A good small-scale model (SSM) is a process performed at laboratory scale, representing a manufacturing-scale process. Establishing an SSM that is truly representative of manufacturing scale is integral to allowing small-scale process characterisation studies. Previously, the analysis of time-course data was performed with an approach that assessed ±3 SD of the mean manufacturing data, though sometimes we do not have access to the required three or more manufacturing runs data.

While JMP Pro’s functional data analysis function enabled us to optimise our small-scale model, it does not allow us to optimise to multiple time-course trends simultaneously (e.g., lactate concentration and viable cell density). Our workflow aims to add this function.

Using a live demo, we display this workflow on a data set with multiple time-course profiles. Using time-course data from a DOE study, the functional data explorer is used to determine shape functions of multiple trends. These shape functions are then combined into a results table, and factors desirability is set. A model is then generated with the shape functions as an output of the DOE, allowing us to optimise to match several target trends. We are now able to determine optimal conditions without subjectivity.