The biopharmaceutical industry is undergoing significant digital transformation. While this shift brings exciting opportunities, it also introduces increasing complexity and growing expectations for data integrity. To address these challenges, companies must operate with greater precision, consistency, and efficiency throughout all stages of process development and manufacturing.
One area of digital transformation gaining particular momentum is the use of bioprocess digital twins (BPDTs). These multi-scale, virtual models represent either the entire manufacturing process or specific subunits. They strategically combine mechanistic, first-principle models with machine learning (ML), stochastic methods, and real-time process data. Unlike static models, this hybrid approach can represent intricate biological systems as well as scale-dependent effects and extrapolate beyond training conditions with high precision and accuracy. At Samsung Biologics, BPDTs are built on an integrated approach combining three core technologies: computational fluid dynamics (CFD), kinetic-based first-principles mechanistic simulations from digital manufacturing simulation (DMS), and data-driven monitoring and machine learning forecasting models from multivariate data analysis (MVDA).
Samsung Biologics’ integrated modeling approach
The creation of BPDTs begins with an accurate digital representation of the manufacturing system—through 3D scanning of the physical equipment or computer-aided design (CAD). This system is combined with CFD models to simulate and optimize the physical environment within bioreactors and other processing equipment.
By linking key process parameters such as agitation, aeration, and volume to virtual sensing for key indicators—including volumetric oxygen transfer coefficient (kLa), shear stress, and mixing behavior—CFD simulations provide insights on fluid flow, turbulence, heat transfer, cell distribution, and energy consumption. This approach is valuable when scaling up a new process or adjusting equipment and parameter settings.
Dynamic physical information from CFD is then integrated with DMS, which applies first-principle cell kinetics to simulate biologics behavior, including cell growth, nutrient consumption, and metabolite secretion. This modeling enables prediction of oxygen demand, carbon dioxide accumulation, pH, viable cell density, and titer trends in the bioreactor. These insights help determine downstream mass balance projections and uncertainty levels, as well as simulate final drug substance (DS) yield and quality before manufacturing begins.
Additionally, advanced in-line sensors, such as Raman spectroscopy, provide real-time, non-invasive updates to the simulation. These data streams enable the digital twin to operate as a live “shadow” of the physical process. The system supports automated control strategies like feeding or temperature/pH shifts to improve productivity, and can be used for product quality monitoring to prevent batch failures.
Running in parallel, MVDA models monitor multivariate process parameters across operations using a unified interface. This enables quick detection of abnormalities, allowing engineers to act early and prevent deviation. It also supports productivity forecasting and the early identification of quality risks, enhancing decision-making through data-driven simulations.
Samsung Biologics’ integrated hybrid modeling improves operational efficiency, minimizes batch failure risk, and accelerates technology transfers. Clients benefit from early detection of drifts from golden batch behavior, as well as predictions of process outcomes. These capabilities support proactive interventions, reduce time-to-market, and give clients full visibility into manufacturing procedures through digital platforms. The approach also facilitates ML analyses that combine lab and manufacturing data with CFD and DMS simulation results—enhancing productivity and maintaining consistency in product quality.
Overcoming integration challenges in digital twin implementation
Implementing advanced digital technologies requires a thorough, phased approach. Samsung Biologics is actively addressing two major challenges in advancing digital twin implementation.
The first is upgrading the manufacturing execution system (MES) from a paper-on-glass format to a fully integrated digital platform. This shift enhances system connectivity and strengthens data integrity. Previously, any operational tasks that were recorded manually could have led to inefficiencies or variability. The new MES automates these tasks, improving speed and accuracy.
The second challenge involves closing the gap between modeling tools—CFD, DMS, ML, and MVDA—which have traditionally operated in silos. The company is focused on continuously refining and integrating these tools via real-time data connectivity. This holistic connection will provide seamless insights and more reliable process predictions.
This transformation also extends beyond manufacturing. Samsung Biologics’ development labs are adopting advanced technologies such as self-validating ensemble models, functional design of experiments (DoE), Raman process analytical technology modeling, boosting ML, and automated process control strategies. These tools expand the reach of digital twin applications across the full product lifecycle. By aligning systems and fostering innovation, the company is building a more predictive, responsive, and robust manufacturing ecosystem.
About Samsung Biologics
Samsung Biologics is a fully integrated contract development and manufacturing organization (CDMO) offering end-to-end services from cell line development to aseptic fill/finish and laboratory testing. Its CGMP-compliant facilities include multiple scales of bioreactors to meet a wide range of client needs. The company offers a combined 784 kL total capacity across five plants. On top of 604 kL across four plants at Bio Campus I, Samsung Biologics added 180 kL with the completion of Plant 5 and the launch of Bio Campus II in 2025.
Samsung Biologics continues to invest in digital innovation, including major upgrades to its high-performance computing infrastructure. These investments enable faster simulations and more accurate DT modeling, accelerating process development timelines.
By integrating CFD, mechanistic DMS modeling, and MVDA into a unified framework, Samsung Biologics is enhancing process understanding, accelerating development and reducing risk—positioning itself at the forefront of digital transformation in biomanufacturing.
References
Miozza M et al.: Digital transformation of the pharmaceutical industry: a future research agenda for management studies. Technol Forecast Soc Change 2024; 207: 123580.
Chen, Y. et al.: Digital Twins in Pharmaceutical and Biopharmaceutical Manufacturing: A Literature Review. Processes 2020; 8(9), 1088.
Paik S: Enabling digital twins with computational fluid dynamics modelling. Samsung Biologics. 2025.