Using Truck Probe GPS Data to Identify and Rank Roadway Bottlenecks

Abstract

This paper describes the development of a systematic methodology for identifying and ranking bottlenecks using probe data collected by commercial global positioning system fleet management devices mounted on trucks. These data are processed in a geographic information system and assigned to a roadway network to provide performance measures for individual segments. The authors hypothesized that truck speed distributions on these segments can be represented by either a unimodal or bimodal probability density function and proposed a new reliability measure for evaluating roadway performance. Travel performance was classified into three categories: unreliable, reliably fast, and reliably slow. A mixture of two Gaussian distributions was identified as the best fit for the overall distribution of truck speed data. Roadway bottlenecks were ranked on the basis of both the reliability and congestion measurements. The method was used to evaluate the performance of Washington state roadway segments, and proved efficient at identifying and ranking truck bottlenecks.