A METHOD AND APPARATUS FOR PROVIDING PREDICTIONS OF KEY PER-FORMANCE INDICATORS OF A COMPLEX MANUFACTURING SYSTEM

A computer-implemented method for providing predictions of key performance indicators,KPI^,of a product variant, PV, of a product family, PF, to be manufactured by a complex manufacturing system, MSYS, in a manufacturing process, MP, the method comprising the steps of providing (S1) a manufacturing operation model, MOpM, for each manufacturing operation type, MOT, used to manufacture a product variant, PV, of the product family, PF; providing (S2) by the complex manufacturing system, MSYS, measured contributions to key performance indicators, KPI, process context data, PCD, and process execution data, PED, of manufacturing operations, MO; learning (S3) automatically model parameters, p, of the provided manufacturing operation models, MOpM, based on collected process context data, PCD, collected process execution data, PED, and measured contributions to key performance indicators, KPI, to update the manufacturing operation models, MOpM; and evaluating (S4) an updated production efficiency model, PEM, combining updated manufacturing models, MM, including the updated manufacturing operation models, MOpM, to calculate the predictions of the key performance indicators,KPI^,of the product variant, PV, to be manufactured by said complex manufacturing system, MSYS, depending on a product configuration of the respective product variant, PV.