Monitoring Performance of Process Control Assets

Abstract

The thesis work is focused to help the organizations achieve maximum safety and security along with high production level and minimum expenditure on machine maintenance. The theoretical part explains how the legacy systems needs to be replaced with distributed systems and machine learning algorithms. The industrial control process consists of various components distributed at various locations and interfaced across a network. The data collected from these components are enormous. The practical implementation shows classification of the datasets based on feature extraction using machine learning algorithms, clustering to find exceptions in a dataset and using trained models to make predictions for large amount of data. As industry 4.0 is aiming for optimization of the plant process, the raw data collected is analysed using the Power BI tool for data visualization to discover new business opportunities and KPI models.