Optimization Modeling for Enhanced Network-Level Pavement Management System

Abstract

A new formulation of a network-level pavement management system (PMS) model is described, which includes the identification of specific network links in the optimization. This model uses Markov transition probabilities for pavement condition modeling, as have several previously developed PMS models. The incorporation of specific links into the model, however, allows much easier translation of network-level policies to project-level decisions than has been possible. This formulation also allows the easy incorporation of user costs as well as agency costs and can include a variety of specific constraints on the solution. A case study application to Nassau County, New York, illustrates the use of the model and allows a comparison of the model solution with current maintenance practices in the New York State Department of Transportation.