Developing a practical highway criticality assessment framework based on network connectivity and redundancy

Abstract

Network criticality assessment aims to identify important nodes or links within a network that that are critical to its overall performance. Existing studies often focus on factors such as traffic volume and sociodemographic attributes, giving preference to heavily traveled highways in densely populated areas. Rural highways, particularly those with low traffic volumes, tend to be undervalued, despite their potential significance in sustaining network connectivity and facilitating access to services and opportunities within local communities. This paper addresses this concern by developing a criticality framework grounded in the egalitarian principle. The framework comprises two complimentary measures: normalized betweenness centrality and detour importance. The former measures a road’s relative importance in efficiently linking the local network by quantifying the percentage of origin-destination pairs within the neighborhood that utilize it. The latter, detour importance, reflects the aggregated percentage of trips rerouted to the road during disruptions to other critical roads, thereby highlighting its role in providing the needed redundancy of the network. Given the fundamental role of finding realistic shortest paths in this analysis, a large-scale GPS-based probe vehicle dataset is integrated to provide real-world travel time information. The results from the Kentucky case study validate the effectiveness of the developed framework. The framework utilizes efficient algorithms and readily available data, making it practically feasible for large-scale applications.