Ensuring the desired quality and yield in pharmaceutical manufacturing hinges on establishing optimal control variables. Suboptimal controls can result in batch failures, product recalls, equipment underutilization and delays in market availability. However, existing process design methods in the pharmaceutical industry rely heavily on empirical knowledge, which often fails to find the optimal controls. In this project, we developed a multiphased, data-driven optimization method that can capture the complex effect of control patterns of key process parameters on the key performance indicators (KPIs) and can provide recommendations to achieve the optimal “golden batch” for penicillin production in a bioreactor.
Presenting Company: SAS
Watch the recording
