Friday, November 20, 2009
Explaining the Air Traffic Breakdown
It wasn't the fault of a creaky old radar system, but of high-tech flight-monitoring computers.
The major failure of air-traffic control yesterday was yet another sign that
our outdated radar-based system needs to be replaced with a sleek new
satellite-based one, right? That's the
logical progression of much of the coverage out there.
The reality is that, yes, the system needs to be replaced. But yesterday's failure was a high-tech one that could afflict a system
based on satellites, too.
The problem wasn't directly related to radar, but with the National Airspace
Data Interchange Network, a system for processing flight plans and information
for all flights in the country. It
failed in both of its locations: Salt Lake City
and Atlanta.
This meant that automated regional FAA systems couldn't process flight
information. As a result, controllers
had to enter information manually. This caused delays that rippled across the country. "A satellite-based system would have had the
same problem," R. John Hansman,
an MIT aerospace and air traffic control expert, wrote to me this afternoon.
The Federal Aviation Administration hopes to roll out a Global Positioning System-based control system, called Next Generation or
NextGen, in stages. By 2020 most planes will carry a cockpit gadget that
continuously broadcasts the planes' GPS-derived location, altitude, and
speed to ground controllers. In later years, the system will extend so that this information is picked up by other planes, too, so that pilots can gain more control over their routing and spacing. As they beam their position information to one another they'll be able, to some extent, to self-navigate. However, there will always be an FAA air-traffic system keeping track. It's unlikely that pilots will ever be permitted to make all takeoff, routing, and landing decisions entirely by themselves in the event of failures of national air-traffic computers, as happened yesterday.
Massachusetts Institute of Technology
Comments
chimenti
11/21/2009
Posts:1
Whenever discussing air traffic control systems, it helps to keep some perspective. On a highway, cars are separated by feet. In the air, airplanes are separated by miles if not tens of miles. While rare mid-air collisions do occur, they are usually in areas where air traffic is concentrated near airports where ground based control is least effective.
When driving you are not expected to file a "drive plan" nor accept central control to tell you when to change lanes. Automobiles operate on a simple "see and avoid" system.
See and avoid works in the air too if a pilot can, in fact, see. The problem is that weather and distances can make airplanes virtually invisible. Also, the complexity of the current airspace system tends to keep pilots eyes on their displays inside the cockpit so they spend little time actually looking outside for traffic.
I think it's quite possible to develop augmented reality to greatly improve a pilots ability to see other traffic and use on-board computers to suggest separation vectors when necessary.
The advantage is that each pilot would only have to keep track of one or two aircraft that actually pose a threat of collision whereas a ground controller has to keep track of hundreds of aircraft in his sector. A pilot also has more at stake - I've yet to hear of a controller crashing a control tower.
I'm very much in favor of moving as much traffic awareness, control and responsibly as possible into the cockpit. Taken to it's logical conclusion, ground based control becomes redundant.
bildan
11/23/2009
Posts:20