Researchers at the Massachusetts Institute of Technology, Cornell University, and the Italian National Research Council's Institute for Informatics and Telematics have developed a technique for analyzing taxi trip data and organizing ride sharing in real time.
Many see the next logical step of app-focused car services such as Uber being dynamic ride-sharing apps that can match customers going in similar directions, thus shortening wait times, reducing fare costs, and lowering overall carbon emissions. To test this idea, the researchers analyzed about 150 million trip records gathered from 13,000 New York city taxis over the course of a year.
They found about 95 percent of the trips could have been shared and an optimal combination of trips would have cut total travel time by 40 percent. Even accounting for the realistic limitations that only trips starting within a few minutes of each other could be shared, researchers still found total trip time could be reduced 32 percent.
The study also described a method for carrying out real-time trip matching that can handle as many as 100,000 trips in a tenth of a second.
In addition, the researchers created HubCab, a Web application that enables people to explore the data using a map of New York as an interface.
From MIT News
View Full Article
Abstracts Copyright © 2014 Information Inc., Bethesda, Maryland, USA