Ensuring the Stability of Production Processes Using Statistical Models

The stability of production processes is critical for companies to deliver products to customers on time and to specification. In mass production, where a large number of machines and equipment are used, processes are multistage and product variety is high, the possibilities for effective analysis depend largely on the level of implementation of information techniques that can increase the availability of process data. In turn, increased data availability provides a more complete picture of production processes, the changes that are taking place and their impact on quality and stability. With the increasing number of information techniques available, the amount of data generated by the manufacturing industry (process records, events, images, parameters) is expected to grow exponentially. However, access to data alone is not sufficient. To make the data useful for Root Cause Analysis (RCA), it must be analyzed, interpreted and visualized. This paper presents a comparison, evaluation and selection of the type of control cards used to monitor and control the quality performance of manufactured products. It was assumed that the values monitored on the control cards should be able to be represented as sets of non-conforming products (product fractions) in the whole population of manufactured products.