A predictive model for improving production line performance
Résumé
Diagnosing of the reason for non-performance requires that the behaviour of the line under study be precisely analysed in order to identify potential end-of-line gains. The first section of this article presents existing analytical models for homogenous and non-homogenous line production. The second section applies those analytical models to line production performance diagnosis, injecting input data from a real case study into the models. A predictive model for improving line production performance is proposed and implemented in a diagnosis system. This model generates the production gains that would result from each action and helps the operator to choose the action that will most improve line performance. A case study of a non-homogenous production line is used to illustrate the potential efficiency of our diagnosis system