Estimation of parameters of failure distributions with fuzzy data

In this paper a simple procedure of estimation of parameters of two widely used failure distributions (the normal and Weibull distributions) and a method of finding confidence intervals for the parameters is developed. The computational burden of estimation with fuzzy data as observed by Kruse and Meyer (1988) is easily overcome by using special properties of the estimators. It is also demonstrated that the order statistics procedure of estimation can easily be modified to tackle fuzzy data.