Monoclonal antibody production titers have increased significantly in the last 20-30 years. This is partly due to single-use technology. Disposable bioreactors made it easier to run many parallel experiments (high throughput). This helped to accelerate optimization of media, feeds, and operating conditions that drive higher cell densities and productivity. “But for other modalities such as cell therapy, we haven't reached that point yet,” says Von Stosch. “Some processes are simply more complex. You need to collect more data to understand them better. By combining HT, single-use, and modelling, we can improve processes, but even better results can be achieved by running smarter experiments, integrating more sensors and taking better measurements.”
Hybrid modelling
Von Stosch became interested in digital modelling from an early age. During his engineering studies, he recognized the importance of biology and machine learning in the future. Therefore, he pursued a PhD in hybrid modelling. “I went to Porto, Portugal. At that time this university was simply one of the few and the best to do hybrid modelling”, says Von Stosch. He also learned about innovation and entrepreneurship at MIT (USA).
“Absolutely fascinating experience”, he describes his time there. “Never in my life have I met so many people in such a short time. Everyone there was eager to make something happen.”
DataHow
Von Stosch lectured in chemical engineering at Newcastle University (UK), worked at GSK Vaccines (Brussels), co-founded Novasign, and then went to work for DataHow. DataHow focuses on transforming bioprocesses development and manufacturing to develop better processes faster and operate them robustly. "It was in the beginning of the COVID pandemic that I joined DataHow. It was an awful time for everyone, but it was a great time to advance digitalization." He helped the company grow from 6 to 40 people and established DataHow’s Portuguese branch. His research in hybrid modelling found its way into DataHow’s application. Furthermore, he and his colleagues worked hard on transfer learning, where knowledge gained in one process is transferred to the next. This significantly reduces development costs and time.
"At DataHow, we developed a platform (DataHowLab) that allows you to do digital process development leveraging hybrid modelling and transfer learning approaches. It's intended for process scientists to help them with the development or operation of their process. It is an easy to use tool for which you don’t need much knowledge about modelling."
Von Stosch is now co-founding a venture capital firm, HINA Bioventures, which will focus on financing companies in digital bioprocessing. “As a venture capitalist you can support many companies and influence the development of the sector with financial support but particularly also with expertise.”
In-silico approach
In the future, Von Stosch sees the industry moving towards a digital-first, in silico-first approach. “For instance, before wet-lab work, we will use computational models to simulate the outcome, guiding our decisions on what to pursue and what to avoid.
For me, it’s really the interaction between the human and the machine that will evolve and that will change the way we develop and operate processes. In my view, it will still be humans who interact with this intelligence in the next 5 years, connecting dots that the machine does not necessarily connect, and providing new impulses for future development.”
Moritz von Stosch will be giving a lecture at Single-Use Event Basel 2026 explaining how digital solutions support single use applications and vice versa.
Single-Use Event Basel 2026 is the ultimate networking event for professionals in bioprocessing, biotechnology and (bio)pharmaceutical manufacturing, with a focus on single use. Taking place on March 19, 2026, at Congress Center Messe Basel, the event offers a full day of innovation, insights and interaction.