- mAbxience and HP have developed an AI-based digital twin to optimise monoclonal antibody and biosimilar manufacturing.
- The tool is designed to improve predictability, efficiency and consistency in biologics production for both biosimilar licensees and CDMO clients.
mAbxience, a Fresenius entity, has entered a strategic collaboration with HP Inc. to develop an artificial intelligence (AI) solution designed to optimise the manufacturing of monoclonal antibodies and biosimilars. The new system includes a digital twin of the biological process intended to enhance predictability, process control and large-scale bioproduction efficiency.
The project uses real manufacturing data processed through neural network models and validated in an industrial environment. It allows mAbxience teams to simulate and analyse different stages of the cell culture process, with the aim of increasing yields and reducing variability. According to the companies, the tool has potential application across mAbxience’s global operations, including activities relevant to its CDMO and contract manufacturing services.
HP’s Technical Lead for the project, Sergio Martínez, said: “This is a clear example of technology with purpose: we use artificial intelligence not only to optimize processes but to help critical treatments reach more people, faster and better.”
The initiative forms part of a broader strategic framework between the companies to apply AI across multiple operational areas. mAbxience said the solution is designed to support more efficient, robust and tightly controlled manufacturing campaigns, strengthening its ability to offer cost-competitive biologics development. “Artificial intelligence may open a lot of yet untapped potential for biomanufacturing optimization,” noted Timo Liebig, Chief Innovation Officer at mAbxience.
Iván Sánchez, R&D Director at mAbxience, added that the first prototype developed with HP demonstrates how technology can support improved patient access to essential biologic therapies. The project was designed and validated in León, and may be replicated across other sites worldwide.
