At the research symposium of the group Impol on November 21st 2012 we presented the Analysis of the process-technology bases for the optimization of processes with IMB's SPSS technology.
Modeling is becoming a more and more popular tool for analyzing and optimization of production processes as this method of work provides for cheaper and foremost faster alternative to testing on real processes or machines. Optimization of production processes increases productivity, improves the product quality and lowers production costs.
SPSS Modeler is a comprehensive product in the field of artificial intelligence with advanced analytical tools for obtaining knowledge on the process from the existent data. By identifying causes of variations and problems we better understand the process and connections between parameters. However, despite efficient analytical tool the analysis does not offer real value without consolidated and structured data and we can always be doubtful of the result. Only consolidated and properly mapped data, being the basis for technology cubes and technology analysis, contribute to added value.
As an example the building of models on data bases of the company Impol is shown. As an alternative approach to physical modeling models are built for searching the connections between the chemical composition of the EN AW-6082 alloy (internal code SAC3) and its mechanical characteristics. With models of neural networks it is shown that the company Impol can develop a model for assessing final mechanical characteristics as a function of process parameters. In this way, the optimization of process regarding productivity and quality is enabled.
Used technology: IBM SPSS