ATC Jobs at threat ??

[exraf]

link to old report - New Technology does away with the ATC ??. BAE Full scale trial by 2008.

http://www.telegraph.co.uk/news/main...24/ixhome.html

[BDiONU]

Gave me a laugh that they had all this speil about how great a fully computerised ATC system would be. Then in the body there was a link 'operational gridlock' to a story about the chaos caused when there was a computer failure
From my experience of ATC systems we will never reach the stage where a computer would be capable of taking on this human tasks. Tools it can do and it does provide but a computer is a computer and I cannot ever envisage a stage where a computer programme will be sufficiently well programmed and free from glitches and errors to the extent it could take this task on alone.

BD

[PPRuNe Radar]

Plus it's from British Waste of Space ..... so it will be 15-20 years late and about 10 times over budget

There are echoes here of the FAST programme (Final Approach Spacing Tool). The human knocked the socks off the computer every time simply because humans can think laterally and are constantly assessing options. Humans are flexible and can easilly change tactics. Computers are hide bound by rules and algorithms. Crap in = crap out

[Kirk Biddlecombe]

I must admit, this was something I researched into prior to submitting my application.
The conclusion I found was as has been mentioned above. A few years ago I remember my A-Level Computing teacher telling us computers would never be able to completely take over systems of this nature.
I also agree with it.

Kirk

[CAP493]

Even if a computerised ATM system can be devised that will automatically provide safe separation and uplink instructions directly into the FMC, both flight deck and the ATCO will still need to act as interventionists when a glitch arises.

Problem is...neither will have the requisite 'hands-on' skills any longer.

Furthermore, just who is going to decide which aircraft is # 1 at for example, Heathrow when BAW and BMI both arrive at Biggin at the same time; or at Stansted or Luton when RYR and EZY both arrive at Lorel at the same time; or for departure at Edinburgh when EZY, BAW and BMI (or at Madrid when IBE, EZY and BAW) all start and so are ready for departure (and have the same CTOTs) at the same time???

Any suggestion that the airlines will cooperate or reach a mutual agreement is just fantasy.

Trouble with all these so-called automated systems is that they're designed by engineers working for companies who pay 'em to come up techno solutions; but these so-called solutions can't deal with the wild card componant called "human beings"...

[ATCO1962]

Hi Radar,

Maybe some of our Dallas colleagues can comment further but I don't think that FAST and PFAST were as bad as you may indicate.

I read an article by an ATC journal that described how DFW implemented this system of on-screen "suggestions" and how, initially at least, there was great reluctance to use the info as human interpretation of data and subsequent action were considered superior to any computer analysis. The article went on to say that by the end of the trial period or when the article went to print, controllers were using the "suggestions" 96% of the time.

And BD, you need more faith in computers. When Japan introduced the bullet train many moons ago, the bulk of the work was done by computers. The train driver was just sitting pretty watching the nice lights flash at him.

For myself, I think our generation may not accept a computer-driven system but I would be very surprised if, after 30-40 years, the entire ATM machine weren't run by these gadgets. If you want an ATC career like we enjoy now, don't get your hopes up.

[BDiONU]

Originally Posted by ATCO1962

And BD, you need more faith in computers. When Japan introduced the bullet train many moons ago, the bulk of the work was done by computers. The train driver was just sitting pretty watching the nice lights flash at him.
For myself, I think our generation may not accept a computer-driven system but I would be very surprised if, after 30-40 years, the entire ATM machine weren't run by these gadgets.

My job is in ATC systems, specifically NERC system. The complexity proposed by a fully automated system is mind boggling. The glitches and errors which will appear in the millions upon millions of lines of code does not bear thinking about.
However even before any coding happens the requirements would have to be written and they would take many many man years, assuming there was a 'group' who had the executive authority to make decisions and they could reach concensus quickly.
Computer systems can and will provide tools but they will not be in the driving seat.

Bullet train analogy is not the best one, trains don't travel in 3D and they run on fixed rails

BD

[CAP493]

Maybe some of our Dallas colleagues can comment further but I don't think that FAST and PFAST were as bad as you may indicate.

ATCO1962 - here you're talking about advanced computer-support systems for the human controller, not a fully automated ATM system that operates without intervention from the flight deck or Ops room. To that extent, I agree entirely with what you say.

As to the notion that the system will in 30 to 40 years be run entirely by computer-based automation (BDiONU), this is precisely the rationale that one Duncan Sandys as the then Minister of Defence propounded around 40 years ago, for the scrapping of all military interceptor aircraft in favour of guided missiles; and if total automation is such a brilliant idea, why do both the US and Russian space programmes still require human control staff on the ground and human crew on the flight decks???

These systems are only as good as the software programers and if a human - whether singly or in a team - is involved with that task, somewhere, somehow, sometime, there will inevitably be minor sofware faults incorporated some of which will be entirely latent, until one day, under very specific circumstances...

[ATCO1962]

Hi BD and CAP,

You're absolutely right in indicating that these tools are currently just that....tools to be used by very real humans to assist in the task of assimilating information that will sort out some complicated airborne problems.

However, when I think of the massive strides that have been made in UAV technologies, space technologies and the huge changes we've seen in the last 3 decades in computers, it tends to make me think that we will ultimately have some kind of "thinking" software that will enable those millions and millions of code lines to be prepared by an "aware" machine on the fly, so to speak. Software that will, ultimately, be able to repair flaws in its own logic as it develops. You can laugh, but there's only one way that technology is heading and its not backwards.

We introduced an automated ATC system some 10 years ago where I work and even 1 and a half million lines in a relatively simple environment was too much to bear at times. But that's thinking in yesteryear's understanding.

Admittedly, my bullet train analogy falls down but I just wanted to show that the Japanese were prepared a long time ago for computer controlled machinery that had the ability to kill you if it went wrong.

I went to a university a few years ago in Bangkok and was taken down to the area in the library where they store all the finished PhD and masters theses. There were thousands of these on all manner of specialised topics, very few of which had been acted upon. Multiply that by the thousands of seats of learning around the world. Then imagine when someone/thing can tie these pools of info together and totally new fields of knowledge come to the fore. I have no trouble visualising an ATM future almost entirely devoid of human interference. I'm very happy to be proved wrong but just by saying "It'll never happen" is just the challenge that some young buck/doe needs to change the world. Never say never, you dinosaurs!!

[BDiONU]

Originally Posted by ATCO1962

However, when I think of the massive strides that have been made in UAV technologies, space technologies and the huge changes we've seen in the last 3 decades in computers, it tends to make me think that we will ultimately have some kind of "thinking" software that will enable those millions and millions of code lines to be prepared by an "aware" machine on the fly, so to speak. Software that will, ultimately, be able to repair flaws in its own logic as it develops. You can laugh, but there's only one way that technology is heading and its not backwards.
I'm very happy to be proved wrong but just by saying "It'll never happen" is just the challenge that some young buck/doe needs to change the world. Never say never, you dinosaurs!!

A 'thinking' ATC computer system? You need to think about how conservative people involved in ATC are from the safety perspective. Artificial Intelligence is still a long way off, we're still pretty much in the fuzzy logic stage. So you'd need to have a very advanced AI system which was totally 100% proven to be reliable and error resistant. One which had been running a safety critical system for many years. One which had sufficient back up's to cope with ANY minor system failure. One which had some back up in the event that the whole system crashed.
I know of no complex computer system currently on the market which gives the reliability which would be required in an ATC system, which would (I think I have the numbers right) be somewhere around 10 to the minus 7.

In addition there is a vast amount of effort required in writing a specification.

We're developing some additional tools to add onto our ATC system. It has taken years to get to where we are now and its only an addition to an existing system.

"It'll never happen" says this dinosaur (who has been involved in ATC for over 30 years, who works in ATC systems and who's halfway toward a degree in computing & IT), just look at Mode 'S'!!

BD

[Nimmer]

All this talk of fully automated ATC systems!!! Lets talk about fully automated planes, the technology is there now, it is possible. Only the general public won't fly without two people at the front monitoring computer systems, however when things go wrong etc, the human takes over.

Same thing applies with fully automated ATC.

[ATCO1962]

Fortunately for me, I had a grandfather who used to work on Dehavilland aircraft way back when and he preserved a large number of some very old (@1918) Flight International magazines. One vivid memory I retain of some very pleasant reading was a quote from one of the leading aerospace experts of the time who said something like, "The idea being advanced that we will have flying machines that can carry several hundred people at one time at speeds in excess of 250mph is absurd. The laws of aerodynamics simply preclude this possibility."

I'm sure what this esteemed gentleman meant to say was that current understanding of the laws of aerodynamics precluded the thought of these novel ideas but he failed to remember that current thinking never equals future reality. Advances are being made every day that will mean, in a 100 years, our world will look, in a technological sense, nothing like what we see today.

The technology is on the way. Our current experts and generation will never allow full ATC guidance to be conducted solely by computers. But our kids and their kids??? They don't share our innate sense of distrust in computers. Increasingly, they are learning to trust greater levels of service and reliability from a huge variety of computers that we don't even think about, ranging from in-car computers to kitchenware.

I'm truly glad that I've enjoyed the hands-on ATC career that I have had but I don't think that the task will look anything like it does now in another 30-40 years.

[ebenezer]

They don't share our innate sense of distrust in computers

They will do the day after a midair takes place because a line was in error or there was a glitch or corruption in the uplinked or downlinked data.

[BDiONU]

Originally Posted by ATCO1962

The technology is on the way. Our current experts and generation will never allow full ATC guidance to be conducted solely by computers. But our kids and their kids??? They don't share our innate sense of distrust in computers.(

Its not an innate sense of distrust in computers which I 'suffer' from. Its the involvement I have when the systems don't perform as it says on the tin due to a programming error, or a lack of a correctly specified behaviour, or a bit of hardware going wrong or simply the data connection failing.
In other words, computers are not sufficiently developed nor mature enough to be trusted, yet.

BD

[Jerricho]

To echo what has been posted here and in previous threads regarding the issue, IMHO full automation will not be feasable when you factor in unknowns such as weather (both enroute and terminal), blocked runways and mechanical problems both on the ground and in the air. A CB parks itself over an airfield for an extended period is just the start. CAT is another one.

Sure computers can learn and mimic behavious, but (and this is a BIG but) if you read through accident reports from year dot, how many situations have been saved by pure instinct and a little bit of luck?

[airac]

Mr BDionu
Its not an innate sense of distrust in computers which I 'suffer' from. Its the involvement I have when the systems don't perform as it says on the tin due to a hardware going wrong, or a lack of a correctly specified behaviour, or a bit of hardware going wrong or simply the data connection failing.
In other words, computers are not sufficiently developed nor mature enough to be trusted, yet.


 
   

    

  
 
  
 

(marlett )

[Hold West]

Originally Posted by ebenezer

They will do the day after a midair takes place because a line was in error or there was a glitch or corruption in the uplinked or downlinked data.

Brings new meaning to the "Blue Screen of Death".

[Scott Voigt]

Howdy;

PFast died here in the US... The trials at DFW showed that it had bugs and without a LOT of money thrown at it to just make it as good as the controllers. Now that said, the other portions of CTAS do work very nicely and are used daily. However, the FAA found that CTAS like any other bit of automation isn't one size fits all. The found that it had to be tailored for each site to make it work correctly. It was like writing new software for each facility with only the core engine being the same.

As to what was written in the Journal of ATC, you have to remember that most of that stuff is written by the contractors who are trying to sell the stuff, by the contractors who are trying to advise the different agencies what they need, and then by folks who are working for the agencies and who want to retire and go work for the contractors <G>... They seldom publish the stuff that we have come up with that talks about some of the bunk that they have put out <G>..

As to computers eventually doing our job. Sure it will happen some day. I am convinced that someone is going to develop a computer and software to go with it that will mimic the human thinking function and do it well. I have faith in man to make that achievement. I however know that today we aren't there, and I will NOT see it in my carear. I also know that we will see in the not so distant future frieght carriers go to UAV's and passenger carriers go to one pilot operations. People say never, but think again, it wasn't all that long ago that we flew with four man crews then down to three and now two. Automation took over those other two positions. If you have a computer flying the plane, then you only need one person to monitor everything. It is going to happen and passengers will be ready for it in due course.

regards

Scott

[Flaps ten please]

Don't worry - we will have run out of aviation fuel and invented teleportation before that happens!

[atcSA]

The technology for computers has grown drastically over the couple of decades but speaking from personal experiences (ATC and computer programming), computers cannot "think" or compute the amount they need to, to actually be a 100% safe. Anyone can notice this!! Have you ever installed or downloaded a program that never needed an update or patch? Computers could possibly do this but the fact that a machine which does NOT think and runs by rules and limitations is going to have the responsibility of thousands of lives a day including mine and yours, makes me rather whip out my bicycle.

Heaven help us if they're going to run Microsoft!?!

G