Supervisão de dados utilizando Node-Red, Node.Js e ReactJS para adaptação de sistemas legados ao contexto da indústria 4.0

Abstract

In this paper a cyber ecosystem was developed as a proposal for small and medium automations, adapting legacy systems to concepts of Industry 4.0. The Siemens Programmable Logic Controller is used in gateway communication using the Node-RED tool to send data obtained to the application developed with Node.Js, which in turn is hosted on the Amazon Cloud service. To read the data, a private web supervision system has been developed in ReactJS able to graphically display the samples obtained in real time, as well as downloading your history. In order to assess information from the collected samples, the K - Nearest Neighbors (KNN) classifier was used to diagnose failures that may occur in the process studied, so as to alert the user. Finally, the ecosystem was applied to a device in the university’s control laboratory, and using a set of 82 samples from a dataset it was possible to successfully run the proposed ecosystem. After the system was executed, it was observed that all 82 purchased samples were collected and stored without any loss and the failure classifier showed 85% accuracy. The follow-up of system parameters happened easily, practical and intuitively through the supervision system developed without any financial investment in external software licenses or training.