Decision-Making Problem for Managing Pavement Maintenance and Rehabilitation Projects

Abstract

A typical pavement management system (PMS) should help a decision maker select the best maintenance program—that is, which maintenance treatment to use and where and when to apply it—so that maximal use is made of available resources. To provide decision support, a PMS needs a process for setting maintenance and rehabilitation (M&R) priorities. The quality of the prioritization directly influences the effectiveness of available resources, which in most cases is the primary goal of the decision maker. A network-level project selection procedure that uses an operations research optimization method was developed. An integer linear programming model was established to select a set of candidate projects from the highway network over a planning horizon of 5 years. In the model, constraints of available annual budgets and minimum requirements on pavement conditions were satisfied where state transition of each road section was formulated. Two optimization objectives—maximization of the total M&R effectiveness and minimization of the total M&R disturbance cost— were pursued over the planning horizon. A small network of 10 road sections was used as a pilot study for the established model. Pavement management functionalities at the network level are summarized, formulation of the integer linear programming model for the project selection problem is presented, differences in decision makers' value systems and the resulting different undominated solutions are discussed, and a sensitivity analysis of the annual budget constraint is conducted.