Preface

Modern companies are increasingly trying to reduce costs while maintaining the quality of their services and the time spent producing them. This practice is associated with strict control of maintenance management. To this end, maintenance management information systems have been developed to scan each step of the processes, measure quantities, and plan interventions integrated into the production process to optimize production and maintenance procedures.

The main objective of maintaining maintenance processes is to ensure the quality and continuity of production processes. It should be done by examining all aspects of the manufacturing routine, identifying bottlenecks, and evaluating the criticality of each component or subsystem in the overall manufacturing process of the final product. This management should consider repair costs (or replacement) and miscellaneous expenses, planning the work effectively, with reduced repair times and clear instructions for the procedures to be performed.

Several computational tools for managing maintenance processes help the maintenance manager make more proactive decisions. These tools provide data and performance hints from the various processes, the compilation and analysis of data, trend reports, history of past maintenance, and history of similar equipment (or processes), among others.

Maintenance processes have undergone many changes over time. The first phase was simply corrective maintenance ("repair it when it broke"), in which a component is replaced when it breaks. This type of maintenance causes severe problems for production because the breaks usually occur during the production peaks, thus generating instabilities in quality. The second maintenance phase involves preventive maintenance (also called programmed maintenance), which makes exchanges of components after a specific time. There are two main types of preventive maintenance: time-based maintenance and usage-based maintenance. Time-based preventive maintenance is typically applied to equipment critical in the production process that, once damaged, severely impacts production. In usage-based maintenance, the exchange (or even only one inspection) is done based on production hours or cycles of production. This type of maintenance, which still occurs today in industries, also causes financial problems and unnecessary costs because parts and components can be exchanged in good condition, or unnecessary inspection actions can be performed. The advantage of this type of maintenance process over the previous one is that this process can be programmed, thus avoiding creating problems for production.

Then, with the cheapness of the measuring equipment, there was the possibility of better system observability being used. The next phase of the maintenance process is predictive maintenance. In this process, several quantities are observed to monitor each component's wear. By monitoring the amounts, this type of maintenance

mitigates the number of untimely interruptions, reducing unnecessary exchange costs, stops for dispensable inspections, and production disorder.

With the advent of intelligent techniques and the application of higher-order statistical processes, one more step has been taken. Possessing a large number of data and algorithms that enable its interpretation, predictive maintenance is transforming into the maintenance of prognosis. In this last type of maintenance, the fault horizon of a component can be more easily determined according to the quantity and type of its use, always considering the data coming from the measurement system. This maintenance process is aligned with the new guidelines of Industry 4.0, where automation, data exchange, cyber and physical systems, the Internet of Things, and cloud computing are applied.

It is known that there are many maintenance process management strategies and that there is no single formula or a general model for implementing a maintenance process in an industry. Strategies vary depending on the company's size, the maintenance models used, the technological level of the industry, and the degree of observability available in the company. Another aspect that should be considered is the integration of maintenance processes with other processes and areas of the company, such as supply, production, and safety. One way to start or even increase a maintenance process in a company is through the observation and example of other companies with similar characteristics.

This book contributes to the insertion of predictive maintenance in industries and the creation of prognostic maintenance. It includes chapters that address the techniques of managing maintenance processes and presents several topics of interest in the implementation of maintenance, in addition to examples and how they can be applied in industries.

> **Germano Lambert-Torres, Erik Leandro Bonaldi and Levy Eli de Lacerda de Oliveira** Gnarus Institute, Itajuba, Brazil

Section 1

Theory and Main Definitions

Section 1
