The biotechnological utilization of CO₂ offers the opportunity to produce bioproducts without the use of plant-based resources and with a reduced CO₂ footprint. In this project, the development of a gas fermentation process for producing a palm oil substitute is specifically supported through process modeling. The aim is to enable efficient process control and optimization using models.

A key focus is on the application of machine learning methods to develop data-driven models that efficiently leverage experimental data and are capable of capturing complex process dynamics. This shortens model development time and allows the models to be iteratively improved using new experimental data.

By linking mechanistic and data-driven approaches, the application of hybrid process models for gas fermentations is investigated. In this way, models can be developed that reliably capture process dynamics while also being more robust and better transferable to new process conditions through the combination of both approaches.

Based on the developed process models, approaches for model-based process control and optimization are developed, enabling proactive control of the fermentation process and the identification of optimal operating parameters.