Insider Series: Could Computers Replace Human ATC?
This post contains references to products from one or more of our advertisers. We may receive compensation when you click on links to those products. Terms apply to the offers listed on this page. For an explanation of our Advertising Policy, visit this page.
TPG Contributor “Vic Vector” is an air traffic controller at a major ATC facility in the United States. In this installment of our “Insider Series,” he talks about the limitations of technology and how you shouldn’t plan on computers replacing the human element of aviation anytime soon.
Between 1970 and 2014, the worldwide number of passengers who traveled by air increased from 310 million to 3.2 billion — a factor of more than 1,000%. In that same time period, the world population only increased by 97%. As safety and convenience in air travel increase, so do passenger levels.
The FAA’s share of the Department of Transportation budget totals around $15 billion and almost half of that goes toward the Air Traffic Organization (ATO). The ATO is unlike most government agencies in that it’s a performance-based organization — in other words, it’s designed to operate like a business, complete with all the common titles such as COO and CFO. The ATO employs more than 35,000 people, including controllers, technicians and other support personnel.
Like many businesses, payroll constitutes a large percentage of operating costs, and the ATO is no exception. Therefore in today’s climate of fiscal awareness, it’s no great leap to suggest a good way to cut costs in the federal budget would be to do what many other industries have done: outsource some human responsibilities to technology. Put simply, could a computer replace air traffic controllers?
The short answer is no, or at least not for a very long time given our current technological constraints. Now, before you accuse me of being biased simply because I like my job and don’t want to be replaced by a machine (which — full disclosure — is true), let me explain.
On some days, the work of controlling the national airspace system can seem quite routine to the well-trained and experienced professionals who occupy the towers and man the radar scopes. There are times when there are no adverse conditions to overcome, traffic volume is slow or steady and the clearances we issue are ordinary and unremarkable. But those days are the exception, not the rule, and there are a number of factors which regularly throw a wrench in the works of the national airspace system.
The hub-and-spoke method that most airlines have adopted has more consequences than the occasional long layover or missed connection. It also means that there are periods of increased traffic volume during the arrival and departure pushes of major airline hubs, which creates a bottleneck not unlike the rush-hour traffic jam near highway on- and off-ramps. Given that aircraft simply can’t stop in place like cars, our job is to sequence and space these arrivals so the flow continues as smoothly as possible.
There’s technology currently in place, known as Time Based Flow Management (TBFM), which seeks to allow a computer to sequence arrivals to major airports. The idea is that each aircraft en route to a particular busy destination is assigned a scheduled landing time. That time is then projected backward and requisite times are assigned to various arrival fixes along the route of flight. On a radar scope, the controller sees a number associated with each data block headed to an airport where TBFM is in effect. If that number is zero, then the airplane is right on time. A positive four means the aircraft is four minutes early and must slow down or be vectored to hit the scheduled metering time, and vice versa for a negative number.
TBFM is a great tool in theory, but at best all it can do is create a time to gain or lose — not a strategy to gain or lose that time given the other traffic constraints of a particular sector. Given the current technology, only a human brain can formulate a means by which to slow down or shortcut an aircraft to meet a metering number.
Weather is another common hurdle that both pilots and air traffic controllers deal with on a regular basis. Things like thunderstorms in the summer and icing in the winter can be dynamic and unpredictable. Oftentimes, we find ourselves scrambling to reroute aircraft or simply keep them separated as they all try to hit the same hole in a squall line. There’s no computer algorithm in place that can separate aircraft as they deviate randomly around adverse weather.
It’s not just ATC that faces these questions regarding the necessity of humans. In an ideal situation with no weather or adverse conditions, automation can do a great job flying an airplane. However, when an engine explodes or a hydraulic system ruptures, there are certain subjective decisions that must be made based on a variety of situationally specific information. Emergency situations are the CAPTCHAs of aviation, if you will.
We don’t currently have the technology to replace the dynamic and variable decision-making ability of the human brain in real time, and I don’t foresee us having that ability any time soon. For all we know, instead of Chesley Sullenberger and Jeffrey Skiles working together to land US Airways 1549 in the Hudson River back in 2009, a computer algorithm might have tried to land at the closest runway at Teterboro, killing everyone on board as well as 50 people on the ground in Weehawken.
Computer technology has revolutionized the way we work, communicate and consume information, but there are still some areas where it can’t approximate the work of a well-trained human brain. Controllers and pilots alike earn their pay not just for the routine work they do, but for their ability to handle the abnormal situations in ways computers just can’t — at least not with today’s technology.