Hiding in 172M Cab Rides: A New Frontier for the ‘Sharing Economy’
Splitting a cab fare isn’t usually what you’d call a high-tech activity, even with millions of people carrying sophisticated Internet portals in their pockets and purses. Researchers at MIT think it’s about time that changed.
By analyzing a year’s worth of New York City cab rides, MIT’s Sensable City Lab was able to see just how many taxis passed by another passenger who was headed in the same direction at the same time.
As you might have guessed, there were a lot of missed connections. So many, in fact, that researchers believe a digital dispatch system linked to passenger smartphones could significantly reduce the number of cab trips in the city, lab fellow Michael Szell said.
“It looks like if people are willing to prolong their trips up to three minutes, then we can really save 40 percent of the trips,” he said. “And this is just if two people share a cab.”
The MIT researchers are now looking for more taxi data from cities around the globe, including Singapore, Vienna, and Boston. “We are also interested if this is a general finding that applies to all kinds of cities, or is it just Manhattan-specific?” Szell said.
In the meantime, digital startups are trying to tackle the ride-sharing problem from the ground up.
New York-based Bandwagon, which launched last year, already lets users in its home city pair up for rides around town or cab fares home from the airport. Other companies, like California-based Lyft and Sidecar, were conceived as ways of letting non-professional drivers pick up digital hitchhikers.
All of these efforts hint at a not-so-distant future where a few taps on a smartphone screen makes getting around the city a bit easier and cheaper, not to mention more efficient—cutting down on carbon emissions and soul-crushing gridlock.
The MIT research started by collecting data on all of the taxi rides performed in New York for 2011. Szell was able to get the information by filing a freedom-of-information request with the city taxi authority, which gave him plenty to work with: 172 million trips by 13,586 registered cabs, with the unique vehicle identification, GPS coordinates of the pickup and drop-off locations, and the times associated with the trips.
To help make sense of all that data, the researchers built a visualization tool that shows the possible nearby matches for any cab ride that occurred that year. The public version, called HubCab, lets anybody play with the pickup and drop-off locations of cab rides around the city to see how many matches might have been made.
A trip from Central Park to the Lower East Side, for example, was made about 1,000 times that year. The interactive tool says cab riders could have saved more than $25,000 in fares and 9,000 miles driven if the right sharing matches were made.
Of course, not everybody wants to share a seat. “It’s not enough to have the incentive of splitting a fare. Some people just don’t want to ride with a stranger. And that’s completely OK,” Szell said.
But the MIT group says there are still clear benefits, even if you simulate a lot of people declining to share cab rides. Their data model indicates that the maximum number of shareable rides could be filled pretty fast—if a quarter of the roughly 400,000 daily cab rides were splitting fares, it might be enough to capture nearly all of the possible ride-shares.
“So you have high returns on investment, in business language,” Szell chuckled.