Approach to design space from retrospective quality data

Abstract

Context: Nowadays, the entire manufacturing process is based on the current GMPs, which emphasize the reproducibility of the process, and companies have a lot of recorded data about their processes. Objective: The establishment of the design space (DS) from retrospective data for a wet compression process. Materials and methods: A design of experiments (DoE) with historical data from 4 years of industrial production has been carried out using the experimental factors as the results of the previous risk analysis and eight key parameters (quality specifications) that encompassed process and quality control data. Results: Software Statgraphics 5.0 was applied, and data were processed to obtain eight DS as well as their safe and working ranges. Discussion and conclusion: Experience shows that it is possible to determine DS retrospectively, being the greatest difficulty in handling and processing of high amounts of data; however, the practicality of this study is very interesting as it let have the DS with minimal investment in experiments since actual production batch data are processed statistically.