Current project

Bioproduction is a rapidly growing market with significant  potential for optimization. Vanillin (4-hydroxy-3-methoxybenzaldehyde), a valuable metabolite used worldwide in the food, cosmetics and pharmaceutical industries, is an example of a compound whose biosynthetic processes could greatly benefit from improved scalability and efficiency. Existing methods such as plasmid stabilization, CRISPRi gene silencing, and two-compartment systems face limitations in terms of industrial scalability and operational complexity.

Our innovation addresses these challenges through precise monitoring of microbial cultures, enabling data-driven process optimization. For vanillin-producing E. coli (chosen for their genetic flexibility), we have developed an AI-supported bioreactor system with the following features:

• Modular sensors that measure key parameters such as temperature, pH, gas concentration, and gas flow
• Wireless, bidirectional communication for real-time transmission of measurement data to adaptive control algorithms
• Integration of machine learning, which further refines predictions through optical density measurements

Our goal is to identify and suppress 'escapers' (non-productive "cheater" subpopulations) that consume resources without contributing to vanillin synthesis. Our model calculates optimal time points for the addition of inducers to specifically control the microbial community. This approach aims to maximize production efficiency, minimize metabolic waste products, and provide a scalable foundation for industrial biosynthesis.

By combining real-time analytics with targeted metabolic interventions, STREAM offers a sustainable way to optimize microbial metabolite production on an industrial scale.