Fully automated flight

[fireflybob]

Ok it's Christmas, a time to reflect methinks!

How far are we from fully automated flight for passenger jet transport operation? (I have my own views on this!).

Will automation ever be able to land in crosswind components of 30 kts on a gusty day?

Will passengers be happy to go flying in a fully automated aeroplane?

How many of our passengers think it's all automatic already and we just sit there all day pushing a few buttons and drinking the coffee?

[NWSRG]

Don't think that passengers will be happy in the knowledge the machine is doing it all until they're prepared to let their cars do it all for them as well!

We're now on the verge of the first autonomous vehicles...once a top range BMW or Merc gets the technology, the rest of the world will be about ten years behind. So maybe in 15 years autonomous vehicles will be the norm.

Allow another ten years for full market penetration and acceptance...so in 25 years the world will be travelling in cars that drive themselves. At that stage, folk will accept an aircraft that can do the same, BUT, they will still want to know that a real person is up front, just in case!

As for the technology, it's probably not on the critical path...it will come before the passenger acceptance.

[411A]

Will automation ever be able to land in crosswind components of 30 kts on a gusty day?

One type has been doing this for well over thirty years, in regular line service.
It was manufactured by the Lockheed California Company.
Its name is....TriStar.;)

Will passengers be happy to go flying in a fully automated aeroplane?
They do this now...in the above type.
And, it was the first widebody type to do so.

Pilot(s) push the switchlights, airplane flys automatically.
Works good, lasts a long time.

Of course, others have 'caught up' now, but Lockheed was the first.

[CaptSeeAreEmm]

Good question.

Boeing and Airbus are still pushing out hundreds of "piloted" aircraft every year - There are thousands af A320's and 737NG's, new models like the A380 and 787 are just now coming on to the market - most of them will last atleast 30 years. So I dont think we will see anything major the next 30 - 40 years.


Not to mention the probably extensive certification period an "pilotless" commercial transport aircraft will have to go through - is there even a daft CS-25 underway to cover this?

[Rainboe]

OK- someone has to ask, what's a 'CS-25'?

Full automation- full generation away still. I can't even get properly reliable radio control of a model plane even now. UAVs regularly just 'get lost'. No computer yet made can handle all the separate complexities of different faults and factors, sometimes happening in combination. They all do very weird things, and all the pathways have not been checked because it would be impossible. We haven't even got automation for mainline trains yet, or cars- how on earth anybody can expect it for aeroplanes in the next 25 years I cannot imagine!

But once it appears, there is no point in having a 'fallback' pilot present! His skills would not be of much assistance on a full scale emergency landing.

[172_driver]

OK- someone has to ask, what's a 'CS-25'?

Certification Standards 25, new term for JAR 25.

[BOAC]

OK- someone has to ask, what's a 'CS-25'?
Let me google that for you (http://lmgtfy.com/?q=CS-25)

Jingle bells and all that.

Nice quote from Michael Caine this am:

"When you are going through hell - keep going"

EDIT: Bah humbug, 172:ok:

[Wunwing]

Depends on where the aircraft is to fly. Making an assumption that it can technically be done in the next few years who gets ground controll of the aircraft? Not a big decision in say domestic USA, but a big political, security and legal problem for an aircraft flying Austrlia to UK.I'm sure the countries being overflown will want some input but security questions alone would cause some serious heartburn.In the end I suspect the status quo will remain for the foreseeable future.

[parabellum]

As has been said before, many times, it will never happen due to security considerations.

Imagine the outcome if suicidal terrorists took over any sector of ground control, either physically or by using equipment that could 'block' the legitimate ground control, 9/11 would pale to insignificance by comparison.

[Exaviator]

Doors closed - engines started - cabin crew doing last cabin preflight checks.

Announcement over P.A.

"Welcome aboard ladies and gentlemen, today is a worlds first. There are no pilots aboard, you are on a fully automated flight."

"We invite you to enjoy the flight in the meantime sit back, relax and remember, nothing can go wrong - nothing can go wrong - nothing can go wrong - nothing can go wrong - nothing can go wrong......."

Yea Right :D

[Old Smokey]

How far are we from fully automated flight for passenger jet transport operation?

About 25 years.

Will automation ever be able to land in crosswind components of 30 kts on a gusty day?

They can now (Refer to 411A's post.)

Will passengers be happy to go flying in a fully automated aeroplane?

If an aircraft load of passengers were told today that they were on a fully automated flight, they would all get off. If an aircraft load of passengers were told in 25 years that they were on a fully human controlled flight, they would all get off. Acceptance will be gradual.

How many of our passengers think it's all automatic already and we just sit there all day pushing a few buttons and drinking the coffee?

ALL of them. (But they stay on board because they still believe that we're a good back-up system).

Regards,

Old Smokey

[potkettleblack]

Its already happening now in a different guise. We originally had a cockpit full of pilots, an engineer and a navigator. Then we lost the navigator with INS etc. Then we lost the engineer when the cockpit was automated and we got glass. That left us with two pilots. In some countries today only 1 of these pilots has a type or instrument rating.

At the moment we have situations in Europe where only the person sat in the LHS is being paid to go to work. Maybe the next move will be to dumb down the role of the F/O so that they are only qualified to "relieve" the commander in the cruise, work radios etc. Then the next natural progression may be to get rid of the costly captain and have two of these super duper radio operators. They will be in voice or data contact with the ground and can pull any necessary cb's in order to rectify faults and suchlike.

After a while we might only need one of these radio operators. Eventually "testing" will prove that we don't need anyone up front and the previously airborne operator can now sit behind a computer on the ground controlling a whole host of aircraft.

Perhaps ultimately through datalink clearances and suchlike ATC will be flying the aeroplane, or a computer model at least as there probably won't be any "real" controllers anymore.

[lowcostdolly]

Guys pardon my ignorance here but how could you ever take human beings out of the flight deck safely...... ever?

I operate on the Airbus 319 and in normal flight or an auto land the Pilots appear to be just monitoring what the computer is doing. In flight that gives them lots of time to do what most pax think they do.......chat to the CC, drink coffee etc etc ;)

Now I know nothing about the workings of the airbus computer except:

It takes a Pilot to program it pre flight because it doesn't appear to be able to do this itself.
If it has a serious hissey fit on the ground we seem to stay on the ground until a human being sorts it out. That can take a long time and again it doesn't appear to be able to fix itself.
If it plays up in the air on the flights I have been on we have diverted. A Pilot has made that decision not the computer.As long as the airbus computer functions normally all is well but if you take the Pilots off the plane who will take up the slack when it decides to have a "tech" moment?

A few years ago when I was on my first FAM flight I saw for myself just how faillable a computer is. I was on a 737 and at my mob a newbie CC is encouraged to sit in the flight deck for take off/landing to gain appreciation of the increased workload at this time. On take off both pilots really chatty, relaxed etc

On approach something had failed on the computer to do with headings and the FO was flying the plane manually backed by the Capt. The atmosphere in the F/D was very different because at that point he was truly being paid for what he knew and could do in a non normal situation.

After landing the Capt explained to me what had happened none of which our pax knew.

I got off that plane feeling truely grateful for the skill and experience of the flight crew which kicked in when the automation failed.

If they hadn't been there.......

I would never fly on a Pilotless plane for the above reason either now on in years to come. An experience like that stays with you.

[potkettleblack]

Guys pardon my ignorance here but how could you ever take human beings out of the flight deck safely...... ever?

But it ain't about safety. Its about cost. Like it or not. That is the world we live in. Unfortunately there is a trade off between the cost of a human life and safety. No doubt at some point in the future someone will work out (probably an insurance actuary) that it is perfectly acceptable to have x number of hull losses in order to save the x billion per annum in wages, oversight, training costs etc.

Your right about some of the technical issues. There probably will be a need to have some sort of operator on board to pull a cb and do a computer reset. On the ground it doesn't matter as an engineer will be sent out and the aircraft will automatically advise maintenance of any tech issues prior to dispatch. They might even push the boat out and teach the operator how to do a landing in an emergency. More than likely it will be the person selling you a cup of tea who has a few more smarts than the most junior crew member. Everything else can be automatically datalinked into the Airbus eg: flight plans, clearances from ATC. The technology is already there now.

[Rainboe]

To put this one to rest- when I take a train, there is still a driver on board! Now I know the navigation must be a problem (like you have a choice?), and deciding what altitude to fly at (like you have a choice?), BUT, if you can't even have trains safe to run without a driver, how on earth does anybody ever think planes will be safe to fly without a pilot (for at least another generation. And please don't use driverless airport buggies and Disneyland monorails as an example! I get on a British, French, German, Italian, US train, and they all sport drivers....in 2010.

The problem is, pilotless planes will still need a pilot for customer acceptance. Unless that pilot regularly flies, you will probably be better off with a load of Microsoft computers up there all crashed, like mine does every day!

[deltahotel]

Don't forget the dog with big teeth. The cockpit of the future will have a pilot and a dog. The pilot is there 1. to make the passengers feel safe, 2. to feed. the dog. The dog is there to bite the pilot if he dares to try to touch anything in the cockpit.

Happy automation.

DH

[Sky Wave]

Rainbow

The Thameslink line should have a fully automated section between London Bridge and West Hamsptead by 2015. They seem to think that automation is the only way that they will get the required trains per hour through the core section. So it's coming!!

SW

[Rainboe]

'Toy trains'! Why aren't the thousands of trains running across Europe and the US/Canada running driverless if unmanned is so good? If they can't sort out complex enough technology to control driverless trains, how can they do it for aeroplanes in the next 30 years? How would they handle the Heathrow 777, Sully's A320 and the many other peculiar incidents beyond the remit of computerisation? Remember the Taleban have hacked into the video systems of UAVs now- ground control is out. Hackers will set out to break in. Making a computer system complex enough to cope will take years. They say all the computer pathways of the Airbuses haven't even been tested. Combination failures will overwhelm a computerised system, which will have to be so complex it will cost more than 2 pilots paid less then traindrivers! It's absurd. Customer acceptance will be zero.

And if anybody tells that corny pilot and dog joke again here, I think I will scream..........

[Anansis]

Why aren't the thousands of trains running across Europe and the US/Canada running driverless if unmanned is so good?There are. the Docklands Light Railway, the Kuala Lumpur Metro and the Dubai metro to name a few List of driverless trains - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/List_of_driverless_trains). It is much easier to incorporate such technology into new systems built from scratch than it is to add it to existing systems. The only reasons other railways do not operate driverless are-

(1) Cost of retrospectively installing software and hardware.
(2) Public confidence (traveling on an automated low speed commuter system is one thing, taking a 200mph driverless Eurostar another, even though the principles behind the technology are the same).

Of course, it is not really fair to compare automated railways with aviation. The infrastructure of a railway is fixed whereas, as Rainboe correctly points out, there are many more variables to consider when flying a plane. But it does go to show that proven technology to run automatic trains exists- the only thing stopping it being rolled out on high speed railways is acceptance.

[Brian Abraham]

If the trouble Honeywell is currently having getting some of their electronic cockpit displays in the A139 to work the answer is NEVER.

[Piltdown Man]

...or the the complete and utter rubbish that Honeywell have put into Embraers. If I made that stuff I'd make sure that my name wasn't on it. And if Honeywell avionics do ever make it into pilotless aircraft we'll all have to live like troglodytes. Unless of course, these planes are built by Embraer when we'll be perfectly safe. Because they won't actually get airborne.

PM

[411A]

Because they won't actually get airborne.

Harsh words, to be sure.
What exactly is the problem, as....in the past, Honeywell turned out some mighty fine avionics.
Usually....:}

[Brian Abraham]

Here's a sample 411A http://www.pprune.org/rotorheads/204240-agusta-139-a-48.html

[Pugilistic Animus]

YouTube - Boeing 737 Gear Up Landing (http://www.youtube.com/watch?v=FOtDcvQuhUs)
YouTube - Chaka Khan- Through the fire. (http://www.youtube.com/watch?v=6QUwDBF53bc) I have a musical mind:)

I'm not sure I'll ever be aboard with that idea:eek::\:eek:

Of course I'm inextricably biased:}

[Piltdown Man]

What exactly is the problem, as....in the past, Honeywell turned out some mighty fine avionics.

There are two separate issues here. The Honeywell avionics I do battle with everyday are poorly specified, designed and built (Embraer 190 to be precise). Jeez, do I HATE the Guidance Panel (aka Flight Mode Panel). The MCDU has some nice features but the FMS has absolutely no concept of the aircraft's performance. The OEM documentation is poorly written and each module within the system appears to have been put together by a different team who refuse to speak with any other. As for the aircraft - well if it was a car, I'd be insisting on my money back. It is temperamental, unreliable, heavy (both in control loads (fly-by-wire?) and DOW) and built out of a peculiar plastic that breaks when looked at. I should be impressed flying a brand new aircraft rather than sitting there wondering "why?"

Honeywell have lost the plot on this thing and instead of glistening on the top of a pile of excrement, their products merge into the background. Embraer can and should have done better.

PM

[Brian Abraham]

From Aviation International News today

SAIBs Address Honeywell FMS Software Glitch
The FAA late last week issued special airworthiness information bulletins NM-10-12 and NM-10-13 to address a glitch in software for Honeywell NZ-2000 and FMZ-2000/IC-615/IC-800/IC-1080 Primus Epic flight management systems. Affected units are installed on a wide range of business jets manufactured by Cessna, Bombardier, Dassault Falcon, Gulfstream, Hawker Beechcraft and Embraer, as well as several regional jets and airliners. The problem was discovered during an Rnav approach on a Bombardier Challenger programmed with NZ-2000 FMS software version 5.2. According to the FAA, the airplane was flying the WAATS Three Arrival STAR into Salt Lake City International Airport. (The FAA incorrectly identifies the arrival as “WATTS” in SAIB NM-10-12.) “A different runway was selected in the FMS during the STAR, and resulted in the FMS navigating the airplane toward the initial STAR waypoint instead of the next sequential STAR waypoint,” the FAA said. As a result of this incident, Honeywell issued a service information letter on October 15, to provide operators with guidelines for flight plan modifications made on the “arrival” page on the FMS, including a destination runway change. The FAA said this airworthiness concern is not an unsafe condition that would warrant issuance of an Airworthiness Directive.

[Halfbaked_Boy]

Somebody (a lot of people) mentioned costcutting as a point...

... Forgetting about how much maintenance a fully automated flight control (and decision making) system would require?

I can just see it now - the regulations requiring intricate and thorough inspection of all 'artificial human' systems after every completed day, if not every flight.

Add to this the certification required to get the whole thing online (as previous posters have mentioned), and you're not going to be saving much (if any) money at all. Certainly nowhere near what would justify 'x number of hull losses'...

[The SSK]

Just think about it for a moment …

The certification process – how long would it take for a pilotless passenger aircraft to be certified? How much would the process cost? Who would pay for it? Boeing? Airbus? What would be in it for them? Which certification agency would have the balls to say ‘we guarantee this will be 100% safe’?

Because at the end of the day, who would buy it? Seriously – who would be the first airline to operate pilotless? What incentives would they have to offer to get anyone on board? €9.99 tickets? Ryanair do that already. Regular airlines would match the pilotless fares and NoCrew Airlines would be out of business quicker than Varsity Express.

Or if they did stay airborne, the very first accident and that would be that – project cancelled, 25 years of development and billions of development dosh wasted.

Oh – and by the way, they wouldn’t be pilotless. The pilot would be sitting on the ground, in front of an instrument panel with flight controls at his disposal. Just as highly paid, just as expensively trained, all you would save are the hotel bills and the overseas allowances.

[169west]

... and still pilots required... and two needed... maybe with a different ATP (Astronaut Transport Pilot)!

Space Ship Two

http://cache.gawker.com/assets/images/gizmodo/2009/07/WK2_cockpit_close.JPG

[davidjohnson6]

Why aren't the thousands of trains running across Europe and the US/Canada running driverless if unmanned is so good?
There are. the Docklands Light RailwayOn each DLR train, there's a person who spends most of their time opening/closing the doors and checking tickets. Every now and again, the train control system plays up, and the door opener has to unlock a panel at the front of the train with buttons underneath and become a train driver.

[gcw01]

If you look at what has happened with the introduction of automatic trains you can get an idea of what could happen in aviation.

Train drivers on automatic lines are now called 'train operators' who are there primarily to close the doors when they can see it is safe to do so on a cctv monitor. They are trained to try and fix the train when it develops a fault (by following instructions on the screen) and if the train really pulls up in a heap to manage evacuation of the passengers. On the 'driverless' docklands line in London there is still a train operator on board the train but they perform other duties as well such as ticket checking. This is the way the industry is going across metro and commuter services, certainly in Europe.

So by analogy in aviation you could see 'fully automatic' passenger aircraft but they will not be pilot-less. They will have 'aircraft operators' on board who are trained how to sort it out in an emergency but do other things like serve the pax in the meantime. They will still need to sit in the front with windows and make excuses on the PA as to why the aircraft is running late - as do train operators!

G

[Helix Von Smelix]

I have used the DLR several times, and not seen a "Train operator" yet :uhoh:

[Wojtus]

I have used the DLR several times, and not seen a "Train operator" yet
I've spoke to DLR "operator" when I've been riding it once. Her primary task seemed to be pushing "close door" button. I've asked if it's so crucial that they couldn't automate it. She answered - of course, they did automated. But unions fought for their jobs.
Go Unions!

[flyvirgin]

I have used the DLR several times, and not seen a "Train operator" yet :uhoh:
LOL I get on the DLR everyday to work and there is always one on there, without fail.:ok:

[What-ho Squiffy!]

If you haven't heard of it - Google it now, and see your future.

At an indeterminate point in our future - some say well within 20 years - computers will exceed human capacity and intelligence. And here's the kicker: It is from this point that we - as humans - can in no way predict or even comprehend what will happen in the development of this new super-human intelligence.

Technology will increase logarithmically, as hyper-intelligence begats hyper-intelligence.

So, you can predict all you like, and say "Never!" or "Not in my lifetime!", but the only think that is certain is that we have no idea what the future holds for us in terms of technology.

The biggest challenge I see for any automated flight will be mastery over the atmospheric conditions. Turbulence, ice, microbursts, lightning, thunderstorms, tornados - the list goes on. Once the sensor technology can reliably sense the environment at a distance to afford avoidance, the game's up.

And who knows what else will come to fruition at the hands of the hyper-intelligentsia? Anti-gravity? Immortality as an upload? If you're around, it will be quite exciting!

[davidjohnson6]

At an indeterminate point in our future - some say well within 20 years - computers will exceed human capacity and intelligence. And here's the kicker: It is from this point that we - as humans - can in no way predict or even comprehend what will happen in the development of this new super-human intelligence. I thought Judgment Day was meant to have already happened and that we are now in the middle of the war against the machines ?

The Skynet Funding Bill is passed. The system goes on-line August 4th, 1997. Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern time, August 29th.

[Matt101]

I thought Judgment Day was meant to have already happened and that we are now in the middle of the war against the machines ?

Nah didn't all that change in Terminator 3 when they discovered killing Arnie and his mate in Judgement Day had only delayed events rather than prevented them....

Mainly I am just annoyed that you beat me to the joke!

In a more related manner I was watching a programme yesterday which discussed such matters. The biggest problem computers face in terms of automation, except of course all the bugs Bill Gates has put in them, is their ability to interact and learn.

You call a travel agent over the phone and tell them you need to go to New York, firstly the travel agent knows as you live in Cleveland that, flying is probably the best option so will ask "when do you need to fly', to which you reply that you have a meeting at 9 am on the 4th. So the travel agent starts looking at flights to get you into New York the night before or at least sufficiently before 9 am for you to shower, change and travel to your meeting. At no time did you say I want a flight, or I want the flight to depart at 9pm on the evening of the 3rd, but the squidgy brain in a person's head knows from more obtuse information what to do.

Can you imagine ATC passing on information to a pilot-less aircraft? It would be like traintracker in the UK; "After the beep please say the name of the station you are travelling from". "Eastbourne". "Did you say Glasgow?". :ugh:

Meanwhile at a UK airport by the time the automatic jet has deciphered what the Geordie accent is saying about the comms out aircraft heading towards it (No squawk from this GA aircraft so no lovely TCAS), and has worked out how to deal with multiple failures in ever worsening weather, it's too late as Microsoft fails again and the Stewardess is greeted with a blue screen error when she goes to the cabin terminal to find out the flight status. This error simply says "Please PA for a Pilot".

Will it never happen? Oh I doubt it, but is it close? I doubt that even more.

[Brian Abraham]

See the threads now running on the latest CX incident with the the A330 at Hong Kong. Do you see fully automated flight in your future? I don't think so.

[cdtaylor_nats]

Air Traffic Engineers have long spoken of the stupidity of having a computer in the air talking to computers on the ground via a human to human interface. Its the humans mishearing or misunderstanding communications that cause the problems.

As for checking the AI after every flight there is no need - simply reboot it and its back to the original state already tested.

[wiggy]

Um, it's all very well this techy talk of autolanding in 30 kts across (40 for the Triple)..but we do realise, don't we, that in the non-virtual world a lot of runways aren't ILS equipped - even in the US of A, certainly in Europe and Africa? Never mind passenger acceptance, we're a long way off full automation 100% of the time until we have GPS/GNSS autoland -how far off is that (serious question)?....

[Denti]

Well, in a way we do have GPS Autoland via the GLS system. However that still requires a ground installation as a ground based augmentation system is the heart of it. It is not operationally certified for better than CAT I, but autolands work like a charm.

[pb365]

Rainboe,
Your statement regarding train drivers pay is nonsense as is a good deal of what you say about automation for trains. From other posts elsewhere on this board it is all too clear that you are not a pilot so why don't you shove off and post your crap elsewhere?