Predictive maintenance is a series of actions that are taken, and techniques that are applied, with the aim of detecting possible failures and defects of machinery in the incipient stages, to prevent these failures from manifesting into a larger one during operation, preventing them from causing emergency stoppages and downtime, causing negative financial impact. Its mission is to maintain a certain level of service on the equipment by scheduling reviews at the most opportune time. It usually has a systematic nature, that is, intervention is carried out even if the equipment has not shown any signs of having problems.

The requirement for a predictive technique to be applied is that the incipient failure generates signs or symptoms of its existence, such as high temperature, noise, ultrasound, vibration, wear particles and high amperage, among others.

Techniques for detecting faults and defects in machinery vary from the use of human senses (hearing, sight, touch and smell), to the use of process control and quality control data, the use of statistical tools and fashionable techniques such as vibration analysis, thermography, tribology, motor circuit analysis and ultrasound.

The ISO 17359 standard establishes general guidelines for machine monitoring and diagnosis. It defines parameters such as vibrations, temperature, lubrication and inspection techniques, providing a standardized framework for the implementation of predictive maintenance programs.[1]​.

For a predictive maintenance program to be considered effective, it must increase the reliability and operational status of the machinery while at the same time reducing production costs, including maintenance costs.

To design and incorporate an effective predictive maintenance program, it is necessary to determine the equipment that will be used in this maintenance, as well as the machines and processes that justify the implementation of the program both technically and economically.

To achieve this you require:

Advantages:.

Disadvantages:

The minimum steps necessary prior to the implementation of predictive maintenance are:

They consist of the following: