Integrating our disconnected traffic systems — from traffic lights to personal on-board navigation — could reduce urban congestion and even air pollution, according to Liviu Iftode, a computer science professor at Rutgers University. Iftode, in collaboration with Rutgers colleagues and Mario Gerla of UCLA and his colleagues, recently received a three-year, $1.94 million grant, to study how wireless technology could improve traffic control and air quality. Below are excerpts from my recent interview with Iftode.
Explain your technology and why it’s needed.
Urban congestion and air pollution are two big problems for cities today. We’re trying to use wireless technology to connect drivers to traffic centers where this information is available. [It would use] on-board navigation. At this point, there is technology like this to control or reduce traffic in urban settings. But most of these technologies are disconnected. The traffic center collects the information continuously. They can also control the traffic lights. But at the end of this chain are the drivers who have on-board navigators. They might have information about traffic, but they likely have no information about pollution. We’re trying to explore whether wireless technologies can be used to upgrade the communication between the traffic center and the driver.
On one hand, the municipality’s goal is to reduce air pollution, to improve the city walkability. From the driver’s perspective, the important thing is to reach a destination as soon as possible. Part of this is to create incentives and see whether they’re working. [Another] issue is how to make this as non-intrusive as possible for the driver. If the driver has to constantly inspect this information about air pollution and congestion and make a decision based on that, it’s not going to work. We’re looking at how to improve this interface between the driver’s on-board navigation, so most of the time the on-board navigation can choose the best routes.
There’s no connection between our current traffic systems?
They’re disconnected. There’s virtually no way for the traffic centers to suggest alternative routes to drivers based on their destinations. There’s no way to incentivize drivers to take those suggestions. Today, as you know, there are on-board navigation systems that get traffic information. You might have information on certain roads, but not on every road. They are not frequently updated. Therefore, choosing alternative routes may often put the driver in even higher congestion.
Congestion and pollution are related, but not exclusively related. There are other environmental factors — air flow, the layout of buildings on the street — that make air quality difficult to model. The municipality might be interested in returning the city to its citizens. Therefore, it can decide to move the traffic from a certain street to make it more livable.
In what locales would this be implemented? Where would it be most useful?
New York, Los Angeles and other large metropolitan areas where traffic is a problem and pollution is high would be the best candidates. But that’s a long shot. At this point, we’re still in an exploratory phase. We’re trying to understand the problem. It’s a huge problem in terms of modeling pollution and simulating traffic. Our long-term goal is to have impact on the design of future traffic control systems in the big cities.
Are you targeting a certain type of driver? Is this best for the daily suburb-to-city commuter?
This is another complicated problem. It’s about what we call incremental deployments. Who is the driver that would adopt this sooner and how can we incentivize drivers to adopt this technique? The traditional incentive model for controlling driver behavior with respect to congestion is the congestion fee. You impose a certain fee if drivers take a route through a congested area.
We’re looking at creating a positive incentive, instead of a negative incentive. A driver could get a certain discount. The daily commuter might be interested in benefiting from this over time. It could be significant. Modeling the incentives to figure out which ones work is another part of our work.
What stage of the project are you in now?
The first step is to fully understand the problem and connect these pieces together. This is an interdisciplinary effort. At a later time, we’ll start to build models and simulate those models. We’ll start looking into building the technology components.
One piece that is critical for our project is the traffic light. Traffic lights have been around for 150 years and have not changed much. They provide simple visual information. Traffic lights are a good place to collect and send air quality data. They can be used to get information from the cars to pass onto the traffic centers or to broadcast information from traffic centers to cars in that proximity. We’re looking at whether we can upgrade the traffic light. Cars have already incorporated a lot of computer technology. The traffic light was left behind.
It’s a long-term project. It’s impossible to complete it within a time frame of three years. This will take much longer.
Photo: Liviu Iftode