(223e) AI-Powered Digital Twins in Cell Culture Process Development: Two Case Studies

The integration of artificial intelligence (AI) and machine learning (ML) in biopharmaceutical process development has emerged as a transformative approach to enhancing reliability, efficiency, and agility. Bioprocessing’s inherent complexity—characterized by unknown mechanisms and parametric uncertainties—challenges the application of a single modeling approach across development stages.

Here, we present two case studies demonstrating the transformative potential of ML models in biopharmaceutical processes. In the first case study, we leverage ML digital twins to mathematically describe complex or unknown mechanisms in cell culture enabling in silico exploration of design spaces. These models will allow rapid evaluation of media compositions and altered process conditions to maximize productivity. The second case study demonstrates how an internally developed product quality (PQ) model enhances manufacturing process consistency by correlating process variables with critical quality attributes thereby mitigating risks in process optimization while ensuring key quality parameters remain stable.

Together, these advancements underscore the potential of AI/ML in driving innovation in biopharmaceutical manufacturing. By integrating custom data-driven PQ modeling and hybrid cell culture modeling, we are building a digital framework that enhances process understanding, accelerates development timelines, and ensures product quality consistency.