Measurement Uncertainty in Decision-Making

Abstract

In this paper, the authors examine a common issue concerning the influence of measurement uncertainty on decisions. In fact, in some practical applications, it can be necessary to put in comparison measurement data with thresholds and limits. It occurs when the conformity with fixed specifications has to be verified or if warning and alert levels have to be not exceeded. In such a circumstance, to take reliable decisions in presence of uncertainty is a concrete problem. Measurement uncertainty may reasonably be the cause of unreliable decisions. In order to manage properly the uncertainty effect, the authors have developed a decision making procedure based on a methodical approach to measurement uncertainty. In detail, a fuzzy logic algorithm estimates the probability to take a wrong decision because of the uncertainty. Such information is so used in order to optimize the decisional criteria, improving the consistency of the final computing results. Risks and costs associated to the possibility to take a mistaken decision are minimized. Consequently the algorithm singles out the most reliable decision.