Digital Twins of Continuous Integrated Biomanufacturing Processes

A lot of industries already realized Industry 4.0 and Industry 5.0 technologies into their manufacturing. This enables smart automation and model-based control strategies. In biopharmaceutical industry digitalization is still not common. The product is mainly performed in batchwise manner and the focus is on in-process and release analytics, which are executed off-line and time consuming. The imminent potential for process optimization control and predictive maintenance cannot be fully utilized without manufacturing in continuous mode and a digital representations of bioprocesses. Models for bioprocesses can be based on understanding of underlying biophysical mechanism e.g. heat/mass balance, adsorption/desorption kinetics or steric hindrance or similar and combined with computational fluid dynamics. Such mechanistic models are available to describe and predict chromatography, filtration, and centrifugation. Strategies will be shown how to convert a batch into a continuous integrated bioprocess and the benefits and challenges will be discussed such as reduction of floor space, high productivity but also the problem of propagation of process disturbances will be addressed. For a continuous end to end antibody manufacturing process the residence time distribution model will be shown and how process disturbance propagate. Finally, a roadmap will be presented how we could arrive at a digitized or virtual bioprocess.