Statistical process control enables to keep an industrial process under selected and controlled conditions. The traditional analyzers and statistical tools are useful generally in an univariate way. In a multivariate approach, as required for complex batch processes, they allow the monitoring but do not explain and correct eventual detected changes. The project will develop a very promising approach: on-line intelligent analyzers (spectrometers) providing good estimations of the products quality coupled with N-Way methods, allowing a decomposition of spectral data closer to the process and products knowledge. The final deliverable will consist of a methodology and software to be used at operator level to enable a product quality control and associated actions in case of detected changes.
Funding SchemeCSC - Cost-sharing contracts