Stochastic analysis of maintenance process costs in the IT industry: a case study

In this contribution we propose a method for determining the type of computer component whose replacement minimizes repair costs depending on the warranty period. Components that most often cause a dysfunctional state of a computer are first identified using records of warranty repairs from the customer service department. Repair costs covered by warranty are then calculated using a semi-analytic method and simulation of the most critical component lifetime in different user environments. It was assumed that the use of a cheaper critical component generally results in higher costs for computer warranty services due to shorter failure-free runs. For this reason, the company should choose a compromise between price and quality. Input data was taken from the records supplied by the computer assembling company (times between failures) and from the records from time tracking software (computer up-time).