Optimal Spatial Sampling of Infrastructure Condition

Infrastructure management is the process through which inspection, maintenance, and rehabilitation decisions are made with the aim of minimizing the total life-cycle cost. The inputs to decision making include current facility condition as assessed from field observations, and future facility condition as forecasted based on the current condition and possible maintenance and rehabilitation (M&R) actions using a deterioration model. The quality of the estimated current condition depends on the accuracy of the inspection technology, the spatial sampling employed, and the nature of correlation between condition variables and different locations. The uncertainty associated with spatial sampling has not been recognized and quantified in the infrastructure management literature. However, its impact on arriving at the optimal M&R decisions can be appreciable. This paper motivates the importance of sampling, focuses on quantifying the spatial sampling uncertainty, and investigates the effect of incorporating spatial sampling as a decision variable on the expected minimum total life-cycle cost.