Uncertainty assessment for availability: importance measures

This paper provides an uncertainty assessment for the availability of a two-state repairable system. During the design stage of a system it is very important to allocate scarce testing resources efficiently. Although there are a variety of importance measures for the reliability of a system there are limited measures for the availability of systems. Our study attempts to fill the gaps on availability performance measures and provide insight for techniques to reduce the variance of a system-level availability estimate. Our technique utilizes the delta-method in order to provide an estimate of the variance for the systems availability. The variance importance measure developed provides a measure of the improvement in the variance of the system availability estimate through the reduction of the variance of the components availability estimates. In addition, a cost model is developed that utilizes the cost of unit variance reduction in the importance measure. The variance importance measures for several series-parallel systems are developed and the results are consistent with the availability importance measures developed by Cassady et al. as well as available reliability importance measures