Modularized PAT Online Training Platform to Accelerate the Workforce Innovation in Biopharmaceuticals Manufacturing

In this project, we will create a modularized extensible online training platform on leading-edge process analytical technologies (PAT).

Project status

100% Completed

Industry Need

The biomanufacturing industry is growing rapidly and becoming one of the key drivers of personalized and lifesaving medicines but faces critical challenges, including: (1) deviations

in manufacturing caused by human error and lack of process knowledge on the manufacturing floor; (2) the lengthy lead time for process development and manufacturing as well as quality testing

before release; and (3) highly regulated commercial product manufacturing systems that are "hard to change" after approval.

Solution

Creation of a modularized extensible online training platform on leading-edge process analytical technologies (PAT). The platform can provide large-scale, low-cost, and high quality life-long customized training for industry, non-profits, and college students, support workforce innovation, and facilitate biomanufacturing 4.0. Integrated with existing professional training certification programs on process control, design of experiments, and data analytics, this PAT

training platform includes: (1) end-to-end bioprocess hybrid models, which can leverage the existing mechanistic models from each unit operation, provide risk- and science-based production process understanding, and quantify the spatiotemporal causal interdependencies of critical process

parameters (CPPs) and critical quality attributes (CQAs); (2) bioprocess risk and sensitivity analyses, which can guide process specifications and troubleshooting, reduce release times, and support quality-by-design (QbD); and (3) risk-based prediction, which can provide a reliable guidance on decision making, accelerate process development, and support manufacturing automation.

In addition, the digital twin-based virtual lab (vLab) and realistic case studies were developed to reinforce the biomanufacturing mechanism learning, support the understanding of bioprocess uncertainties, provide experiential learning, and facilitate real problem-solving skills development.

Outputs/Deliverables

OUTPUTS

Conducted literature review on case studies for digital twins in biomanufacturing
Developed three real-world derived case studies.
Developed an integrated biomanufacturing simulator, including both upstream and downstream operations, and utilized calibrated parameters for improved accuracy.
Developed a N-linked glycosylation mechanistic model and an integrated biomanufacturing simulator.
Developed a dynamic Bayesian network hybrid model with sensitivity analysis capabilities.
Developed a PAT library for predictive analysis and a Raman spectra sensor monitoring simulator.
Implemented Raman spectroscopy-based process monitoring and data analytics techniques.
Developed Python code for global sensitivity analysis, which provides additional insights into the relationship between input parameters and process performance.
Developed a process control library and global sensitivity analysis code.
Documented and designed a user-friendly vLab interface.
Deployed vLab on an AWS server.
Developed training material on risk analysis, predictive analysis, process modeling, Raman spectroscopy, and process control.

Impacts

Creation of a modularized extensible online training platform on process analytical technologies (PAT). The platform can provide large-scale, low-cost, and high-quality life-long customized training for industry, non-profits, and college students,