Making Better Maintenance Decisions With Statistics

The natural lives of machines and components vary tremendously! Maintenance managers struggle constantly to maximize their equipment lives while minimizing their maintenance costs, but they normally lack the tools and techniques they need to do this properly. This paper proposes a methodology based on probabilities of failure and expected values for repairs that recognizes the variation among individual machines and components as well as the high costs associated with catastrophic failures. It enables a user to determine the optimum change-out intervals for many types of equipment. The approach proposed here is sophisticated enough to be reliable but simple enough to be applied by the maintenance staff on the shop floor. It will complement existing preventive and predictive maintenance programs.