A PROBABIUSTIC APPROACH TO PREVENTIVE MAINTENANCE

Sometimes it is impossible to know beforehand when particular machine parts may break down. If every break-down could bring about a grave damage, it would seem unwise to use such machine parts over the whole span of their natural life. Throwing away the long-lived parts before their lives expire and replace them by new ones will usually cut down production cost. Some typical models of preventive maintenance are showm in Fig. 1. In this paper, assuming the life of machine parts to be stochastic variables, some formulae for determining the optimal time of parts replacement in order to minimize the total cost are discussed from the stand-point of probability theory, and graphs for practical use are also prepared. Notations used in this paper are as follows: a == the price of a machine part. al == loss of labour and time in replacing a broken part by a new one. a2==10ss of labour and time in replacing a long-lived part by a