(738g) Accelerating Process Development in Biomanufacturing via Digitalisation

The Life Sciences sector, including pharmaceuticals and medtech, is at the forefront of the global economy, accounting for a total value of 285 billion U.S. dollars in 2020 (Mikulic, 2024). When it comes to process development and innovation, the sector’s leading position is not reflected in the day-to-day operation. Manufacturers develop processes by relying on time- and cost- intensive wet-lab experimentation. This increases the use of solvents and raw materials, negatively impacting the environmental footprint, while delaying the time-to-market of novel therapeutics (Kasemiire et al., 2021). For this, systematic approaches are required that can offer a cost- and time- efficient platforms to screen materials and operating conditions sustainably. In this context, digital tools offer a cost- and time- effective solution to perform in silico experiments that investigate the system dynamics and assist with decision-making during process development.

Model-based approaches are dependent on the existence of a reliable digital twin that describes the process dynamics. This can be either a mechanistic model that is based on first principles, a black-box model derived directly from data or a hybrid formulation. Each model structure requires diverse types of data and dataset sizes to be available for development and validation (Michalopoulou & Papathanasiou, 2024). Irrespective of the model structure, however, biopharmaceutical systems are often challenged by complex dynamics that are poorly understood and further impaired by unavailability of measurements suitable for model development. The choice of modelling approach becomes therefore conditional to the specific requirements and capabilities of the unit operation and/or system at hand. As a result, a model-based identification of a suitable operating region (design space) should be tailorable to the model type and the system specifications (Sachio et al., 2023; Sachio et al., 2024).

In this talk, present a portfolio of digital tools for modelling, optimisation and design space identification in the area of biopharmaceutical manufacturing. We focus on biomanufacturing separation systems and we assess different modelling approaches and their capabilities in serving as reliable digital twins. We further present a computer-modelling framework for design space identification that caters for both systems for which a mechanistic model is available and for occasions where fully data-driven approaches are employed. Lastly, we discuss how our tools are applicable to other sectors within Chemical Engineering, such as energy systems.

References

Kasemiire, A., Avohou, H. T., De Bleye, C., Sacre, P. Y., Dumont, E., Hubert, P. & Ziemons, E. 2021. Design of experiments and design space approaches in the pharmaceutical bioprocess optimization. European Journal of Pharmaceutics and Biopharmaceutics, 166, 144-154.

Michalopoulou, F., & Papathanasiou M. M. 2024. Assessment of data-driven modeling approaches for chromatographic separation processes. AIChE Journal, e18600

Mikulic, M. 2024. Global value of life sciences sector deals 2013-2021. Statista, [online] URL: https://www.statista.com/statistics/398216/total-value-of-global-deals-in-chemical-pharmaceutical-industry/

Sachio, S., Kontoravdi, C. & Papathanasiou, M. M. 2023. A model-based approach towards accelerated process development: A case study on chromatography. Chemical Engineering Research and Design, 197, 800-820.

Sachio, S., Likozar, B., Kontoravdi, C. & Papathanasiou, M. M. 2024. Computer-aided design space identification for screening of protein A affinity chromatography resins. Journal of Chromatography A, 1722, 464890.