Modeling Construction Project-Induced Freight Reliability for South Carolina’s Interstate System

Abstract

This paper focuses on the freight component of the federal rulemaking for performance management that assesses freight travel time reliability on the interstate system. As this is the first time agencies (departments of transportations and metropolitan planning organizations) have been tasked with establishing targets for freight reliability (and other measures), the paper proposes a modeling approach to systematically account for the traffic and roadway characteristics and predict possible changes in the truck travel time reliability (TTTR) ratio due to construction projects. Although different trendlines can be utilized to provide a range of TTTR indices for future years, the subjective nature of predicting a target value in a future year demands a more systematic approach to capturing the effect of construction projects on the variability of the TTTR ratio. This study uses South Carolina’s interstate system to develop the two generalized linear models (GLMs) and a Bayesian GLM. Length and type of traffic message channel, type of construction project, area type, average annual daily traffic, and number of lanes were found to be among the significant predictors associated with a change in the TTTR ratio as a result of construction activity. This paper discusses the association of these causal factors and their influence on freight reliability. Owing to the contemporary nature of TTTR, there is no existing research predicting this particular type of freight reliability; as such, this work is thought to be relevant and beneficial for transportation engineers, planners and forecasters.