Citation

Lambe TA. Oper. Res. Q. 1974; 25(1): 99-110.

Copyright

(Copyright © 1974, Operational Research Society, Publisher Pergamon Press)

DOI

unavailable

PMID

unavailable

Abstract

A simple algebraic formula for combining repair data with prior experience determines the time when a machine should be replaced in order to minimize the expected cost of equipment purchase and maintenance. A random sample from an exponential distribution represents the cost of each repair, and a time-dependent Poisson process represents the intervals between repairs. Bayes' formula provides the basis for combining data with previous judgement about the characteristics of the equipment. Automobile maintenance records support the basic assumptions of the model, and illustrate the method of deciding when to scrap a given machine.