Generally, correlation analyses are used to calculate the connection between two variables in order to identify possible causal connections. On the one hand, variables can be physical measured variables, such as temperature, pressure, or torque. On the other hand, derived variables that describe, for example, the quality of a product, or event data, such as alarm or status messages, can be used for this type of analysis.
The strength of the connection (linear, monotone) of two variables is expressed by the correlation coefficient. This can have a value between 0 and 1. However, connections with high correlation coefficients must be investigated more thoroughly. This is because correlations do not describe cause-effect relationships, but, at the most, provide indications of them. Whether a correlation is really a causal relationship can be decided only with specialist knowledge or scientific investigations.
What influences product quality?
One use case for correlation analyses investigates which factors influence the quality of a product (see illustration). Specifically, this raises the question of the strength of the correlation between product quality and possible influencing factors. In this example, only the temperature and the number of alarms are suspected of having a relationship with product quality (correlation strength |k|=0.68). Slight fluctuations in pressure have a negligible effect.
In this simplified use case, warnings about temperature fluctuations could be given in good time using anomaly detection or the exact triggers for the alarms could be analyzed with the help of an event correlation analysis. The objective of correlation analysis would then be to identify all critical processes so that they could be monitored optimally with machine learning methods.
Katana, the data analytics segment of the USU Group, has extensive experience in the application of machine learning methods in the industrial sector. Take advantage of our knowledge and our solutions for building your data-driven business models to reduce your costs and improve your quality and value added.