Control Strategies

Improvements to the control of bioprocesses will have multiple benefits such as increasing the reliability and robustness of manufacturing with a higher assurance of molecule product quality. 

The control of process and product attributes will be enabled by various feedback strategies. For example, the careful control of particular media analytes can be used to maintain the monoclonal antibody glycan profile during an upstream process. This will require a hybrid modeling control approach alongside the necessary process analytical technology to monitor the required analytes. 

Sartorius has been building the concepts around hybrid modeling by combining multivariate data analysis with a mechanistic model and deep learning. This will ultimately lead to the combining of hybrid models for each unit operation into an integrated overarching control of the complete bioprocess. In this way, a pathway towards the autonomous running of bioprocesses that can self-regulate deviations within a defined design space will be provided. This will drive efficiency improvement to manufacture including the reduction of batch failures.