Optimum Decision Policy for Management of Pavement Maintenance and Rehabilitation

Abstract

An effective practical decision policy has been developed for use in the selection of an optimum maintenance and rehabilitation program. Its main objective is the optimization of pavement condition under constrained budgets. The developed policy utilizes a discrete-time Markovian model with five condition states labeled a, b, c, d, and f. State a represents pavements in excellent condition, and State f indicates pavements in bad condition. Several decision options have been introduced based on either maximizing the proportion of “good” pavements or minimizing the proportion of “bad” pavements. State probabilities at some desired future time have been used as the main objective functions in the development of optimum maintenance and rehabilitation programs. The unknown variables in these programs are those representing improvements to pavement condition through implementation of maintenance and rehabilitation work. The resulting optimum programs are nonlinear in form, and therefore the penalty function method with functional evaluations has been successfully used to yield optimum solutions. The optimum solution to a particular program defines the type and extent of maintenance and rehabilitation work required for annual or biennial implementation. Pavement maintenance is mainly defined as routine maintenance consisting of filling cracks, patching potholes, and other applicable techniques such as chip seal coat or slurry seal. Pavement rehabilitation is defined as major rehabilitation actions to include resurfacing (overlay), resurfacing with partial reconstruction (localized reconstruction), and complete reconstruction applied to pavements in States c, d, and f, respectively.