On the importance of variability when managing metrology capacity

In-line quality control is a crucial and increasingly constraining activity, in particular in high technology manufacturing. In this paper, we study a single metrology tool assigned to control the production quality of multiple heterogeneous machines. We introduce, model and study the tradeoff between the quality loss resulting from the sampling policy, and the quality loss induced by delays in the metrology queue. An iterative approach is proposed to optimize sampling periods using the solution of a relaxed problem which assumes full synchronization between production and metrology, and which has been previously formalized and solved. Based on computational and simulation results, and a prediction model, the paper ends with recommendations to better manage metrology capacity utilization under various levels of variability.