Multidimensional MSP: detecting and identifying the “invisible”
In this presentation, Mr. Daniel LAFAYE DE MICHEAUX presents a problem encountered. Traditional MSP is based on individual monitoring of parameters to be controlled; this so-called “one-dimensional” approach has the disadvantage of not taking into account the links between the parameters and therefore too often tends to implicitly assert “I see nothing = nothing happens”.
Daniel Lafaye de Micheaux and Didier Vieux rely on the industrial experience of GPC System to show how multivariate vision improves the profitability of production lines. If the serious phenomena affecting them (long stoppages, rejects, customer returns, etc.) are often invisible on the monovariate control charts, they are on the contrary clear and clearly identifiable in multivariate vision.