Optimization of Factor Settings for Pharmaceutical Filling Process by Factorial Design of Mixed Levels

Abstract

Product and process variations can be costly to manufacturers in terms of high rework expenses, scrap, and inspection. We studied the variability of a generic pharmaceutical filling process (i.e., the fill weight and its related four factors). Firstly, we used mixed level factorial design to carry out the experiments and collect the data. The significance of the process factors and their interactions was determined using analysis of variance (ANOVA). Next, process capability analysis and optimization process were performed. The ultimate goal of the study was to develop the optimal level settings of controllable factors to minimize the quality loss caused by the deviation of process mean from the target value (nominal fill weight). The optimal level settings of the process factors were obtained for high and low viscosity products. As presented in this paper, significant quality improvement in the filling process can be achieved by reduction in fill weight variations. The approach may be generalized to other similar filling processes.