Design and Implementation of Predictive Maintenence in Mechatonic System

Abstract

This project monitors the performance of a mechatronic device in a cloud platform called IBM Cloud. If the monitoring value reaches a certain condition, then it will automatically alert the user to call for maintenance duty by doing predictive learning analysis. The goal of this project is to create a cloud application for the data from the mechatronic system to perform the predictive maintenance operation in real time using machine learning tools such as SPSS modeler, streams flow, Watson studio and Watson IoT platform. This is achieved by sending the mechatronic system data through IoT2040, which is an interface that transforms data from S7-1500 PLC to IBM cloud by using Node-Red programming. The SPSS model is created by training the sample data using a neural network framework and connected with mechatronic system data in Streams flow. Then the streams flow analyses the data in Real-time and shows the alert in the Dashboard. This alert signal is sent back to PLC to generate a maintenance alert in the system before it affects the mechatronic system.