Optimal Sampling of Infrastructure Condition: Motivation, Formulation, and Evaluation

Abstract

Infrastructure management is the process through which inspection, maintenance, and rehabilitation decisions are made to minimize total life-cycle cost. Condition measurement, spatial sampling, and forecasting introduce uncertainty into the process. The first and third uncertainties are captured in the infrastructure management literature. However, the second has not been recognized and quantified. This study focuses on motivating the incorporation of the spatial sampling uncertainty in decision-making, quantifying this uncertainty, and investigating its effect on the expected minimum life-cycle cost. Evaluation measures are defined to assess the contributions of the developed framework, namely incorporating uncertainty due to spatial sampling in the decision-making process and selecting the sample size optimally. A numerical analysis indicates that in general both contributions are important under a wide range of realistic scenarios and the implications of the combined effects on minimizing life-cycle cost and long-term budget planning are marked.