A method for modelling a coating process (160) comprising a plurality of coating parameters is disclosed. The method comprises the steps of: dispensing, by means of the coating process (160) and during K work cycles, a coating on each of K pieces of objects to thereby obtain K pieces of coatings; recording (190), during each of the K work cycles, coating variable values (X i,j) of p coating parameters at M instances to thereby obtain recording results (formula (I)); and measuring (170) at least one coating property at m locations of each of the K pieces of coatings to thereby obtain measurement results (formula (II)). The method is characterized by the step of determining (210) a digital twin (220) of the coating process (160) on the basis of the recording results (formula (I)) and the measurement results (formula (II)). By using results from a large amount of classical quality control measurements together with corresponding coating parameter information, a digital twin (220) of the coating process can be determined through statistical processing of such big data. The digital twin (220) may be used either for automatic adjustment (230) of the coating parameters to obtain an improved coating quality, for prediction (240) of the coating quality right after a work cycle to obtain an improved quality control, or for both.
|IPC||G01B 11/ 06 A I|
|Publication status||Published - 13 Aug 2020|