Skip to content

Beating the traffic

iStock image

While on sabbatical, Javed Aslam, a professor in the College of Computer and Information Science, is collaborating with MIT researchers on a project in Singapore, studying traffic flows through data from taxi cabs to reveal congestion patterns and determine how to make city driving a less painful experience.

Tell us about the project, and where you come in.

The Future Urban Mobility project is part of a larger Singapore-MIT Alliance for Research and Technology enterprise.  Singapore wants to figure out how best to get people from place to place. They want to develop, in and beyond Singapore, new models for the planning, design and operation of future urban transportation systems, including public and private transport, land use, environmental impacts, urban planning, and transport-related policy issues.

I signed on as a senior research associate. My wife, Daniela Rus, an MIT professor in the Department of Electrical Engineering and Computer Science, signed on as the first principal investigator in-residence. Our goal was to kick-start all of the research activities for the project, in collaboration with students, post-docs, researchers, and faculty in Singapore.

Why study Singapore’s traffic as opposed to any other big city?

Singapore has a tremendous amount of data available with which to study traffic. For example, as part of our study, the taxi company Comfort Delgro provided us with data for August 2010. The data included roughly 16,000 taxis logging their identification numbers, GPS locations, speeds, status, and date and timestamps every 30 seconds. In essence, I have a roving network of 16,000 taxis, and I know exactly how fast they’re going at any point in time all day for a month, in roughly one-minute intervals.  These 16,000 taxis effectively comprise an enormous roving “sensor network” with which to study traffic, congestion, mobility patterns and so on. We are in the process of obtaining similar data from the local bus system, as well as static traffic sensor data from over 1,000 road intersections in Singapore.

Based on your experience studying traffic patterns and congestion in Singapore, what advice can you give city drivers on how best to reach their destinations on time?

One seemingly counterintuitive piece of advice is that the fastest route isn’t always the best. While most route planning tools will suggest the fastest route based on historical data or expected travel time, if you have a hard deadline to meet, it may well pay to take a somewhat longer route that is more consistent, in that it is less likely to be unusually delayed.  Such routes may actually increase your probability of reaching your destination on time.