OPTIMIZATION OF INDUSTRIAL MAINTENANCE: Use of Machine Learning in effective predictive maintenance

Uso do Machine Learning em manutenções preditivas eficazes

Abstract

This article aims to analyze, through a literature review, how Machine Learning (ML) optimizes industrial predictive maintenance by impacting traditional indicators (OEE, MTTR, MTBF) and operational costs (OPEX). The theoretical framework explores the concepts of predictive maintenance, the evolution of performance indicators, and the applications of machine learning algorithms. Case studies and industrial applications highlight the effectiveness of integrating IoT sensors and Machine Learning algorithms for fault prediction and resource optimization. The research adopts a mixed-methods approach, combining qualitative and quantitative analysis, with emphasis on indicators such as OEE, MTTR, and MTBF. The findings indicate that the use of ML contributes to cost reduction, extension of equipment lifespan, and improvement of production efficiency. In the end, even with structural and cultural challenges, the trend is that these technologies will be more and more adopted, requiring technical training and a mindset switch from industrial managers and everyone else involved.