Abstract
A secure supply of raw materials plays a vital role in the rapid growth of emerging technologies. The criticality assessment has been introduced to evaluate the supply risk of materials from economic, social and environmental aspects. From business perspective, multiple critical metrics should be involved, so that the decision makers can focus on the different metrics and put efforts to minimize the corresponding risks. However, due to the complexity of assessment and uncertainties in data sources, some metrics cannot be evaluated quantitatively, but qualitatively. This paper introduces a fuzzy linguistic approach to evaluate multiple risk metrics for material criticality assessment. The risk levels and the importance weight of metrics, expressed in linguistic terms, are modeled by triangular or trapezoid membership functions. We apply this method to evaluate the criticality of three materials: Cobalt, Tungsten, and Yttrium. We use matrix operation to aggregate the MFs of the multiple metrics in order to represent the overall criticality. As a result, the three materials are ranked according to their critical levels. The proposed fuzzy linguistic approach shows the advantage to evaluate the criticality with multi-criteria when only qualitative data is available. The membership function is an appropriate way to represent linguistic terms with imprecision, which are commonly used to interpret risk terms. The definition of multiple metrics provides flexibility for the users to choose risk categories they are interested in, and the aggregation of the metrics supports them to compare the criticality of materials. Further study may focus on how to justify importance weight judgments to make tradeoff decision, or integrate temporal factor to predict the future critical risk for business purpose.