Métodos multicritério e mecanismos de aprendizagem para suporte à tomada de decisão em manutenção industrial

Abstract

The current industrial scenario is undergoing transformations that impact the entire world economic context, and there is a growing need for better use of resources, reduction of waste and, mainly, dynamism in the development of activities due to the growth in competition. Industrial maintenance is a sector that plays an essential role for the operation of industrial parks, and it is important to have a system that assists the decision making process in relation to this area of industry. With technological advancement, there is the possibility of implementing applications through cloud computing platforms, developed according to the needs and specificities of each scenario. This work presents a system for prioritizing maintenance orders, developed using multicriteria decision support methods and machine learning algorithms. This system aims to improve the asset management activity, optimizing the process of meeting maintenance orders, prioritizing and allocating them according to criteria established on an information basis. The resulting system has a parallel approach between the multicriteria method and machine learning, allowing to evaluate the performance of both and identify characteristics of their application. In addition, commercial platform implementation aimed to explore the applicability of current technology platforms in the industry.