problems, ...)? I

The airlines will have to buy insurance to cover the manufacturers just as they must cover the pilots today. There will be a 'power by the hour' clause in the sales contract to cover such issues.

Quoting bueb0g (Reply 102):
No it isn't. Computing power isn't the problem. It's the fundamental limits of computers, i.e they are dumb as rocks.

For pilotless aircraft, I have been posting in these threads how computing power is a fundamental issue. There hasn't been a better *low power* CPU available to aerospace engineers for over a decade as even laptop CPUs went to really high watt power consumption which doesn't work in an environment where the aircraft must fly after the loss of cabin pressure.

Think about it, today's FBW software is built to operating on a 80186 (yep, no typo on the CPU #), or in the case of a 777, the 80486. Think how little one could do with a 486 computers. Now have available only a quarter of the capability due to conservatism and then add all the redundancy of flight software. Now replace that capability with a multi-core i7 and the quantity of software that may be run goes up by orders of magnitude.

Compare say a Garmin navigator of 5 years ago with a cell phone today where the navigation uses 4X the CPU. But that navigation now constantly calculates alternates so that re-routing is smoother.

Or take Kiva robotics. There robots seem simple, but they are running off very complex differential equations to optimize their operation. The limit on the robot/human interaction is the server CPU power. Think of how much safer those robots are than a traditional warehouse full of conveyor belts or forklifts run by yahoos. (Anyone who has ever spent much time in a warehouse has almost been run over by a forklift operator refusing to follow the rules.)

The software is being developed. It is a question of when to launch.

Quoting tomlee (Reply 103):
The biggest hurtle I think to having automated flying is that will the mfg's (and their insurers) accept the liability risks of being completely at fault for any "automated pilot error" indicated crashes (computer bug, hacking, reliability issue, training problems, ...)?

That is the least of the issues. The airline's insurance would be at fault. If it were an airframer error, they would be sued as they are today. With self driving cars, the initial estimate is a *reduction* in insurance costs of 30% due to the lower likelyhood of an accident.

I've found those against a technology just keep raising the bar for what needs to be done. What is required is slightly lower chances of an accident than a piloted aircraft.

The legal issues might have pilotless aircraft start somewhere other the the US or EU. Cest la vie. Embraer is a big proponent of the technology... This might be like encryption where a short sighted legal framework pushes the development to nations with a more ready legal system.

Quoting tomlee (Reply 103):
A stand alone computer system as you said can be made secure as long as nothing has remote control authority. (no backdoor remote piloting function for example)

Technically speaking the flight deck is already "connected" it just has no practical means of hacking the fly by wire system because the attack surface is basically non-existent which maintains the airgap.

One would keep the same security protocols as with a FBW aircraft. "Hacking" is of major concern to airliners which is why the new UAVs are not remotely pilotable in the traditional sense. One pilots them by sending "general commands" (e.g., go to waypoint X, enter landing pattern at airport YYY). The detailed commands are left to the multiple onboard computers. It is the older UAVs that might be pilot controllable that have so many issues. Thanks to autonomous.

I'm more worried about the phone network being hacked than aircraft. I've watched technicians upload new software builds to an aircraft. It is a tedious process that first requires physical access to the aircraft as the allowed paths for a new software build are very limited. The aircraft must also have the computers set to maintenance mode to even begin to allow the process to start. Even then, they are built to reject any software change unless the process is followed.

Quoting Pihero (Reply 105):
Ah ! Those arrogants in their ivory tower !

How is UCLASS an ivory tower? Or the numerous other projects at Boeing, Lockheed, Northrop, and Airbus. Embraer is developing technology too. This will be like driverless cars. Two years ago, everyone was scared of them. Now? I'm one of hundreds of thousands looking to offload my commute and as I posted before, it looks like the insurance will be 30% less than human driven due to removing the 'randomness' people impose.

What is there to reply? Those who don't want the technology keep raising the bar for what is required. Others of us just watch the technology being developed. Certain enabling technologies have happened.   

Every one of this threads have seen those excited about autonomous flight just dismissed. But the software development is happening.   

Lightsaber
 
FlyPNS1
Posts: 5560
Joined: Wed Dec 01, 1999 7:12 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 12:10 pm

Quoting lightsaber (Reply 106):
Every one of this threads have seen those excited about autonomous flight just dismissed. But the software development is happening.   

But software development is only part of the equation in moving toward pilot-less aircraft. That's what people seem to be missing.

Quoting lightsaber (Reply 106):
Compare say a Garmin navigator of 5 years ago with a cell phone today where the navigation uses 4X the CPU. But that navigation now constantly calculates alternates so that re-routing is smoother.

But even with all that improvement, GPS road navigation is still quite poor. Many of the "fastest" routings are still slow, the time to travel estimates are still way off, their understanding of traffic congestion is poor and understanding of traffic intersections even worse.

Quoting lightsaber (Reply 106):
"Hacking" is of major concern to airliners which is why the new UAVs are not remotely pilotable in the traditional sense. One pilots them by sending "general commands" (e.g., go to waypoint X, enter landing pattern at airport YYY).

The DoD has already has some of their most advanced UAV's hacked. The DoD itself gets hacked on a routine basis.

Quoting lightsaber (Reply 106):
I've watched technicians upload new software builds to an aircraft. It is a tedious process that first requires physical access to the aircraft as the allowed paths for a new software build are very limited.

That's not how hackers would attack an aircraft though. There would be much easier ways than going through a software build.

I certainly believe the technology for pilot-less aircraft is here, however I do believe there are a lot of other considerations that most engineers like to ignore.

Quoting lightsaber (Reply 106):
Over the years as this topic was debated on a.net, UAVs have gone to waypoint navigation, logic based avoidance loops, new sensors (e.g., multi-spectral cameras).

All true, but the safety/reliability record of UAV's remains abysmal. No where near good enough for commercial passenger operations.
 
Pihero
Posts: 4318
Joined: Mon Jan 31, 2005 5:11 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 1:14 pm

Quoting lightsaber (Reply 106):

My objections, one by one, just for starters.
And I see your still into computing power...

Quoting lightsaber (Reply 106):
When a discussion is 'arrogance of engineers?' I've heard that a thousand times in my career, always from those who want technology to stop

Funny that your side of argumenters are not mentioned. Btw, they started it :

Quoting planemaker (Reply 18):
that a surprisingly high number of people don't know where we are at technologically (but think they do) or simply have [b ]closed minds.[/b]
Quoting planemaker (Reply 32):
It is really interesting that almost all "deniers" on here are posting through a rearview mirror and are not aware of what the state of the art is, let alone comprehending what will be coming down the pike in 10 to 15 years.
Quoting planemaker (Reply 40):
so many years (during which you have contributed great content to these threads) so many people still don't get i
Quoting planemaker (Reply 69):
the majority of posters are not knowledgeable about state of the art, let alone being able to comprehend or understand where it will be in 5, 10 ,15, etc. years. As I mentioned earlier, they are invariably looking through a rear-view mirror
Quoting lightsaber (Reply 99):
this is the Luddite argument.

...and you think we could have a discussion ?  

This thread is close to being the meeting of dissatisfied people, re-hashing over and over something that has been known for some seventy years : automatons can be designed for specific tasks, with a degree of self-correcting immediate situations, like weather avoidance.
In 1944 /45 V1s and V2s were crashing into London and the English countryside : Accuracy of a few miles.
In 2001 cruise missiles were destroying bunkers : Accuracy ~1 meter
They can achieve better, now...
So what ?
The 1m accuracy had been reached seventy years ago, with unguided projectiles, crappy rockets or antiquated guns : see how Typhoon pilots destroyed the nazi armour in Normandy.

Back to civil aviation :
Blind landings : any airline cadet has done dozens under hood, and on light aircraft, too ! I did, 45 years ago, just using depression gyros, IAS and a two-needle ILS. All airline pilots I know have done blind landings - zero/zero - in the sim. There is nothing an automaton can do that I can't... except better fly straight and level for hours.
If one looks at the modern instrument panel, what you are talking about is the PFD and the ND on Map Mode : immediate, short-term adjustments of the flight path on a memorized route plus some amendments for weather, traffic and ATC... These are only add-ons to the stored flight plan.
Outside that short-term PFD and that magenta line...bye bye ! there is nobody.
Introduce then an unforeseen scenario ( like the one you lot have never solved ) and the more likely outcome is a fireball at some destination airport.
There is above a small short discussion about flight planning and fuel use.
Outside the US, we agree with dispatch on the departure fuel quantity, filling all the squares : burn-off, hold, diversion... etc... Once airborne, for the flight commander, all the squares are erased : he has fuel in the tanks he is absolutely free to use as he pleases, therefore opening a vast array of choices of different management strategies... his/her responsibilities.

When planemaker foretells of a *boring* future, thinking that we just slavishly follow stored plan, he totally ignores the fact that many of us still play the game of "screw George", changing cruising levels, programmed CIs, speed schedules, route changes... Most of the times we are successful : George is not the ultimate authority... and by far.
Humans have a strange way of using their tools in ways that the conceptors have never thought of.
 
bond007
Posts: 4428
Joined: Mon Mar 14, 2005 2:07 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 2:10 pm

Quoting bueb0g (Reply 102):
Interesting how the pro-pilotless airliner lobby are resolutely not responding to the fundamental arguments

I really don't waste my time on factual explanations to somebody who says the following and obviously has no respect for anyone with a differing view (and WE are the arrogant ones LOL):

Quoting Pihero (Reply 96):
The same arrogance from engineers ( ?????????????) who claim to have ALL the solutions but have not the first clue on air transport environment
Quoting Pihero (Reply 105):
Ah ! Those arrogants in their ivory tower !

...and anyone who describes computers as...

Quoting bueb0g (Reply 102):
i.e they are dumb as rocks.

Obviously shouldn't be commenting on anything of a technical nature.

For the record, as has already been mentioned, some of us have been in your same "air transport environment" for all of our working lives, and have thousands of hours as pilots ..as well as perhaps being educated engineers and scientists.

Typically, when folks start running out of valid arguments, they resort to ad hominem attacks on qualifications and education ...instead of simply providing opposing views or facts.

Quoting Pihero (Reply 108):
...and you think we could have a discussion ?

Ummm....he mentioned being not knowledgeable and not comprehending. If you think that is similar to saying we are all arrogant, then so be it. I have no interest in continuing with this.

Jimbo

[Edited 2015-04-02 07:13:51]
 
Pihero
Posts: 4318
Joined: Mon Jan 31, 2005 5:11 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 2:37 pm

Quoting bond007 (Reply 109):
.he mentioned being not knowledgeable and not comprehending.

If so, where is the point of this thread ?
Most of the *engineers* who foretell the demise of the pilots have absolutely no clue on what managing a flight is about.
Computers ? Certainly not the answer.
AIs ? We haven't begun to touch the aspects involved in flying an airplane, bar the short-term piloting.
See my problem... The very fact that it is ignored / unanswered... points to a real inability to envisage a way around the amount of lateral thinking involved in flying.
...and it requires a rather simple solution which doesn't even demand specialized knowledge of a particular aircraft systems... What happens when it is required ?

Quoting bond007 (Reply 109):
some of us have been in your same "air transport environment" for all of our working lives, and have thousands of hours as pilots

If it were the case,
1/- My problem would have been already solved. It has been around for four years already.
2/- An IT solution would have been revealed. Has it ?

That reminds me of a huge discussion years ago about an automated device to bring back VFR pilots lost in instrument conditions : The solution was...
1/- to shut the engine down   
2/- to then glide the aircraft to a dead-stick approach and landing to an airfield (unequipped with any navaid) in the vicinity.

Some ( they are still on this thread ) saluted an extraordinary step to aviation safety !
The cost of the device was similar to what one would pay for an instrument rating at one of the best reputed flight schools.

Quoting bond007 (Reply 109):
I have no interest in continuing with this.

Welcome to the club.
 
fastmover
Posts: 419
Joined: Mon Mar 30, 2015 5:37 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 6:43 pm

I think the big problem here is Engineers vs Pilots

Pilots by their very nature do not trust anything, that goes back to your very first flight, check everything don't take anyone's word that something is done, question atc, question mx and so on. You are in command if something goes wrong its your fault so check check and check.
Yes you have to have an amount of trust in your aircraft but we have a big red button to turn the autopilot off for a reason.
I fly the 190 and I can tell you for a fact if its a windy day most guys turn the autothrottles off since they lag and have a hard time keeping up with gusts, they go full power just to go idle in the next second, meanwhile that "pilot" anticipates the gusts far better than that computer can, you actually fly by the seat of your pants. I don't think anyone has said to the engineers it will never happen I just feel most of us who deal with current automation and the airline system believe they have a long way to go. That is all I have said, but a note of caution to our engineer friends this is airliners.net full of planes piloted by people so you most likely won't get a warm reception when you talk about taking our jobs away in an almost giddy tone while ignoring people that work in the system and know just what it takes to make it work. One day it will happen no doubt its how long, that is the question.
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 7:40 pm

Quoting lightsaber (Reply 99):
Most of the anti-automation posts just do not understand where the technology is today. I understand the passion for piloting, but this is the Luddite argument.

I think you are being generous here.   Luddites recognized and comprehended the technology that replaced them. The deniers on here are more like flat earthers.

Quoting lightsaber (Reply 99):
Again, I am in awe of the new low power i7 CPU from Intel. It is an enabling technology.

And Intel already has the next generation in the pipeline for distribution later this year.  
'

Quoting bond007 (Reply 109):
For the record, as has already been mentioned, some of us have been in your same "air transport environment" for all of our working lives, and have thousands of hours as pilots ..as well as perhaps being educated engineers and scientists.

Typically, when folks start running out of valid arguments, they resort to ad hominem attacks on qualifications and education ...instead of simply providing opposing views or facts.

In some cases, even before our working lives!  
'

Quoting fastmover (Reply 111):
I think the big problem here is Engineers vs Pilots

It isn't pilots vs engineers... it is pilots vs pilots (some of whom are also engineers).  
 
WPIAeroGuy
Posts: 324
Joined: Tue Aug 21, 2007 11:52 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 8:09 pm

I remember one of these threads several years ago where a lot of the same points were brought up. I am an engineer in the commercial aviation industry, a pilot, and I believe we will see pilotless commercial aircraft - at least on the personal level, within the next 20-30 years.

Quoting Pihero (Reply 96):
Four years ago, I just posed a very simple problem of flight management and only pilots and one brilliant engineer from Germany solved it. The proponents of the pilotless airliner didn't even begin to grasp the question.
Two years later, I did ask the very same question... and guess what ? we have to wait until our world would be unrecognizable to the progress they are bringing to us.


Now I'm curious - what was the question?
 
Pihero
Posts: 4318
Joined: Mon Jan 31, 2005 5:11 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 8:30 pm

Garbage as usual.
Where is the engineering airline pilot amidst your lot ?
... and it's still only about computing power.
Funny you can't even realize that with just human feelings and a few mental calculations, pilots have flown faster than to-days' airplanes full of automatics ( see 707s / 727s / DC-8s / 880s/ 990s / Caravelles / Tridents... to name just a few.
Fuzzy logic, after fourty years of research has still not be brought into flying crafts.
In my career, I have had quite a few brush-ups with engineers to know that the arrogance is certainly not on our side.
Among a lot of experiences : " Don't disconnect the EDP if you lose a generator.".. I did and prevented a fire.
Procedure was quickly changed later by the OEM after my report.
An experience with a nicely designed windscreen heater. A two-position switch : ON and AUTO . Electrical arcing after a birdstrike, bubbling glass, melting screen in a big way. Descent at minimum speed and only 5mm thickness left on the melted area.... The station engineer started to tell me off for not respecting the procedure - which was to continue... I told him in no uncertain terms that first he needed to use his frakking knife to see how deep the window was and then go and have a masturbating conversation with his boss...
We now have a way to switch the power off to these heaters.
So, to accept that you have everything covered in your design, I'd tell you to go back and read some fairy tales.
Most of the times they failed to bring the plane back was due to some engineering marvel that turned to be a death trap.

History of aviation is full of bad ideas getting into airplanes : outward opening cargo doors with a dodgy locking system, Hydraulic systems not protected by plugs, three-pointer altimeters, shoddy electrical wirings over the pilots'heads, batteries that are less safe than the previous ones and augmented fire risk in an unprotected area...... etc...

I'd be a fool to trust you to design a pilotless paper aircraft.
 
User avatar
boeingrulz
Posts: 521
Joined: Thu Sep 30, 1999 2:55 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 02, 2015 8:43 pm

Quoting czek6 (Reply 42):
I've heard many people say that flying will be automated and that the pilot will be there in case of emergency. But where and how will said pilot get his or her emergency flying experience?

In the future there will be a pilot and a dog in every cockpit. The pilot to monitor the systems and a dog to bite him if he touches anything.
 
sierrakilo44
Posts: 680
Joined: Tue Dec 13, 2011 1:38 am

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 12:55 am

Quoting boeingrulz (Reply 115):
In the future there will be a pilot and a dog in every cockpit. The pilot to monitor the systems and a dog to bite him if he touches anything.

I've heard that joke before. It wasn't funny the first time.
To make it more accurate the pilot will be there to fly the plane, and the dog will hunt down any fool who still drones on about pilotless passenger airliners in our lifetimes.

For all of you who rail on about how aircraft are already pilotless, please read Patrick Smith (Askthepilot.com) and his latest column where he shows you how "automated" aircraft actually are in this day and age, by giving you an example of a relatively simple descent from altitude:
http://www.askthepilot.com/germanwings-crash/

[Edited 2015-04-02 17:56:41]
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 3:44 am

Quoting sierrakilo44 (Reply 117):

No one on here has suggested turning a +30 year old airliner into a pilot-less aircraft.
 
KCmike
Posts: 531
Joined: Mon Jul 19, 2004 10:13 am

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 4:07 am

Quoting fastmover (Reply 41):
I have to ask in a respectful way how many of you guys really fly airliners?I feel that those of us who fly aircraft in the airline world know that there is a long long way to go before this happens.I know its fun to say we don't really need pilots or computers are better at flying planes but until you sit in that seat and really understand what goes on I think many would change their tune.

I sense not many on here.

Everyone seems to still be spooked by the unfortunate accident last week. As sad is it is, there will never be a perfect solution. Guess what? When the first pilotless airplane is hacked and used in a horrible fashion every one will be back to same argument only on the otherside saying how there should be pilots. It will never happen. There have already been examples of drones being hacked and I'm sure it wouldn't be too hard to do to an airliner. There are far too many risks going fully computerized. I know everyone is sad about recent events. As a pilot myself, I personally have lost friends in accidents. We try to learn what we can an improve from them and move forward. Pilotless aircraft is not a step forward in safety.

PS - Do you want a computer at the sticks when ingest birds and have a dual flameout at 1500 agl. Nah me neither
 
tomlee
Posts: 610
Joined: Sat Aug 21, 2010 9:01 am

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 4:31 am

Quoting KCmike (Reply 119):
When the first pilotless airplane is hacked and used in a horrible fashion every one will be back to same argument only on the otherside saying how there should be pilots.

While I agree that there should always be human pilots regardless of automation levels, it is certainly possible to have a isolated computer that flies the plane because no one has to date managed to practically hack the FBW system of a plane which already runs on triplex computers. If the pilotless or more correctly highly assisted flight control system has no backdoor, no remote control, no attack surface then it can't be hacked. (having the control room on the plane makes it so there is no significant control center link to break into)

Quoting KCmike (Reply 119):
It will never happen. There have already been examples of drones being hacked and I'm sure it wouldn't be too hard to do to an airliner.

Drones typically don't have good quality INS so they have to rely on GPS (which can be spoofed although if they just let people use an encrypted band you could resist spoofing attempts almost completely) and don't use vision to guide motion automatically. If a plane was equipped with thermal/vis cameras it could perform isolated navigation on sight and dead reckoning alone.

Quoting KCmike (Reply 119):
There are far too many risks going fully computerized. I know everyone is sad about recent events. As a pilot myself, I personally have lost friends in accidents. We try to learn what we can an improve from them and move forward. Pilotless aircraft is not a step forward in safety.

PS - Do you want a computer at the sticks when ingest birds and have a dual flameout at 1500 agl. Nah me neither

Neural nets (deep learning, basically a crude model of a brain) have the damage accommodation bit going years ago, all you have to do is dynamically update the plant equation (the plane) so that you optimize control.

http://ti.arc.nasa.gov/m/pub-archive/1341h/1341%20%28Nguyen%29.pdf

However there are far too many risks to go fully computerized as dropping pilots entirely is like dropping an entire layer of safety. To mitigate pilots intentionally crashing planes Airbus method of computers first then human would work. (Damage control would also be mostly automated as well)

Fully automation isn't remotely ready at this point for industry wide use but, Automatic ground collision avoidance should be looked into as there are already existing products as that would provide plenty of time for people to respond to malicious control of the flight deck and would be useful in many situations providing a safety net. (Provided the door can be unlocked).

[Edited 2015-04-02 21:40:14]
 
sierrakilo44
Posts: 680
Joined: Tue Dec 13, 2011 1:38 am

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 5:19 am

Quoting planemaker (Reply 118):
No one on here has suggested turning a +30 year old airliner into a pilot-less aircraft.

No, but they use these aircraft as examples about how pilot's today don't really fly aircraft (which is BS)
And like Patrick Smith said in his article, there isn't much difference into how you would manage that descent in a 767 to how you would do it in a 787/350. The advocates of pilotless airliners insist the airliners of today are already fully automated, which is not the case
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 5:50 am

Quoting sierrakilo44 (Reply 121):
The advocates of pilotless airliners insist the airliners of today are already fully automated, which is not the case

There has not been a single poster who has said that. Crossing the Atlantic in a 777 still requires around 17 keystrokes.  
 
spacecadet
Posts: 3584
Joined: Thu Sep 20, 2001 3:36 am

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 6:03 am

Quoting tomlee (Reply 120):
While I agree that there should always be human pilots regardless of automation levels, it is certainly possible to have a isolated computer that flies the plane because no one has to date managed to practically hack the FBW system of a plane which already runs on triplex computers.

"Possible", "probable" and "wise" are three different things more or less independent of each other.

I will postulate that while there may someday be pilotless planes, there will never, ever be pilotless planes that are completely isolated from ground-based human control. (At least not until Skynet takes over.) At the very least, there's going to have to be a mechanism for human overrides during unforeseen emergencies or other scenarios that the on-board computer is not programmed to deal with. We've seen how unintended consequences can rear their ugly head with the design of the post-9/11 cockpit doors over the past month and a completely isolated, autonomous, pilotless airliner is an absolute Mount Everest of unintended consequences just waiting to happen.

If you accept that it will always be *possible* for a human to control a pilotless airliner, then you have to accept that it will always be possible for a human to intentionally crash one. The questions brought up at the start of the thread are moot.

If you *don't* accept that it will always be possible for a human to control a pilotless airliner, then you have to accept that the humans that program this airliner's computers are themselves fallible and one or more of these airliners *will* crash either because those human programmers did not foresee a certain situation in advance, or because poor coding (aka bugs) led to problems that no human was available to overcome. No new FBW system that I know of on any airliner has ever launched without bugs, some of which were quite serious. But the humans on board were usually able to overcome them.

And when one of these airliners crashes, you can forget about the computer trying to avoid populated areas. A GPS map doesn't know an empty field from an apartment building or airline terminal, nor could you ever hope to program that information for every single area around the world.

Things that are possible are not always wise. That doesn't mean they won't happen, just that they probably shouldn't.
 
Max Q
Posts: 9143
Joined: Wed May 09, 2001 12:40 pm

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 6:10 am

Quoting Pihero (Reply 114):
Garbage as usual.
Where is the engineering airline pilot amidst your lot ?
... and it's still only about computing power.
Funny you can't even realize that with just human feelings and a few mental calculations, pilots have flown faster than to-days' airplanes full of automatics ( see 707s / 727s / DC-8s / 880s/ 990s / Caravelles / Tridents... to name just a few.
Fuzzy logic, after fourty years of research has still not be brought into flying crafts.
In my career, I have had quite a few brush-ups with engineers to know that the arrogance is certainly not on our side.
Among a lot of experiences : " Don't disconnect the EDP if you lose a generator.".. I did and prevented a fire.
Procedure was quickly changed later by the OEM after my report.
An experience with a nicely designed windscreen heater. A two-position switch : ON and AUTO . Electrical arcing after a birdstrike, bubbling glass, melting screen in a big way. Descent at minimum speed and only 5mm thickness left on the melted area.... The station engineer started to tell me off for not respecting the procedure - which was to continue... I told him in no uncertain terms that first he needed to use his frakking knife to see how deep the window was and then go and have a masturbating conversation with his boss...
We now have a way to switch the power off to these heaters.
So, to accept that you have everything covered in your design, I'd tell you to go back and read some fairy tales.
Most of the times they failed to bring the plane back was due to some engineering marvel that turned to be a death trap.

History of aviation is full of bad ideas getting into airplanes : outward opening cargo doors with a dodgy locking system, Hydraulic systems not protected by plugs, three-pointer altimeters, shoddy electrical wirings over the pilots'heads, batteries that are less safe than the previous ones and augmented fire risk in an unprotected area...... etc...

I'd be a fool to trust you to design a pilotless paper aircraft.

Well said.




There are simply too many non predictable variations in aviation to allow a 'program' to resolve every one.
The skill, human judgement and experience of a Pilot is simply invaluable and cannot be replaced.



Every engineer and / or 'pie in the sky' nerd that believes pilotless aircraft are possible should be made to
sit in the cockpit of one for a few hundred hours just to see what routinely crops up.


It might open their eyes a bit !
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 6:28 am

Quoting Max Q (Reply 124):

No one is saying that today's aircraft using today's NAS will be pilot-less.
 
frmrCapCadet
Posts: 5220
Joined: Thu May 29, 2008 8:24 pm

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 1:48 pm

Another analogy comes from the sea. The captain is the person who 'doesn't do anything'. He (for a long time, now also shes), is the stoic who steadfastly through boredom and peril knows where the ship is going, what it is or needs to be doing, and ensures that everyone is 'making it so'. Whether managing computer or persons a modern airline pilot is becoming more like a ship captain than a fighter pilot.

Not to say that a pilot does not benefit by knowing the skills of a fighter pilot. I have wondered it a few dozen hours a year in small planes set up to do emergency sorts of maneuvers might not be useful to an airlines piloting skills, and likely greatly enjoyed by them to boot.
 
KCmike
Posts: 531
Joined: Mon Jul 19, 2004 10:13 am

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 2:20 pm

Quoting frmrcapcadet (Reply 126):

On a perfect day flying from A to B is routine. Most training programs/sim sessions are not flying a routine flight. You get put through a series of malfunctions and emergencies. This is why flight decks will never be pilotless. Yes, it's not too tough to get to the destination but it is when things go wrong that you need someone up front that knows how to deal with it. Also you can be guaranteed that the folks upfront will be give every effort and try everything possible thinking outside the box, because its their life on the line too.
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 3:44 pm

Quoting frmrcapcadet (Reply 126):
Another analogy comes from the sea.

You mean Rolls-Royce Marine's design for autonomous ships...  
.

(Notice no bridges in the picture?)

There is even a consortium in Europe, with the EU as a partner, for advancing autonomous ships. It is not just airliners that are going to be autonomous.

Quoting KCmike (Reply 127):
but it is when things go wrong

You mean like Apollo 13 when the engineers on the ground came up with the solution.  

Joking aside, you have to disuss this topic taking into account what the state of the art will be in 20-30 years in all areas.
 
tomlee
Posts: 610
Joined: Sat Aug 21, 2010 9:01 am

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 9:31 pm

Quoting spacecadet (Reply 122):
I will postulate that while there may someday be pilotless planes, there will never, ever be pilotless planes that are completely isolated from ground-based human control. (At least not until Skynet takes over.) At the very least, there's going to have to be a mechanism for human overrides during unforeseen emergencies or other scenarios that the on-board computer is not programmed to deal with. We've seen how unintended consequences can rear their ugly head with the design of the post-9/11 cockpit doors over the past month and a completely isolated, autonomous, pilotless airliner is an absolute Mount Everest of unintended consequences just waiting to happen.

Well automatic trains with no drivers in my city have a drivers console under a locked cabinet so it is generally a good idea to have a manual fallback. If you had a mostly automatic plane you could have a locked room with a pool of keys needed to open it so that if something goes wrong and it needs a few humans to decide that. The train works in the same manner as you have to dispatch a team with the correct keys and the control room disables automatic control for the line.

Quoting spacecadet (Reply 122):
If you accept that it will always be *possible* for a human to control a pilotless airliner, then you have to accept that it will always be possible for a human to intentionally crash one. The questions brought up at the start of the thread are moot.

Possible to control but the automated assists would have priority so that even if a person tried it would take a few minutes to many minutes to take manual control unless the computer handed control over itself. (Hence the damage control automation bit so you can't just disconnect the system with the press of a button) Coupled with a pool unlockable door and you could stop a person from crashing the plane.

Quoting spacecadet (Reply 122):
If you *don't* accept that it will always be possible for a human to control a pilotless airliner, then you have to accept that the humans that program this airliner's computers are themselves fallible and one or more of these airliners *will* crash either because those human programmers did not foresee a certain situation in advance, or because poor coding (aka bugs) led to problems that no human was available to overcome. No new FBW system that I know of on any airliner has ever launched without bugs, some of which were quite serious. But the humans on board were usually able to overcome them.

Which is where the legal liability things crops up and why it is just on that aspect alone nice to have a human layer to serve as a fallback.

Quoting spacecadet (Reply 122):
And when one of these airliners crashes, you can forget about the computer trying to avoid populated areas. A GPS map doesn't know an empty field from an apartment building or airline terminal, nor could you ever hope to program that information for every single area around the world.

You can program that in most cities are almost completely 3d mapped with lidar data from low flying planes and cars scanning the ground areas (don't need the image data so the actual database wouldn't be too massive). In areas lacking the data you can use less fine grained satellite data or just contract one of the big tech companies to scan every major airport (it can be done with planes/helicopters/drones so maybe one day we will have google flight view), just ask Google to help out. Building 3d charts isn't that much harder now than 2d charts.

The issue you did ignore which is a problem is what happens if there are only a set of bad options how does a fully automatic plane choose, you could also encode priorities into the map for things like airport terminals/hospitals/schools/... but the plane would be automatically picking who lives/dies or the 3d chart maker more specifically. (legal issues)

That aside having a non-overridable ground collision avoidance system would greatly help in general. (It would just be the continued development of EGPWS)
 
BoeingGuy
Posts: 6749
Joined: Fri Dec 10, 2010 6:01 pm

Will We Have Pilot-less Aircraft In The Future?

Fri Apr 03, 2015 10:00 pm

Quoting bond007 (Reply 66):
ATC puts total trust in computers every time they tell you to turn right 20 degrees to avoid traffic. Same as when TCAS gives you a RA.

The computer is doing a calculation to give the crew guidance commands to avoid another airplane during an RA. This is not a good example. It's one small alerting system on an airplane doing a fairly simple calculation, not having computers manage and fly every aspect of any airplane from gate to gate.
 
Max Q
Posts: 9143
Joined: Wed May 09, 2001 12:40 pm

Will We Have Pilot-less Aircraft In The Future?

Sat Apr 04, 2015 7:10 am

How about the computer nerds that believe in pilotless aircraft develop a fleet of them and only fly on them for say, a year ?


Then we can compare how safe they are compared with piloted aircraft.



They just might want to have their 'affairs in order' first..
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Sat Apr 04, 2015 7:26 am

Quoting tomlee (Reply 128):
That aside having a non-overridable ground collision avoidance system would greatly help in general. (It would just be the continued development of EGPWS)

Following is some related information:

Improved Ground Collision Avoidance System

NASA-Pioneered Automatic Ground-Collision Avoidance System Operational

Ground Collision Avoidance System ‘Saves’ First F-16 In Syria


And if you are interested, an article on the development of GPWS and the evolution of EGPWS:

GPWS – Innovation That Has Saved Lives


One point which has been raised before is that people on here are not looking at where we will be at in 20-30 years in all areas of the industry (in a few cases their reference point is 40 year old technology).
 
tomlee
Posts: 610
Joined: Sat Aug 21, 2010 9:01 am

Will We Have Pilot-less Aircraft In The Future?

Sun Apr 05, 2015 5:04 am

Quoting planemaker (Reply 131):

Terrain avoidance that has authority to cause corrections even if to just decrease collision energy would be good (as in prohibit flying directly into a mountain but allow a crash landing at low speeds and shallow angles) It makes zero sense to unrecoverable nose dive a plane right into the ground in any normal operating condition. So in normal operation pilots control over the flight would not be impeded but if someone took over the cockpit or went nuts it would ideally stop a crash from occurring or at least make it more than 0% survivable. (Delaying the crash would also help people fighting back for control)

This also would not make the plane pilot-less just ground fearing and I think a plane that doesn't want to slam at speed into the ground would be a big improvement. Given the technology already exists and has been demonstrated to work it is an existing not future technology.
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Sun Apr 05, 2015 7:46 am

Quoting tomlee (Reply 132):
Given the technology already exists and has been demonstrated to work it is an existing not future technology.

I wasn't referring to ground avoidance technology seeing that I linked to the existing NASA/Air Force example.  
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Wed Apr 08, 2015 3:52 am

DARPA is moving right along. From a month ago...

Quote:
ALIAS PREPARES FOR ACCELERATED TAKE-OFF

March 10, 2015

Phase 1 research includes planned ground and flight demonstrations to help mature capabilities faster

In the latest step in a decades-long process through which automation has taken on increasing responsibilities in the cockpit—allowing pilots to focus on flight tasks demanding their unique capabilities—DARPA has awarded three contracts for its Aircrew Labor In-Cockpit Automation System (ALIAS) program. ALIAS envisions a tailorable, drop‐in, removable kit that would enable high levels of automation in existing aircraft and facilitate reduced need for onboard crew. The program intends to leverage the considerable advances that have been made in aircraft automation systems over the past 50 years, as well as the advances that have been made in remotely piloted aircraft technologies, to help shift and refocus pilot workloads, augment mission performance and improve aircraft safety.

DARPA has awarded prime contracts for ALIAS to the following companies:

o Aurora Flight Sciences Corporation
o Lockheed Martin Corporation
o Sikorsky Aircraft Corporation

These performers are working with DARPA’s technical team, which includes experts from the National Aeronautics and Space Administration (NASA), the U.S. Air Force, the U.S. Army and the U.S. Navy.

In Phase 1 of ALIAS, DARPA intends to focus on three critical technology areas:

o Development of minimally invasive interfaces between new automation systems and existing aircraft

o Knowledge acquisition on aircraft operations, to support rapid adaptation of the ALIAS toolkit across different aircraft

o Human-machine interfaces that would enable high-level human supervision instead of requiring pilots’ constant vigilance over lower-level flight maintenance tasks

“Because we want to develop a drop-in system for existing aircraft, we chose performers who could conduct actual ground and flight demonstrations at the start of the program instead of at the end,” said Dan Patt, DARPA program manager. “We’re excited to have a lot of Phase 1 hardware ready to test, which we hope will steepen our learning curve and mature the capability faster. We will also be working closely with members of the crew-operations community, and will be integrating their feedback to help ensure that the move towards greater automation takes best advantage of machine and human skill sets.”
 
User avatar
lightsaber
Moderator
Posts: 23362
Joined: Wed Jan 19, 2005 10:55 pm

Will We Have Pilot-less Aircraft In The Future?

Wed Apr 08, 2015 5:54 am

How to eat an elephant? One bite at a time.
The reason I talk about software as all situations are broken down into software routines. But these routines do not trigger actions per se, they trigger MODES. Unless you're into rather advanced software logic, which few pilots except for test pilots are, you wouldn't have ever been exposed to the difference in programing.

The issue in programming for different modes is each mode is composed of a large number of subroutines. While some subroutines can be shared across all modes (e.g., raising or lowering landing gear), not all software behaves the same.

For example damaged aircraft modes must be incorporated that change the flight laws (basic operation of the aircraft). There also must be multiple navigation modes for the navigation is constantly tested and re-tested and sometimes certain sensors must be ignored (e.g,, test for pitot tube icing).

But the software is able to ADD navigation modes a human pilot is not able to process. e.g., Tri (or more) spectral vision with tri-polar vision (bipolar with a 3rd camera for added accuracy and error checking). With enough processing power, add a 4th or 5th camera.

Today I was attending a presentation on a new bipolar vision system. Stuff I wish I could share in more than the general detail I'll write about now: A real breakthrough in camera accuracy (3-D real time tracking of an impressive number of objects). I'm sure the presenter has a public web page. Find out about it... before one of the big aerospace companies buy them to sell the technology as part of their product lines. This presents a *major* leap forward in aviation safety due to the situational awareness. In particular when combined with other software already commercially available.

How many objects can a pilot or pair of pilots track accurately. e.g., distance to a runway, aircraft in front of them, behind them, and other potential flying objects... What you thought was possible just leapt forward. When we were quoted the price we almost gagged (on how cheap it was)! When they noted how much of their code was from open source video games we smirked as when code is good... use it! I'm under NDA, so I can't say which aircraft, but there are aircraft out there flying with code from open source or licensed video games. Not the full game, just subroutines to do certain flight logic.

Again, everything done by a human pilot is broken down to modes, actions, and the detailed response. It is a challenge of writing all the software, V&V, and having the computational power to run it in flight with all the fail safe redundancy.

Quoting Pihero (Reply 108):
My objections, one by one, just for starters.
And I see your still into computing power...

I've seen the software that should be in commercial aircraft running on demonstration aircraft. The logic allows much more functionality. That functionality is what is needed and the current certified computing power for commercial aircraft is negligible! Same for the UAVs where that information is public.

Look at warehouse robots. The upgrades this year in the server farms allowed added logic which tremendously improved the warehouse flow and efficiency. This allowed the warehouses to directly simulate the differential equations for optimum operation instead of doing simplifications. The same is true of aircraft control logic.

You're objections miss the big picture. They just say 'it can't be done.' Engineers take very complex problems and break them down into tens of thousand (or more) small steps. Trying to answer each bit in detail just would take time and misses the big picture. I'm not saying we're ready today. But there are business jets with server racks flying commercial airspace as demonstrators... for the last few years.

Quoting Pihero (Reply 114):
Where is the engineering airline pilot amidst your lot ?

My last project that I worked related to this had over a dozen TEST pilots, all with commercial pilot certs. Two of the test pilots were ex-commercial pilots.

Quoting FlyPNS1 (Reply 107):
But software development is only part of the equation in moving toward pilot-less aircraft. That's what people seem to be missing.

It is more than development, it is the verification and validation. Those two terms *drive* aviation development. Nothing flies without being verified and validated. V&V determines if a new engine or in flight entertainment system goes onto an aircraft. We engineers talk software as that is where the 'rubber meets the road' for autonomous flying.

I wish I wasn't bound by NDA and could describe how aircraft software goes through V&V. What I can share is *all* new aircraft are tested as software devices. They really aren't treated the 'old school' way anymore. I get a chuckle out of some of the issues airframes have as a proper software V&V program has made all but a few areas of flight test boring. Why boring? The aircraft just behaves as expected now. I had the pleasure of going through a flight test program where all the issues were in areas they skipped doing proper software and made the aircraft old school (e.g., hydraulics instead of electrical). There is a reason the A350 flight testing went so smoothly. They spent the money to simulate the aircraft well before flight and thus there were few surprises and issues to fix.

The 787 had issues as they skipped too much of the software and basic mechanical V&V and *assumed* it would go well and instead had to fix it later when hardware designs were fixed instead of just computer models.

I say this as someone who knows my designs work and just fly.

Quoting FlyPNS1 (Reply 107):
But even with all that improvement, GPS road navigation is still quite poor. Many of the "fastest" routings are still slow, the time to travel estimates are still way off, their understanding of traffic congestion is poor and understanding of traffic intersections even worse.

You really need to try Google maps. Also, flying is never just GPS for a high end UAV. Oh... for a drone it might be pure GPS, but a proper UAV navigates by:
GPS, inertia, pitot tubes (many more than today's commercial aircraft), terrain following (visual and radar), and other 'proprietary tricks.'

Seriously, your comment makes it sound like we're talking drones and not the latest technology. Commercial flights will be highly scrutinized. This is a conservative industry.

Quoting planemaker (Reply 131):
Following is some related information:

And the publicly available information trails the R&D labs by 5 years (or more).

Quoting tomlee (Reply 128):
Which is where the legal liability things crops up and why it is just on that aspect alone nice to have a human layer to serve as a fallback.

Liability ends up at the insurance company. Swiss RE and the other re-insurers will end up deciding this, not personal accountability. Does Kiva robotics take personal liability over every one of their warehouse robots?    The warehouse insurance policy does.

Quoting planemaker (Reply 134):

DARPA is moving right along. From a month ago...

  

It amuses me how we'll have these threads and during the discussion there is always an advancement or a program announced (such as this DARPA) during or right before the discussion that helps show the progress being made. Yet some pretend the technology is struck back in 1980.

Quoting Max Q (Reply 130):
How about the computer nerds that believe in pilotless aircraft develop a fleet of them and only fly on them for say, a year ?


Then we can compare how safe they are compared with piloted aircraft.

Are you familiar with swarm technology? That will prove the concept first. There will be mistakes, but the technology is advancing. There are prototypes flying. Mostly business jets with server farms onboard. They've done well. By the time there is a fleet the decision is done and we'll be watching the fleet conversions.

Oh, if you meant 'computer nerds' to be derogatory, that era passed decades ago. Wake up, too many jobs in the USA are computer related for that to mean what it did say in the 1980s. I wish I was a true 'computer nerd,' they make much more money than I do! Every one I know puts commercial pilot salaries to shame. But I'm a 'mechanical guy' who just likes computers as a tool to make stuff happen.


Lightsaber
 
Max Q
Posts: 9143
Joined: Wed May 09, 2001 12:40 pm

Will We Have Pilot-less Aircraft In The Future?

Wed Apr 08, 2015 6:02 am

Quoting lightsaber (Reply 135):
Are you familiar with swarm technology? That will prove the concept first. There will be mistakes, but the technology is advancing. There are prototypes flying. Mostly business jets with server farms onboard. They've done well. By the time there is a fleet the decision is done and we'll be watching the fleet conversions.

Oh, if you meant 'computer nerds' to be derogatory, that era passed decades ago. Wake up, too many jobs in the USA are computer related for that to mean what it did say in the 1980s. I wish I was a true 'computer nerd,' they make much more money than I do! Every one I know puts commercial pilot salaries to shame. But I'm a 'mechanical guy' who just likes computers as a tool to make stuff happen.

How about just meeting my challenge ?



If you have so much faith in Pilotless Aircraft, design and build one which you and all others so convinced of their safety exclusively travel on for a year.


We can all see how safe they are then.


Proof is in the pudding !
 
User avatar
lightsaber
Moderator
Posts: 23362
Joined: Wed Jan 19, 2005 10:55 pm

Will We Have Pilot-less Aircraft In The Future?

Wed Apr 08, 2015 6:09 am

Quoting planemaker (Reply 127):
You mean Rolls-Royce Marine's design for autonomous ships...

Now that just makes sense. We remove people from spacecraft as the weight/cost penalty is silly. But now that I think about it, such a huge part of a ships costs are repairs to keep the parts the humans need going... Build the engines to aerospace/oil gas turbine standards (even if reciprocating) and you just have a vessel that needs a periodic C and D check... Ghad, the cost savings would be immense. Getting rid of the bridge and crew quarters would allow another row of containers which on the big ships is 26 TEUS wide, 20 high or another 520 TEUs for little to no cost. So cost savings *and* revenue...   

http://www.bbc.com/news/technology-26438661
"Given that the technology is in place, is now the time to move some operations ashore? Is it better to have a crew of 20 sailing in a gale in the North Sea, or say five in a control room on shore?" he asked.

The same is true of aircraft. I've seen UAVs land under conditions no human pilot could have landed (they simulated for months to see if any pilot could land the unstable airframe... nope). But I do like the idea of the copilots comfortably in a 0-g ground environment ready to step in if required. But it won't be hands on the stick, it will be sending commands (e.g., land at this airport, priority, summoning an ambulance). It will be by mode switching, not stick and rudder. That the computer does much better already.

Lightsaber
 
FlyPNS1
Posts: 5560
Joined: Wed Dec 01, 1999 7:12 am

Will We Have Pilot-less Aircraft In The Future?

Wed Apr 08, 2015 12:09 pm

Quoting lightsaber (Reply 135):
You really need to try Google maps.

I use Google maps all the time, it's still not that good. It's getting better, but still inadequate. Google maps has poor understanding of the impacts of weather, traffic volume and road construction on a particular route.

Quoting lightsaber (Reply 135):
Seriously, your comment makes it sound like we're talking drones and not the latest technology. Commercial flights will be highly scrutinized. This is a conservative industry.

It is a conservative industry which is why it will be a long time before we have pilotless aircraft in part 121 operations. I have no doubt that it will happen, but not as soon as some are predicting.

Remember, this type of technology will eliminate thousands of high-paying pilot jobs. You think the pilots unions won't lobby to delay it? Eventually, the pilots will lose, but they alone can delay this technology a decade or more.

You want to know why Amazon won't be using drones to deliver packages anytime soon, it's not because of technology. The technology is clearly there. However, the likes of FedEx and UPS aren't going to let it happen anytime soon and lobbyied behind the scenes to make sure the FAA doesn't allow it either.

Quoting lightsaber (Reply 135):
It amuses me how we'll have these threads and during the discussion there is always an advancement or a program announced (such as this DARPA) during or right before the discussion that helps show the progress being made. Yet some pretend the technology is struck back in 1980.

This research by DARPA is far from cutting edge. Not to mention that for every DARPA success, there are tons of failures or programs that stall out. Of course, the DARPA failures don't get nice press releases.
 
fastmover
Posts: 419
Joined: Mon Mar 30, 2015 5:37 pm

Will We Have Pilot-less Aircraft In The Future?

Wed Apr 08, 2015 2:53 pm

Quoting lightsaber (Reply 137):

These "drones" that landed in conditions that pilots could not, what were these conditions and what was the landing like?
Will we all be flying around in unstable airframes in the future?


Again have you ever ever followed a pilot around for a trip to see the ENTIRE system he works in not just the hands on flying part?


We get it computers can do great things but again you ignore ATC, FAA, new aircraft types YOU HAVE TO CHANGE THE ENTIRE SYSTEM. We added I pad mounts to our aircraft it took almost a year to get the FAA to sign off on it and that is just a thing to hold an I pad. Sure you can go build a drone and fly it over IRAQ and if it crashes who cares, but the same can't be said for a brand new airliner, you have a long long way to go.
 
Pihero
Posts: 4318
Joined: Mon Jan 31, 2005 5:11 am

Will We Have Pilot-less Aircraft In The Future?

Wed Apr 08, 2015 5:37 pm

In fact, this discussion is about displacing the pilots - i.e breathing humans at the pointy end of the airplane - in favour of operators on the ground.
As a matter of fact, the proponents of the pilotless aircraft are only displacing the human variables of the OPS equation, without removing them. At all.
It goes, obviously, without saying that the guys on the ground will be less prone to errors and mistakes than the pilots.

Are they ?

Situationnal awareness is a big part of air transport safety... and I doubt the the ground operator will have an improved one, especially when the chips are down and the airplane is suffering a major break-down inside its highly computerized systems.

As for progress and the mistake / error-less engineers, the following excerpt proves that mistakes -even small ones - could happen, with dire consequences :

Quoting lightsaber (Reply 137):
. But I do like the idea of the copilots comfortably in a 0-g ground environment ready to step in if required.
A zero-g ground environment ? like weightless ?
With a simple lapse like that, and assuming the airplane takes off in direct law, what happens when normal law slides in ?... to me it will look for a 1g acceleration from it's programmed zero... or the contrary ?
What would be the consequences ?
Whatever they'd be, it wouldn't be pilot error, this time.
 
Tangowhisky
Posts: 667
Joined: Thu Jun 29, 2006 2:26 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 09, 2015 12:07 am

Quoting Pihero (Reply 140):
As a matter of fact, the proponents of the pilotless aircraft are only displacing the human variables of the OPS equation, without removing them. At all.

You are correct and therefore a pilot-less aircraft does not work. The way to go is fully autonomous. Your arguments support the facts that all this nonsense talk of remotely piloted aircraft is not the way to go.

A fully autonomous aircraft with the upgrade to the existing ATC system would be the way to go.
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 09, 2015 5:30 am

Quoting lightsaber (Reply 135):
How many objects can a pilot or pair of pilots track accurately. e.g., distance to a runway, aircraft in front of them, behind them, and other potential flying objects... What you thought was possible just leapt forward. When we were quoted the price we almost gagged (on how cheap it was)! When they noted how much of their code was from open source video games we smirked as when code is good... use it! I'm under NDA, so I can't say which aircraft, but there are aircraft out there flying with code from open source or licensed video games. Not the full game, just subroutines to do certain flight logic.
NVIDIA??  
Quoting lightsaber (Reply 135):
Oh... for a drone it might be pure GPS, but a proper UAV navigates by:
GPS, inertia, pitot tubes (many more than today's commercial aircraft), terrain following (visual and radar), and other 'proprietary tricks.'

You have been fortunate to work with "big toys" you can't talk about but even at the "retail toy" level, compared to only a few years ago, the pace of innovation is still pretty impressive...

Quote:
The Phantom 3 quadcopters also gain the ability to hold their position without using a GPS location signal, making them more suitable for indoor use.

The platform has a downward-facing camera that creates a 3D map of the ground underneath it in real-time, and identifies key vectors in that map and notices if there's any movement of those vectors and automatically adjusts itself to hold a position.
Quoting lightsaber (Reply 135):
And the publicly available information trails the R&D labs by 5 years (or more).

Which is pretty incredible. But to get an understanding of the pace and rate of technological progress the publically available stuff is nonetheless sufficient. For example, this was just recently announced. Stuff like this is coming out continuously.  
Quoting lightsaber (Reply 135):
It amuses me how we'll have these threads and during the discussion there is always an advancement or a program announced (such as this DARPA) during or right before the discussion that helps show the progress being made. Yet some pretend the technology is struck back in 1980.

That "pretend" hasn't changed since the first DARPA Grand Challenge was discussed on A.net.. Almost everyone on here was saying that it would never be won. Then that the Urban Challenge would never be won. Then that there would be no commercial interest. And now we have a race by all the major manufacturers plus silicon valley... and yet people are still "pretending". This is a fresh example of what was kick started by the DARPA Grand Challenge.
 
User avatar
notaxonrotax
Posts: 1334
Joined: Wed Mar 16, 2011 2:29 pm

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 09, 2015 6:18 am

Who would decide on a medical diversion?
Who would decide an unruly passenger is just getting out of hand, warranting a diversion?
Who would decide on seat belt signs on / off?
Who would communicate with ATC with regards to the reason of delays for their passengers etc?
Who would troubleshoot in-flight issues like a passenger reporting a loose panel or any other things that appear out of order?

I am guessing that in case of a pilotless plane, the job as a senior purser just became a whole lot more important / stressful.
Would they have to be trained even further for these additional duties? I am guessing yes......
Would they be paid higher? Would this mean we would need 1 more cabin crew member to be on board?

Considering all that: what are the cost savings in the end?

No Tax On Rotax
 
tomlee
Posts: 610
Joined: Sat Aug 21, 2010 9:01 am

Will We Have Pilot-less Aircraft In The Future?

Thu Apr 09, 2015 6:34 pm

Quoting Notaxonrotax (Reply 143):

Not requiring a pilots license or the testing/requirements/constant screening of that would probably be a large cost saving. Still makes sense to have pilots though even with a fully autonomous plane because of defence in depth, especially if it hasn't been in use for a long time to find the unknown unknowns.
 
mrocktor
Posts: 1391
Joined: Mon Jan 24, 2005 12:57 am

RE: Will We Have Pilot-less Aircraft In The Future?

Thu Apr 09, 2015 10:29 pm

Quoting fastmover (Reply 41):
I feel that those of us who fly aircraft in the airline world know that there is a long long way to go before this happens.
I know its fun to say we don't really need pilots or computers are better at flying planes but until you sit in that seat and really understand what goes on I think many would change their tune.

I work in airplane design, and I can tell you for a fact we could do it right now.

But...

Quoting SEPilot (Reply 49):
There is a difference between what can be done, what should be done, and what will be done. I have no question that pilotless aircraft can be done, but I do not believe it should be done or will be accepted.

Where I agree, even though we can do it, this does not mean we should do it. Better task distribution between humans and automation is a field ripe for improvement though. Unfortunately regulation does not allow much innovation (wihtout undue cost and business risk).
 
tomlee
Posts: 610
Joined: Sat Aug 21, 2010 9:01 am

RE: Will We Have Pilot-less Aircraft In The Future?

Thu Apr 09, 2015 11:41 pm

Quoting mrocktor (Reply 145):
Where I agree, even though we can do it, this does not mean we should do it. Better task distribution between humans and automation is a field ripe for improvement though. Unfortunately regulation does not allow much innovation (wihtout undue cost and business risk).

The regulations are important if you move too quickly you can run into many problems you never expected could happen (just because it is demonstrated doesn't mean there isn't some edge case it will fail catastrophically). If you take the slow approach it will be much more likely to succeed and work without major problems. It is all about a proper level of regulation.

Things like using GNSS data is something that is happening now and slowly advances in sensor/control technology will trickle down into production commercial aircraft. You certainly don't want to be flying on a beta-test flight control system platform.

More computer based assists is the way forward as you can introduce new technologies while still operating under the same systems and procedures as before and everything if done correctly will be a smooth transition.

Even with drones having the control room as close to the vehicle as possible is ideal and for a global aviation network the closest place to the vehicle is the cockpit it is safer that way as there is no remote control link required.
 
planemaker
Posts: 5411
Joined: Mon Aug 25, 2003 12:53 pm

RE: Will We Have Pilot-less Aircraft In The Future?

Fri Apr 10, 2015 2:14 am

Quoting tomlee (Reply 146):
If you take the slow approach it will be much more likely to succeed and work without major problems.

It is going to be gradual. As has been discussed in past threads, the progression will start off with SPO in military and cargo before migrating to pax, starting from smaller to larger aircraft, And the process repeats itself with full automation. A big issue will be flight boredom. A recent survey of 777 pilots said they spent just 7 minutes hand flying during a typical flight. Airbus pilots reported half that time. With increasing aircraft and systems automation coupled with NextGen, pilots will have little to do during flights.

Quoting tomlee (Reply 146):
and slowly advances in sensor/control technology will trickle down into production commercial aircraft.

Slow advances? The 787's AHM already transmits 28 times more data than the 777. And the 777X will do likewise to the 787. And it isn't just Airbus and Boeing. GE, for example, has already spent several billion on the "Industrial Internet" and "Intelligent Platforms".
 
tomlee
Posts: 610
Joined: Sat Aug 21, 2010 9:01 am

RE: Will We Have Pilot-less Aircraft In The Future?

Fri Apr 10, 2015 2:31 am

Quoting planemaker (Reply 147):
With increasing aircraft and systems automation coupled with NextGen, pilots will have little to do during flights.

Just give them VR training in flight with full motion simulation (you can probably safely use the plane to induce the required acceleration motions without deviating significantly from the allocated space call it simulation turbulence for passengers, give them a VR headset so it can fed them simulated visual data/audio). You can probably train pilots to handle more situations than existing training programs perform. I wonder if they will ever use touch screens instead of real buttons for the control panels (not the displays), I do like clicky buttons they have a much better tactile feel over a glass screen.

Quoting planemaker (Reply 147):
Slow advances? The 787's AHM already transmits 28 times more data than the 777. And the 777X will do likewise to the 787. And it isn't just Airbus and Boeing. GE, for example, has already spent several billion on the "Industrial Internet" and "Intelligent Platforms".

Well compared to consumer products a plane probably has less control bandwidth than single USB3 port. Even with the Ethernet based network air databus they pay a penalty for all that determinism and redundancy (which is good to have obviously but it is still slower than regular Ethernet)

They do lag the consumer electronics space a lot and it would definitely qualify as slow compared to the retail space where you have a new product line every year basically. Planes don't push that kind of volume and neither to the system mfgs. Not that being slow is bad for a safety critical industry.

Also alternative sensor systems like detecting stall conditions directly on the wing itself are not really that common. (Say a distributed network of digital barometric sensors placed in many locations so that unless the plane is completely covered in ice you always have enough data to detect relevant air parameters) Or putting strain gauges all over the plane in a fully digital network so you can by the moment detect external loading on the plane. I'm sure they do this when they test the aircraft but at some point it would be cheap enough to just have it always built in.

[Edited 2015-04-09 19:37:56]
 
fastmover
Posts: 419
Joined: Mon Mar 30, 2015 5:37 pm

RE: Will We Have Pilot-less Aircraft In The Future?

Fri Apr 10, 2015 3:31 am

Quoting planemaker (Reply 147):

While they are not "hand flying" they are telling the aircraft what to do.
Its not like they turn the autopilot on and go to sleep.
You really minimize what a pilot does and I find it kind of annoying.

Popular Searches On Airliners.net

Top Photos of Last:   24 Hours  •  48 Hours  •  7 Days  •  30 Days  •  180 Days  •  365 Days  •  All Time

Military Aircraft Every type from fighters to helicopters from air forces around the globe

Classic Airliners Props and jets from the good old days

Flight Decks Views from inside the cockpit

Aircraft Cabins Passenger cabin shots showing seat arrangements as well as cargo aircraft interior

Cargo Aircraft Pictures of great freighter aircraft

Government Aircraft Aircraft flying government officials

Helicopters Our large helicopter section. Both military and civil versions

Blimps / Airships Everything from the Goodyear blimp to the Zeppelin

Night Photos Beautiful shots taken while the sun is below the horizon

Accidents Accident, incident and crash related photos

Air to Air Photos taken by airborne photographers of airborne aircraft

Special Paint Schemes Aircraft painted in beautiful and original liveries

Airport Overviews Airport overviews from the air or ground

Tails and Winglets Tail and Winglet closeups with beautiful airline logos