What is Predictive Maintenance?

Maintenance costs are one of the largest factors impacting a plant’s budget. Attempts to reduce these costs have led to the development of several maintenance strategies. Predictive maintenance is a defect inspection strategy that uses indicators to prepare for future problems.

One major mistake made with predictive maintenance is that it is often confused with preventive maintenance, which consists of preplanned, routine measures used to prevent major problems. To better understand predictive maintenance, it is best to draw a contrast between it and preventive maintenance, which is often a more widely recognized term.

Preventive maintenance involves regular, prescheduled service. For example, a machine may be shut down every month to have fluids replaced and refilled. Then, every six months, it may be shut down so its belts can be replaced. This is done in hopes that major problems associated with fluids and belts do not occur.

Predictive maintenance, to the contrary, generally involves observation. Notes are taken of indicators that may signal larger problems. For example, a machine may be checked regularly. If abnormal fluid leakage or moisture accumulation is found, these act as signals that some larger problem may be growing.

Another major difference between these two maintenance strategies is that preventive maintenance may be possible while a piece of equipment is in operation. However, in most cases, the equipment must be shut down for a period. Predictive maintenance almost always involves assessments that are taken while equipment is functioning normally. It will generally render the inspection useless or inaccurate if a machine is assessed for defects while it is shut down or its capacity is reduced.

Predictive maintenance does not rely on hunches and intuition. It does not rely on industry statistics, such as those that say a piece of equipment must have a specific service at specified intervals. Predictive maintenance relies on real signals demonstrated by a single and specific piece of equipment.

This can include comparing statistics for indications of an impending problem. It is commonly noted that one of the most costly mistakes made in the industrial setting is not recording and analyzing data such as equipment performance and heat distribution. The figures that represent an optimally operating piece of equipment are excellent indicators when compared to figures that are drastically different. If a box cutter cut 30,000 boxes every month for a year and then suddenly can only cut 20,000, this should be regarded as a signal.