Predictive Maintenance: An Embedded System Approach

Predictive maintenance is an approach growing in industries, mainly in small and medium enterprises, by improving competitiveness. This paper presents a system developed in a mobile edge computing (MEC) architecture. The objective is to unite the benefits of security and the low cost of embedded devices with the necessities of predictive maintenance. This work compares the performance between the embedded and desktop devices and presents a study case considering a programmable logic controller application. For that reason, the predictive system was coded in Python, taking advantage of the language packages. And the case purpose is a simulation scenario of real-time data containing safe operational and error samples, which vary pseudo-randomly. The study result shows that the system achieves high precision. That indicates the viability and the low complexity of applying an embedded device as a solution for predictive maintenance.