Spc of a near zero-defect process subject to random shocks

In this paper we study the problem of monitoring and control of a type of process in which long series with no non-conformities are observed together with occasional samples containing a large number of non-conformities. We call this a near zero-defect process subject to random shocks. Such processes occur often in practice, and a model is proposed for the identification of real non-random variations of process characteristics. Based on the statistical analysis carried out for this model, a procedure for decision-making in the control of this type of process is suggested, and analysis of some actual cases presented.