Gavin Reynolds Senior Principal Scientist in Process Engineering and Digital at AstraZeneca, Macclesfield, UK. 

Acceleration of the development of new medicines and treatments is a key challenge across the pharmaceutical industry.  Traditional formulation development utilises extensive prototyping and make/test cycles.  Predictive modelling approaches have the potential to reduce the waste and lead time associated with traditional approaches by identifying a suitable formulation space leading to more targeted prototypes with a greater chance of achieving manufacturability and critical quality attributes.  Predicting the behaviour and performance of powders is challenging, and compounded further when predicting multi-component mixtures.  In this presentation, the focus will be on the prediction of compaction properties of powder mixtures.  This is a component of formulation design and improved predictive models in this area will at least ensure that any tablet prototypes will meet the manufacturability requirements of a mechanically robust tablet.  Recent progress in modelling, datasets and workflows enables a successful formulation window to be identified.  These tools can be combined with a system of models for other quality attributes to support rapid formulation development and process selection.