Intelligent manufacturing or “Biomanufacturing 4.0” is the implementation of digitalization, digital platforms, real-time monitoring and control, predictive technologies, and other data-driven tools such as machine learning (ML) and artificial intelligence (AI) to advance process and product knowledge and understanding. The information gained from those strategies can help drug developers design their processes to enhance product quality and optimize their operations. Most industry experts agree that the tools for this “next generation” of biomanufacturing will take time for widespread adoption, but such strategies will need to in place to meet the process and capacity requirements of future complex biologics. In this report, authors from the Alliance for Artificial Intelligence focus on the approaches they use to process large amounts of data (“big data”) and how the knowledge of different types of data can be used to train model parameters. Authors from Haynes Boone highlight the advantages of using AI and the intellectual property implications of AI-created inventions in the biopharmaceutical industry. And authors from Bayer discuss the application of digital twins, ML, predictive analytics, and software platforms for a viral clearance process and suggest future work in cloud implementation. 

Working with Big Data in Healthcare and Bioprocessing Settings
by Oscar Rodriguez, Adam Roose, Andrea Vuturo, and Rafael Rosengarten
Current biomanufacturing processes are producing more data than ever before. The need to interpret, analyze, and apply that data to gain knowledge to improve those process is driving the industry toward data-driven tools such as artificial intelligence and machine learning, key components of intelligent manufacturing. The traditional approach to biomanufacturing is predicated on years of experience and the identification of critical quality attributes. However, as the authors point out, “missed opportunities in early data are poorly interrogated because of constraints in data capture and available analytical systems.” The authors review different types of process data and how  can be applied to train models in both the healthcare and biomanufacturing industries.

Protecting Artificial Intelligence Inventions in Drug Development
by David McCombs, Eugene Goryunov, Dina Blikshteyn, and Kalyani Joshi
Artificial intelligence can be applied in drug discovery to process large amounts of data, to identify patterns, and to identify potential new drugs. AI also can improve clinical trials by finding appropriate subjects and ensuring medication compliance. However, current discussions of AI intellectual property are focusing on whether an AI-created invention would be considered unpatentable and whether AI should be listed as an inventor in patent applications. The authors discuss considerations of AI inventions and the implications they have in drug development. They also present potential solutions to the patent issue, suggesting that “lobbying efforts to change the law and allow AI owners to become inventors might be paths to explore.”

Pathogen Safety Digital Platform for Biopharmaceuticals
by Konstantinos Spetsieris, Michal Mleczko, Shyam Panjwani, and Oliver Hesse
Digitalization and the use of data-based tools such as digital twins have the potential to save biomanufacturers time and resources. Digital twins already are being used in other process industries. Authors from Bayer emphasize that the full benefits of using digital twins can be obtained by embracing “best practices, data science methods, digital technologies, and learnings” from those other industries. They highlight the steps that biomanufacturers can take to advance toward digitalization. The authors also provide a case study about an “exploratory innovation project” involving the application of digital twins and software platforms in viral clearance and pathogen safety. They develop a “pathogen safety model” and digital twin prototype and highlight the factors of cloud architecture for model development.

Automated Process Control Based on In Situ Measured Glucose Concentration
by Silvia Sturzenegger
Glucose is an important critical process parameter in biomanufacturing. High glucose fluctuations and excessive glucose feeding can result from bolus additions of glucose. However, insufficient feeding can lead to nutrient depletion. Because glucose concentration influences glycosylation and antibody production, frequent monitoring of glucose concentration and metabolic feedback control would help increase cell density and viability, boost product yields, reduce glycation, and decrease variability between batches. C-CIT Sensors has developed in situ glucose and lactate sensors for use in different platforms and vessel types. The authors review how those sensors and related software can be used for automated process control. They present results from studies on viable cell density and provide an example of a glucose control strategy.

Realization of Quality By Design and Beyond: The Intelligent Cell Processing System
by Masaki Hosoya, Tsutomu Soma, and Hirotoshi Kawamura
Mesenchymal stem cells (MSCs) and induced pluripotent stem cells (iPSCs) are two major cell types used in cell therapy development and manufacturing. One of most critical steps is cell expansion, a multiparametric and complex process. Automation of this step could reduce deviations from manual processing and help biomanufacturers realize quality by design. The authors describe the features of the company’s CellQualia intelligent cell processing (ICP) system, an automated instrument for cell expansion with process analytical technologies. They highlight the importance of automation and how the system is designed to be a platform device for the expansion of MSCs and iPSCs. They provide data on standard quality control testing and discuss culture-status monitoring, early signs of cellular quality change, and the application of the CellQualia ICP system in cell manufacturing.