Reliability Cost Estimation: Managerial Perspectives

This paper deals with selected sources of uncertainties associated with the reliability related life cycle costs (LCC). Some of the factors responsible for uncertainty in reliability cost estimation are identified and discussed. A practical approach to the use of Beta distribution in estimating the reliability-related costs within the LCC framework is detailed, and is illustrated using a hypothetical example. Improved estimation of reliability related LCC requires better understanding of sources of associated uncertainties and the methods of dealing explicitly with them. Numerous distributions can be used for modeling of cost variates besides the Beta demonstrated in this paper. The 3-parameter Weibull and Gamma distributions are two examples. Whichever distribution is chosen, given a sufficiently large cost data base, it is essential to test the cost variate for goodness of fit and to validate the model by appropriate test of hypothesis. However, in estimation of costs involving large capital expenditure over a period of years, the data base is not likely to be large. Lacking, then, a statistically significant sample which would allow testing for the underlying probability distribution governing the cost variate, modeling via the Beta or other appropriate distribution presents a practical alternative.