Design and Implementation of a Real-Time Industrial Data Acquisition System for Data Analytics and Predictive Maintenance
Epameinondas Orestis Batsis
Abstract
It has become clear that industrial production data is becoming more and more important these days since companies try to predict and optimize production plans. Lots of progress has been made in the development of models that achieve the aforementioned, but without the actual raw information, it is impossible to make accurate real-time assumptions. Precise, continuous, and robust machine data is one of the most important sources of information when trying to make decisions on a production plan. This diploma thesis analyses the design decisions and the implementation methods of a Real-Time Industrial Data Acquisition System, that can connect to a variety of external industrial sensors and monitor key measurements for performance and reliability. Data collected by the system can be then utilized for analytics such as production remote monitoring, real-time optimization, and predictive maintenance. The system offers many different ways of connectivity to cover a large part of the industry and a fully configurable architecture that allows customization, without the need for extra technical skills. It can be deployed to almost any Industrial environment with minimum infrastructure, allowing for a fast and reliable data acquisition plant. It is essentially a platform that can host custom applications, adapting to each solution thanks to the enhanced modularity with which it has been designed. This thesis includes six chapters. The first chapter, “Introduction”, contains a brief description of the System’s key features and overall architecture, as well as some high-level design decisions. In the second chapter, “System Analysis”, the overview of the system’s architecture and the main circuit analysis are described. In chapter three: “Implementation”, the methodology and techniques used in the design of the printed circuit board and the firmware development are presented. In chapter four, “Results”, the outcome of the experiments is presented and the overall performance of the system is evaluated. In chapter five,“Conclusion”, the Thesis outcome is discussed, and the key points of the System are pointed out. Lastly, in chapter six “Discussion and Recommendations for Future Research”, all known issues, limitations, and improvements are presented, as well as elements for further development and upgrade of the system.