Prediction based on executable models
The basic assumption for predicting quality is that rejects or reworking can also occur if all process parameters adhere to the set tolerances. The route to predictive quality therefore involves three steps:
1. Collect process data and correlate with true quality data
2. Develop prediction models
3. Execute model and predict quality based on real time data
In parallel to the actual quality (I.O. or N.I.O.), Predictive Quality also outputs a probability value that informs about the predicted quality.
Live at the Hannover Messe
MPDV will be showing how the new Predictive Quality application works and what possibilities there are for integration into a Manufacturing Execution System (MES) such as HYDRA live at Hannover Messe 2019 (Hall 7, Stand A12).
Further information: http://mpdv.info/pmhmipqen