The ambition in our “Simulation & Control Team” is to apply model based techniques wherever possible and reasonable. Our aim is to reduce the effort and time expense for process development and to improve existent production processes.
Overview and Ambition
Within the Industry 4.0 era, computer integrated manufacturing and digitalization plays a key role. Using computers and smart algorithms to control biotechnological processes accurately at their ideal state is of considerable interest to many biotech manufacturers. Their usage and deploymentcan lead to significant reduction of production costs by optimal scheduling, energy usage and raw conversion, while guaranteeing, high product qualities.
The ambition in our “Simulation & Control Team” is to apply model based techniques wherever possible and reasonable. Our aim is to reduce the effort and time expense for process development and to improve existent production processes.
Laboratories
Due to the high-quality requirements of biotechnological products, bioprocesses are supervised by a growing diversity of sensors and analytical devices. In our Process Analytical Technologies (PAT) Lab new production strains and bioprocesses are analysed on a fully digitalized platform:
• PIMS interconnecting all devices and sensors
• Automated sampling and analytics (e.g. HPLC, enzyme assay)
• Real-time data pre-processing and analysis
• Numerical computing (MATLAB® and Phyton®) interfaces
Research Fields
Process Modelling
The development of a reliable and applicable model is usually the critical step before model simulation and application show beneficial effects. We work on
generic workflows for the generation of applicable models, which are made for a specific application aim and are kept as simple as necessary. We work on:
• Automatic model generation
• Hybrid modelling
• Uncertainty analysis and model validation
Model-based Design of Experiments and Process Optimization
During process development we aim to run highly informative experiments. By the usage of process simulations, experiments can be planned in-silico, before lab
experiments happen. Hereby we aim to decrease the number of needed experiments while increasing the information content per taken sample.
• Optimization of sampling strategies
• Model-based design of experiments
• Multi-objective process optimization
Advanced Monitoring and Control
As process performance is highly sensitive to deviations from the optimal conditions and deviations often lead to irreversible changes, high efforts are invested in
accurately monitor and control all influential process parameters.
• Soft sensors for process monitoring
• Model-based feedback control
• Experimental verification
