Development and Implementation of a Multi-Level Roadway Segmentation Methodology

Abstract

The reliability and applicability of traffic operation analyses depends on their ability to integrate relevant input from disparate databases in a seamless and automated manner. Inputs include information on road geometry, traffic composition, and spatial referencing. These databases are collected by different agencies for different purposes. As a result, a common definition of roadway segments is lacking across various applications. This paper developed a systematic segmentation methodology that considers the needs of various operational and planning studies. A multi-level dynamic segmentation approach has been developed to address different levels of requirements for various studies: at the micro level, referring to the smallest roadway segmentation for traffic simulation studies; at the meso level, representing a combination of several micro segments for traffic operation studies; and, at the macro level, corresponding to planning studies. In this paper, the proposed methodology for the segmentation of freeway and arterial corridors in Ontario (Canada) is demonstrated. At each level, several criteria were selected to identify the locations where the roadway network needs stringent analysis. Next, a pilot study was designed to evaluate the proposed methodology. It was found that the new segmentation methodology can successfully identify areas of congestion and queue growth/dissipation. Finally, the proposed segmentation methodology was implemented for more than 6000 km of Ontario’s roadway network. The results of this study can assist researchers and road agencies with defining a systematic roadway segmentation that can be utilized for different types of projects, ranging from traffic operation to planning studies.