Comparisons of Periodic and Random Replacement Policies

Abstract

For a large and complex system which consists of many kinds of units, we should make minimal repairs at its failures and make planned replacement or preventive maintenance at periodic times. This is called periodic replacement or maintenance with minimal repairs at failures which was summarized in (Barlow and Proschan 1965; Nakagawa 2005; Tadj et al. 2011). However, when the system is successively executing jobs or computer procedures in offices and industries, it would be impossible or impractical to maintain or replace it in a strict periodic fashion, because sudden suspension of the jobs may create losses of production in different degrees if there is no sufficient preparation in advance (Nakagawa 2005). By considering the factors of working times in operations, the reliability quantities of the random age replacement policy were obtained (Barlow and Proschan 1965). Several schedules of jobs that have random processing times were summarized (Pinedo 2012). The properties of replacement policies between two successive failed units, where the unit is replaced at random times, were investigated (Stadje 2003). Under the assumptions of random failure and maintenance, replacement and inspection with planned and random policies, were considered, and their comparisons were made (Nakagawa et al. 2011). Combining planned replacement with working times, the age and periodic replacement policies, where the unit is replaced at a planned time T and at the N th random working