This video is part of a series of three presentations for the pharmaceutical industry.
In the first session, we present the design of experiments, fitting a statistical model, and optimizing the process from the model.
In this second session for drug manufacturing, we explore and visualize the manufacturing data, and then use data mining for root cause analysis.
Suppose we are part of a drug company that is experiencing a high batch rejection rate due to tablet dissolution problems. The lower specification limit is 70%, meaning that the dissolution rate of the tablets must be at least 70, and batches with dissolution less than 70 are rejected.
Rather than optimizing the process using a design of experiments, we wish to identify the main influences on dissolution and identify short-term corrective measures to implement immediately to reduce the rejected batch rate.
Finally, we see how to use a Monte Carlo simulation, to take into account the variation in the manufacturing variables, incorporate the knowledge of the process, and obtain a more realistic idea of how the changes we implement will affect the manufacturing process in the long term.
