Helping Drivers Avoid Lengthy Delays - Brief Article

USA Today (Society for the Advancement of Education), June, 2000

What the Weather Channel has brought to forecasting weather, John Leonard, an associate professor in the School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, hopes to bring to predicting traffic congestion. People plan their days according to the weather forecast, but they can't do that with traffic--at least not quite yet, he says.

Leonard is developing a cutting-edge traffic flow model that analyzes a variety of historical data and a host of variables to predict the next day's traffic conditions. He calls it a congestion "temperature." The model will also estimate current travel times from point to point.

"People need a simple-to-understand number--even if it doesn't have a physical meaning--to represent traffic congestion. We need to publicize it daily so people start to develop a personal understanding of congestion and plan accordingly." He is creating his model using the past two years' data from traffic surveillance systems and commonly used loop detectors on roadways.

One by-product of Leonard's research is a graphic representation of congestion. For each day of data, he is creating "star" diagrams that graphically show travel time estimates from various points of origin to any of the five points on the "star" created from a map of the city's major highways. Then, he will examine the data for trends and try to correlate them with variables such as special events, hotel occupancy, time of day, holidays, schools in session, weather, and vehicle accidents. An analysis of this information, combined with current traffic conditions, would yield a prediction of traffic congestion for the next 24 hours.

Leonard believes if people were to become familiar with a traffic "temperature," they would change their travel behavior and ultimately lessen traffic congestion. "Many people have the flexibility to plan their trips according to traffic congestion. With this sort of pre-trip information, people could decide whether to leave now or wait 10 minutes.... This could spread the peaks out a little and improve travel times for everybody." He sees a particular benefit for people who make a lot of discretionary trips and a benefit to couriers who might be able to reschedule or reroute deliveries.