Architectural and Technological Improvements to Integrated Bioprocess Models towards Real-Time Applications.

Area of Science:

• Bioprocess Engineering
• Digital Twins
• Process Modeling

Background:

• Integrated process models are crucial for digital twins but face limitations in bioprocess development.
• Error propagation and challenges with scale-dependent design spaces hinder current applications.
• Limited experimental data restricts exploration of process design spaces.

Purpose of the Study:

• To improve integrated process models for bioprocessing digital twins.
• To address limitations in scale-up, data handling, and extrapolation.
• To provide a framework for a digital asset throughout the product lifecycle.

Main Methods:

• Developed a simplified data model for multi-unit operation processes.
• Enhanced statistical robustness of process simulations.
• Introduced a novel simulation flow for scale-dependent variables and an extrapolation algorithm.
• Described architectural and procedural requirements for digital twin deployment.

Main Results:

• The improved models demonstrate increased statistical robustness.
• New simulation and extrapolation methods facilitate better design space exploration.
• A proposed real-time workflow enables a comprehensive digital asset framework.

Conclusions:

• The enhanced integrated process models provide a robust digital asset for bioprocessing.
• The framework supports bioprocess development and the full product lifecycle.
• This work addresses key challenges in deploying digital twins for complex bioprocesses.