Innovative Architecture for Industrial Monitoring System

This thesis aims to develop a new innovative architecture for monitoring and SCADA systems able to centralize data and support data analytics for ideally every kind of productions. The core of the innovation is the definition and management inside the platform of a JSON configuration which describe the machine or the plant to be monitored. The state of the art gives a complete landscape about how technologies have evolved since the born of web and which of them have been applied in distributed system such as web monitoring systems. One of the most important topics here are the communication protocols, new generation of high performance databases and the purpose of machine learning algorithms in order to improve maintenance activities and support decision making process. Then the architecture of the system has been presented paying attention on the concept of reconfigurability and customization of the platform. This is done by configuring the machines by means a JSON descriptor that indicates which data collect in the system. An appendix explains in deep how the JSON configuration works and which parameters the user can manipulate in order to increase the precision and the performance of the system. Systems like the one proposed can be used in vary different ways and the thesis wants to analyze some of them and proposing solutions describing real case studies. These case studies regard quality check on the production line (online or offline), data collection for production scheduling and whole monitoring systems for textile application.