AI-Enhanced Continued Process Verification for Ultrafiltration/Diafiltration.

The guidelines from the Food and Drug Administration (FDA) and the European Union Good Manufacturing Practice (EU GMP) Annex 15 necessitate biopharmaceutical manufacturers to uphold continuous control of their processes throughout the product lifecycle, thereby ensuring consistent strength, quality, and purity of the final drug product. As a result, there is enormous interest in continued process verification (CPV) in the biopharmaceutical industry. Typical manufacturing processes generate significant process and analytical data for every manufactured batch. The industry has accepted that manual data collection and statistical trending are labor-intensive and error-prone. In this study, an attempt has been made to streamline CPV for the ultrafiltration-diafiltration unit operation. It entails numerous tasks, including data acquisition using sensors, predictive machine learning models, statistical trending against control limits, process capability assessment (Cpk and Ppk) at defined intervals, fault detection, and a robust process control strategy. We hope the proposed framework will help the biopharmaceutical industry implement CPV and move closer to adopting Industry 4.0.