By Lina Sorg
Anyone that has ever been in a traffic jam where cars slow to a crawl has experienced the frustrating limitations of our traffic system. While road congestion might seem inevitable, math may have an answer to the assumedly interminable delays caused by traffic lights.
The flow of traffic on crowded roadways is part of an intricate network with variable, nonlinear dynamics. Although predicting and managing such dynamics is complicated, traffic lights have been directing vehicles through intersections for about 150 years.
Traffic lights operate via a systematic phased switching process, based on traffic patterns and time and space restrictions. Vehicle transition between the “stop” and “go” phase is interrupted by the transient “setup phase,” marked by a yellow light. The frequent transitions and short phases enable traffic to flow fairly well in both directions, but the length of time and frequency of the “setup phase” significantly reduces intersection throughput and efficiency. Such traffic light limitations cause congestion, delays, and heightened pollution in busy areas.
Researchers at the Massachusetts Institute of Technology (MIT) are pursuing a novel approach to traffic in which vehicles move fluidly through intersections without the help of traffic lights. In the proposed system, a mathematical model allows high-tech “intelligent” vehicles to communicate with other vehicles and relevant infrastructure using sensors; this communication enables cars to maintain a safe tailgating and stopping distance from each other as they navigate a standard four-way intersection.
The model stems from a slot-based control system, similar to those that control air traffic. The system employs a scheduling algorithm to control vehicle speed. A car communicates its intended trajectory to the algorithm, which groups it into a small cluster of nearby cars also traveling in the same direction. Then, cruise control-type software (similar to the autocruise feature available in most cars) would manage the speed of those grouped cars – and cars approaching the intersection from the three other directions – so they can cross the intersection in their assigned slots, and without stopping. Because the algorithm aims to have vehicles in the intersection for the shortest possible time, speed regulation takes place well before the vehicles reach the intersection. Read a more detailed explanation here.
While slot-based intersections (SIs) are not a new idea, there was previously no framework to measure their function against that of traffic lights. According to Carlo Ratti, co-author of the study and director of the SENSEable City Lab in the Department of Urban Studies and Planning at MIT, eliminating traffic lights in favor of advanced vehicle technology ensures a more efficient system because control exists at the level of vehicles, rather than that of traffic flow. The algorithm can then ensure that the vehicles reach the intersection at an opportune time, with safety as the most important parameter. Read more about the authors’ thoughts here.
In highly-trafficked areas, increased velocity does not necessarily correlate with increased efficiency. Instead, a steady, moderate speed keeps vehicles moving consistently, producing a smoother flow of traffic. The researchers refer to this outcome as the “slower is faster” effect. This effect applies to many situations involving pedestrian motion, such as boarding an airplane. When everyone quickly crowds around the entrance, the crowding results in a slow-moving bottleneck. But when they board in steadily-moving smaller groups, overall traffic moves more quickly.
While today’s auto technology can implement features of the slot-based system, it is bound to work even more seamlessly with intelligent, autonomous cars of the future. This type of transportation infrastructure will thus adapt quickly to futuristic smart cars.
The researchers recognize that there is still work to be done with their model. In urban environments where traffic lights are quite close to one another, the dynamics of one intersection would quickly distend throughout the city. Nevertheless, SIs have the potential to double traffic capacity and minimize delays. This exploratory framework can further the efforts of engineers and developers working on intelligent infrastructure, and minimize some of the social and environmental costs associated with traffic.
A corresponding paper published in this month’s issue of PLOS One.