NEXCELL
AI for Tracking and Optimizing End‑To‑End Cell Processing for Clinical Cell Therapy
NEXCELL (2025 - 2029) is a collaborative research project aimed at developing a revolutionary next-generation cell and gene therapy platform for point-of-care therapy. The project aims to produce personalized cell therapies directly at the point of care, thereby improving access to life-saving therapies for cancer patients.
As part of the project, the HAIML group at University of Rostock as well as the Becker Lab at the University of Marburg are developing a probabilistic digital twin of the complex NEXCELL bioreactor system that intelligently monitors technical and biological processes. Using Bayesian AI methods, critical process parameters are reconstructed from large-scale log files and quality controls are automated.
Large language models and advanced bio-informatics algorithms enable the automatic documentation and analysis of cell fluorescence imaging data to characterize the therapeutics produced. The system can predict process deviations, optimize maintenance intervals, and generate regulatory-compliant reports. The innovation lies in the world’s first integration of hybrid probabilistic models that capture both mechanical device components and biological cell growth dynamics. Four years of project work are laying the foundation for a new era of automated, AI-supported cell therapy production. The project strengthens Mecklenburg-Western Pomerania as a biotechnology location and gives patients local access to personalized therapies for serious cancers. The developed AI technologies are applicable in other areas of biotechnology and industry 4.0.