If you live in Boston, Los Angeles or any other major U.S. city, you know this fact: traffic is a nightmare. Sometimes it seems that traffic is all anyone talks about and each delayed meeting or event begins with a story about how bad it was.
The average commuter in the U.S. spends 42 hours in traffic per year. The cost of commuter delays has risen by 260 percent over the past 25 years and 28 percent of U.S. primary energy is now used in transportation. Road congestion is responsible for about 20% of fuel consumption in urban areas. According to one estimate, the cumulative cost of traffic congestion in the U.S. will reach $2.8 trillion by 2030. At the individual citizen level, traffic congestion cost $1,740 per driver during 2014. If unchecked, this number is expected to grow by more than 60 percent, to $2,900 annually, by 2030.
It’s a problem with a classic common and tragic root. No individual driver has an incentive to make changes that would make the entire system better. In other words, each driver seeks to make the best time or take the most convenient route, but no one is in charge of making the system work better as a whole. As a result, traffic just keeps getting worse.
But technology, which in the form of the automobile gave us this problem, may now offer up the faintest hope of a solution for this problem—that is, the global positioning system, the pervasive use of cell phones, and the advent of the self-driving vehicle could bring new solutions to this seemingly intractable problem.
Every vehicle on the road today, carries one or more connected devices, such as the cell phone of the driver, or the vehicle’s own cell or satellite connection. Using signals from these devices, engineers can now collect minute-by-minute traffic information on every road. Analyzing the data can reveal how individual drivers make routing decisions and, using this information, it becomes possible to build very accurate predictive models of traveling patterns and the resulting congestion. One can then leverage these models and optimize overall congestion. An important question is how to induce drivers to make route choices consistent with such a socially optimal solution.
Like many other economic and environmental problems, small changes can lead to big difference in outcomes. Apps such as Google Maps, Apple maps, and Waze. can be programmed to offer drivers choices that are more beneficial to overall traffic patterns, even if these routes are not as direct or convenient for an individual driver. In other words, the apps and the vehicles can be programmed to lead drivers to take routes that favor the overall good of society (minimizing overall congestion) and not their own preferred travel routes. The result could be a better overall traffic system for everyone.
As the world population continues to grow, more and more people are moving to urban areas resulting in growing congestion. One way to make cities more livable and sustainable to is attack the traffic problem head on. Trusting the apps, or even one’s self-driving vehicle, could offer a way to escape the tragedy of the commons caused by individual driver decisions and move to more efficient and socially optimal traffic patterns.
Ioannis Ch. Paschalidis is a former Visiting Professor at the MIT Sloan School of Management and Professor, Dept. of Electrical Engineering at Boston University.