Industrial environment: a strategy for preventive maintenance using Machine Learning to predict the useful life of equipment and Statistical Process Control for Continuous Monitoring of Variables.

• Autor(a) principal: Bezerra, Alessandro
• Data de Publicação: 2023
• Outros Autores: Silva, Katia Cilene Neles da, Nascimento, Elizamary
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: INFOCOMP: Jornal de Ciência da Computação
• Texto Completo: https://infocomp.dcc.ufla.br/index.php/infocomp/article/view/3029

The industrial context, especially those involving technologies, can suffer significant impacts from the operation of equipment on the factory floor. In view of this, several strategies have been used involving the maintenance of equipment so that the amount of corrective maintenance is reduced compared to the execution of preventive maintenance. It is understood, however, that even preventive maintenance requires greater intelligence in the face of changes that arise in the programming of the production sector or even common variations in equipment and the environment. Thus, this article presents the results obtained by implementing a predictive maintenance strategy based on the equipment's remaining lifetime (RUL), combined with real-time monitoring of equipment operating variables with the support of statistical process control tools. In this scenario, three different algorithms (Random Forest, XGBoost, and LSTM) were implemented and tested on a data sample of 60,632 observations, which allowed the results obtained to be displayed in a panel and the user to have access to predictions within their needs.