Model-based methods for identifying optimal process parameters impose specific requirements on the representation of parameters and data, which are not compatible with continuous and time-dependent processes. To optimize processes with more complex parameter and data representations using advanced methods - in particular multi-objective Bayesian experimental design - we developed a comprehensive parameter transformation system allowing engineers and process developers to comprehensively analyze and rapidly optimize their production processes.

The system has been validated on two cases including ultrashort pulse laser structuring and fermentation of plant suspension cells in bioreactors. For laser structuring, the application of our multi-objective process optimization software enabled a more precise optimization by modeling multiple key quality characteristics comprising surface roughness, structuring depth and smoothness of edge areas. In case of the fermentation, a new medium was developed which had a 33 % lower resource demand but led in parallel to a 25 % higher growth rate, thus substantially increasing the productivity and economics of the fermentation process.

In our further research, we will focus on the adaptive optimization of long-term bioreactor cultivation processes for both plant and animal cells. In doing so we emphasize the integration and enhancement of process knowledge through our optimization system.

Get in touch

Contact us to get in touch! With a membership, you’ll gain full access to all project information and updates.