Consolidation of SixSigma by designing a scale error prediction model for steel strip production

voestalpine quality assurance project with Statistica data miner and voestat.



Anwenderbericht - Voestalpine

Forecast model for 34 influencing factors and production monitoring
In total, Statistica and voestat discovered 34 significant correlations. Based on this, the forecasting model was developed in Statistica with the help of the module "Generalized linear models". The explanatory power of the model is extremely high, with 95% of the mill scale errors correctly predicted. The findings of the model have been implemented directly in the production process. The model is also used to monitor production and alert if necessary.

Success Story (only in german, english version follows shortly)