Applied discrete event simulation for root cause analysis and evaluation of corrective process change Efficacy within vaccine manufacturing

Abstract

Digital Transformation of the Biopharmaceutical industry is enabling improved operations through smart manufacturing. One area of interest is the application of advanced data analytics techniques to supplement traditional workflows. The focus of this research was developing a process simulation model to address a defect observed at a manufacturing line at the Sanofi Pasteur Lyon site. This defect entailed a series of Out-of-Trend batches with abnormally low content of a certain attribute, at the end of a two-year process with complex product batch genealogy, which complicated the use of a traditional approaches to Root Cause Analysis. This study performed a statistical analysis of the defect batch attribute content through production stages to determine which contained a Root Cause. Once this analysis identified the Valence Assembly process as a stage of origin, a Discrete Event Simulator for this process was developed based on historical process data and specifications. This simulator was able to model the current process and replicate the defect in-silico. The simulator identified a specific Root Cause in the batch testing protocol as well as the expected incidence rate of the defect over future campaigns. Finally, the simulator evaluated the efficacy of two potential Corrective Process Changes. This work functions as a practical exploration of integrating novel data analysis and simulation techniques into traditional vaccine manufacturing activities.