A novel approach to model traffic on road segments of large-scale urban road networks

The study proposes a novel method for modeling traffic volumes at the road segment level of large-scale urban road networks. This study has been placed in a milieu where existing methods on modeling vehicular traffic volume are hampered by data and cost constraints, especially in developing countries. Emerging traffic modeling methods, based on centrality and space syntax provides a technically-efficient approach to overcome the above-mentioned constraints. Nevertheless, those methods are yet to be popular among practitioners due to limited accuracy and validity. This study modifies the existing methods and validates in five case cities to make them practice-ready. Findings of this study indicated that the proposed method is competent enough to estimate traffic volume of road segments on a par with the internationally accepted standards. The proposed method combines two network centrality measures abstracting the traffic volume on a road segment as the sum of origin-destination trips (i.e., Closeness-Centrality) and pass-by trips (i.e., Betweenness-Centrality). The study modifies the 'distance' variable in the existing formula as 'path-distance' which captures topological and mobility characteristics of roads. The method does not require extensive data and can be implemented by utilizing publicly available open-source network analysis software, hence, ideal for resource-scarce situations. (C) 2019 The Author(s). Published by Elsevier B.V.