A new index of component importance

Several indices of component importance have been proposed in the literature. The index most relevant to system reliability is the Birnbaum importance, but a comparison between two components is hard to prove and even hard to verify by computation. In this paper we propose a new index which is stronger than the Birnbaum comparison, but always verifiable by computing. In many cases, it could even be easier to prove. We also show the relation of this index with other indices.