Data collected from 26,000 sensors buried under the pavement of California freeways is stored in the Performance Monitoring System (PeMS) database at the University of California at Berkeley. The data provide information about the number of cars driving on a certain stretch of freeway and their speeds. This article both reviews how PeMS data are used to study and analyze traffic congestion and bottlenecks, as well as the impacts that ramp metering, travel time advice, and high occupancy vehicle (HOV) lanes have on congestion. The article relates how the analysis of PeMS data has helped to disprove certain long-held beliefs about traffic congestion. In the past, traffic experts believed that an average speed of 40 to 45 mph maximized traffic capacity. The new data show that maximum capacity actually occurs at about 60 mph. In addition, the effect of high occupancy vehicle (HOV) lanes may have the opposite of their intended effect and actually increase congestion.
Samenvatting