Preventive and Predictive Maintenance Predictive maintenance aims to predict the occurrence of problems so that action can be taken to avert downtime. The most common usage of predictive maintenance is with wear parts. Variations in the key dimensions of wear parts are measured and when they go below a predefined level, the part is replaced. In this way downtime and poor machine performance is averted.

There are two stages to predictive maintenance: first identify a characteristic that varies as a component nears its end-of-life and second, define the level at which action should be taken. For example when a bulb is nearing its end of life it may gradually reduce its operating temperature before complete failure. By understanding the failure mechanism and the characteristic associated with its failure, the breakdown can be predicted. The temperature of the bulb at a defined distance is measured at regular intervals and logged. And temperature variations associated with failure are noted. A procedure is set up such that any recurrence of these temperature variations with future bulbs results in the earliest possible replacement of the bulb.

There are 5 steps to effective predictive maintenance:

Step 1: List the significant functions that constitute the machine: For example;