The main objective of this paper is to consider the problem of a system amenable to maintenance and to provide to posterior analysis, Bayesian point estimators and Bayesian availability analysis of a k− out of −m system with geometric failure as well as repair time distribution.
The study considers a Bayesian approach, treats these population models as random quantities and makes good use of old information to construct a prior distribution model for these parameters, and then make use of current data to revise this starting assessment in the form of a posterior distribution model for the population model parameters, while the primary motivation to use a Bayesian reliability method is typically a desire to save on test time and materials cost.
The study clearly demonstrates that, when inspections are performed at specific intervals, time is not continuous and is measured on a discrete scale. It considers the number of successful cycles or operations before failure, then the repair process helps to improve the system reliability.
The proposed methodology represents an efficient way to evaluate the maintenance performance when time is not continuous and measured on a discrete scale.
