Saturday, November 8, 2008

Fallows on the Future of Aviation

Reading James Fallows's post on Michael Crichton, I was reminded of an article Fallows wrote about an air taxi company called DayJet last spring in the Atlantic, "Taxis in the Sky". One of the fascinating aspects of DayJet that Fallows covered was the way the company decided what routes to fly and how much to charge. Below are a couple of brief excerpts.

Their computer models resemble a much more complex version of an “artificial life” computerized game like SimCity or SimLife—or, to explain the nickname they gave themselves, programs that simulate the paths a colony of ants will take across a floor as they discover and retrieve pieces of food. This process is also known as “agent-based modeling.”

Fallows writes that the company's computer programmers are called "ant farmers" because of this. The photo above, from the article, shows some of the ant farmers' equations.

While the ant farmers tried to determine where the company should start its service, the mathematicians from Russia were devising the software on which the company would run. In the end, they came up with plans that, in their view, made DayJet conceptually closer to Google or eBay than to existing airline companies. Before I explain what I heard from Alex Khmelnitsky, a note: most of the time I spend reporting, I spend listening to people describe what they do. It’s the payoff of the job—and through years of transcribing notes, I’ve learned that the typical minute or two or 10 of conversation boils down to a sentence or so of usable thought.

Not so with Alex! He and his colleague Eugene Taits worked for Iacobucci at Citrix and joined DayJet in its early days; another Russian veteran of Citrix, Oleg Kuzedub, joined DayJet recently. During the hour I spent with Alex and Oleg (Taits was out of town), Oleg said only a few words. But Alex more than made up—and I have never listened so hard trying to comprehend.

To spare readers, not to mention myself, undue exertion, I’ll simply say that in the view of everyone I spoke with at the company, the Russian-designed mathematical backbone of the company was its crucial advantage over any competitors, and would mean the difference between the company’s making a profit and not.

No comments: