I was discussing GPS with a colleague who is preparing for the exams to upgrade his commercial licence to an airline transport licence. I mentioned that they like to ask us to recall the frequencies on which GPS operates: 1575.42 and 1227.60. He asked why there were two frequencies, and I started to explain about standard and precise positioning service, and then realized that my explanation was out of date. "You know, I'm not sure how many frequencies there are any more. Every time I read about GPS it gets more advanced, and I don't honestly know what level of technology is experimental, cutting edge, or actually installed in our airplanes. It's not like I'm ever going to be in a position to either tune the frequency, or build a GPS unit out of stone knives and bearskins. I know how to turn on the box and get it to tell me the right direction to fly."
"And," he concluded, "this is how we will be replaced."
For all we can argue that a cockpit must have a skilled pilot or two in it, automation is increasingly able to do most of our job. Automation is far better than people at monitoring a situation that rarely goes wrong, over a long period of time, and then, hours later, doing a task that requires alertness and precision. I told the pilot that he wouldn't be out of work. He'd sit on the ground monitoring five or ten flights for any abnormalities. If the landing gear failed to retract or an elevator trim jackscrew appreared to be misbehaving, the normal flights would be transferred to another operator and he would focus all his attention on the affected flight, landing it by telemetry. Yep, there are situations where only a real live pilot in the cockpit could save the day, but there are also plenty of situations whe real live pilots have messed up where automation wouldn't have.
He got into the speculation. (Anything is better than studying for an exam). "And they'd all have TCAS, but instead of saying 'Climb, Climb,' they'd just climb. The airplanes in communication with each other."
"And with automated ATC."
"And mostly I'd just watch them. And pat the dog that was there to bite me."
We concluded that for every Sioux City there are a dozen Everglades, Silk Airs, or Little Rocks, and that that would be the new price of air travel. And then we went back to work.
4 comments:
Do you think the entire world can ever be THAT automated?
I work in automation. It ain't gonna happen in our lifetimes. Computers are good for lots of things but what they are absolutely terrible at is assembling random bits of data and drawing an independent (i.e. not pre-programmed) conclusion. And since you can't anticipate every permutation there will always be situations the computer can't comprehend.
Real life example...the Airbus that crashed while doing a demo. The computer overrode the pilot's input because it was outside of the parameters the computer accepted as normal and into the trees they went. It was outside of parameters because he was flying a low-level low-speed pass a maneuveur that wasn't "normal".
Come to think of it, my job will be outsourced to Bumfaq, India long before you'll be flying passenger flights by remote control from the ground.
Ah yes, Sioux City. My major college project this term was to do a presentation on air safety. I chose to highlight UAL 232 and the work of Haynes and his team, CRM, and a dash of forward planning and some bloody good luck!
As for automation, while it is mostly good and makes things easy, it can also cause its own problems. What's the saying? - to err is human, to really stuff up takes a computer ;)
Post a Comment