Improving Online Maps, Maybe

By Edward Cone  |  Posted Friday, November 12, 2010 01:11 AM

by Tim Moran

There's a common belief that cabbies know how to get you from here to there the best way possible. They're hip to the shortcuts, the side streets, the traffic patterns, so they can whisk you to your destination in no time.

I am not one who holds to this belief. In fact, on more than one occasion when riding in a taxi in New York City, I've wondered what the heck my cab driver was doing and where he was going. And when heading to JFK from my home on the North Shore of Long Island, I, more often than not, have to tell the driver the best, quickest, and cleverest back-roady way to get to the LIE.

Be that as it may, many people believe in hack prowess, including researchers from Microsoft, who "are mining cabbies' knowledge to create faster driving paths for online maps."

The reality of Web maps is that "current drive-time predictions. . .rely on the length of road and the posted speed limit. Some services will inform drivers that the route takes longer in traffic, but that doesn't help someone who wants to know the fastest route from point A to point B, even if that route might look longer because it takes unexpected side streets."

So...Calling all cabbies!

The Microsoft researchers started their work in Beijing--although the concept should travel to other cities and countries easily enough--by analyzing GPS data from 33,000 Beijing taxis "in hopes of finding faster driving routes that would even be practical for people who don't drive at taxi speed or swerve recklessly between lanes," said Yu Zheng, a researcher at Microsoft Research Asia, who is an author of the paper describing the approach, called T-Drive.

The article maintains that the results of the project are pretty convincing: "The routes suggested by T-Drive are faster than 60 percent of the routes suggested by Google and Bing maps (which provide essentially the same driving time estimates as each other)." In other words, the cabbie-assisted routes offered can shave "about 16 percent off the time of a trip," which they calculate translates into about 5 minutes for every 30 minutes of driving.

Similar projects are underway, including: Nokia and the University of California (GPS data from drivers' cell phones); MIT's CarTel project (call phone data and taxi probes); and Waze, a Silicon Valley startup (social network driving-path sharing).

The one thing that was not mentioned in the article is the other commonly held notion that, sometimes, just sometime, cabbies have been known to deliberately take a route that is less than optimal to add some extra miles to the ride and dollars to the fare. Call it the SuperTramp Drive -- taking the long way home.