SMART High-Throughput AI-driven Fed-Batch E. coli Cultures & Scale-Up for Advanced Bioprocess 4.0 Manufacturing
Summary
Biopharmaceutical Biotechnology
Innosuisse
DataHow AG / Beckmann Coulter AG
Carmen Jungo Rhême
Skills directory
October 2024 - September 2025
This project harnesses DataHow’s cutting-edge hybrid modeling and machine learning capabilities within a Digital Twin platform combined with the power of Beckman Coulter’s BioLector® XT microbioreactors for high-throughput experimentation. The objective: to accelerate bioprocess development and streamline optimization workflows with speed and precision.
As a flagship case study, this project showcases how digital technologies and AI-driven modeling are revolutionizing the biotech landscape. It also highlights the scalability and versatility of hybrid models across different process scales.
The focus lies on optimizing the expression of recombinant Green Fluorescent Protein (GFP) in E. coli using fed-batch cultures, a cornerstone technique in biopharmaceutical manufacturing. Fed-batch strategies enable tight control over nutrient feeds, essential for maximizing protein yield, but demand careful tuning of multiple process parameters.
Using a robust Design of Experiment (DoE) approach, key parameters are systematically varied within the BioLector® XT, an automated, high throughput microbioreactor platform enabling rapid execution of parallel experiments with minimal resource input. Complementary fed-batch runs at the 5L scale are conducted at the Biofactory Competence Center (BCC), representing a critical scale used in bioprocess development prior to full manufacturing scale-up.
All experimental data feeds into DataHow’s Digital Twin platform, where advanced hybrid modeling integrates mechanistic insights with machine learning to identify optimal process conditions, faster, smarter, and at lower cost.