After a slow start in digitalization, biomanufacturing is on a tear. More and more biomanufacturing companies—the major firms, the niche players, and the contract development and manufacturing organizations (CDMOs)—are investing in data-driven production systems. For example, Sanofi recently indicated that it is going “all in” on artificial intelligence across all of its operations, including manufacturing.1 Similarly, GSK and Pfizer have been operationalizing “digital twins”—computer models of production processes.2,3 Roche has been moving in the same direction, endeavoring to make its plants “paperless.” At a facility in Rotkreuz, Switzerland, Roche has been digitizing its production lines and setting up digital twins. Also, at its facilities in the German cities of Mannheim and Penzberg, the company has been “pushing toward full connectivity across processes.”4
The ultimate aim of any digitalization effort is to aid decision making and make the best use of resources, says Cenk Ündey, PhD, head of pharmaceutical technical development data and digital organization at Genentech, an independent subsidiary of Roche. “The benefits of digitalization include enabling seamless data and information flow across the development value chain and into manufacturing,” he explains. “This allows development teams to find relevant data faster, increasing their productivity while advancing the pipeline activities.
“Conventional approaches lock valuable time to daunting manual tasks, whereas digitalized approaches automate tasks that are low value-add for the laboratory or manufacturing operations. It also boosts the staff motivation and helps them focus on innovation and operational excellence.”
Ündey suggests that digitalized approaches are particularly suited to process development. These approaches include in silico modeling, which can accomplish process development goals while reducing the number of “wet” laboratory experiments.
“Successful applications I have seen include support for ‘right first time’ efforts during scale up and technology transfer, process optimization, and troubleshooting,” Ündey notes. “In my career, I have had a chance to prove that when used appropriately, AI and in silico models can significantly improve COG [cost of goods], help reduce inefficiencies, and anticipate developing issues to prevent suboptimal manufacturing performance.”
Ündey concludes that when smart manufacturing concepts are combined with advanced miniaturized sensors and industrial internet of things (IoT) technology, there are opportunities to realize the potential of AI and in silico models, and to increase the efficiency of manufacturing operations.
Investment in digital technology is also increasing in the contracting sector, according to Ran Zheng, PhD, CEO of the cell and gene therapy-focused services firm Landmark Bio. “It’s been well recognized that digital technologies can revolutionize biomanufacturing,” she says. “Digital solutions for operations, maintenance, and process analytics are increasingly designed into many biomanufacturing facilities and demonstrating value and benefits.”
She notes that this investment surge extends to existing manufacturing infrastructure, with a growing number of developers choosing to retrofit plants: “Existing facilities that were not equipped with current digital technologies are catching up by introducing various digital solutions into their operations. Biomanufacturing is being transformed and will continue to evolve.”
The transformation reflects a desire to improve various aspects of data handling. There are, in Zheng’s view, four major advantages of digitalization for any biopharma company: data integrity, operational speed and efficiency, cost control, and process innovation. “Take data integrity and operational efficiency as an example,” she says. “Digitalization allows full digital traceability for every step of the process.”
These steps include analytical method development, statistical analysis, data visualization, process integration, and analytical method transfer from development laboratories to GMP manufacturing and quality control laboratories. “All of these,” Zheng emphasizes, “are accomplished in digitally connected platforms with a few clicks, eliminating resource-intensive and often lengthy processes such as manual data compilation, verification, and report generation.”
Digital supply chains
Offering a CDMO-specific perspective, Zheng says that the ability to track data through development, production, and distribution is vital, and that this ability is a matter of having the right infrastructure. To substantiate these observations, she provides a real-word example: her own company. She says that Landmark Bio unleashes the power of digitally integrated platforms to “enable seamless therapeutic translation and commercialization,” and that the company uses a unified model to “connect operations and preserve the thread of information.”
“Our platform includes a quality management system with [an] organization-wide view of process information, an enterprise resource planning system for end-to-end project cycle management, and a customer collaboration portal to manage workflow, communication, and data sharing,” Zheng details. “Once fully implemented, it will transform the development and manufacturing of cell and gene therapy products and how we interact with our partners and collaborators. It’s an ambitious plan. We are at the very beginning of a multiyear journey.”
The high level of innovation and systems development in the supply space encourages biopharma to take a growing interest in digital alternatives. “There are many technology providers catering to the biopharma industry, and more emerging players are entering the space,” Zheng remarks. “Technology solutions are available from established providers such as Cytiva, and cutting-edge technologies are being developed and ‘productized’ by startup companies.”
“It’s very exciting to see many emerging technologies in this field from both established companies and new players,” she relates. “With novel modalities such as cell and gene therapies rapidly advancing, we need new technological solutions for biomanufacturing.”
Digital demands
CDMOs are investing in digital technologies to respond to customer demands, says Dongweon Yi, director of process improvement at Samsung Biologics. One of these demands is the ability to optimize processes in real time.
“Our clients have recently started asking for a more robust use of digital technologies in the production cycle,” Yi notes. “There are still some areas within the bioprocess—where cells get cultivated to generate noble proteins—whose cell mechanisms have not been clearly defined. The digitalized manufacturing processes enables drug manufacturers to identify and develop faster and more efficient methods for process optimization through multivariable analysis.
“The traditional sampling methods used in bioprocess industries have several limitations that can affect the accuracy of the process in terms of sampling frequency, time delay, and representativeness. When sampling is done infrequently, it may not capture critical variations or fluctuations in the bioprocess.
“Real-time process monitoring and control may get interrupted when samples are transported to a laboratory for analysis, and samples may undergo changes during transport or storage, which would then compromise the accuracy of the analysis. As bioprocesses are inherently heterogeneous, sampling from a single point may not adequately represent the overall process, leading to potential inaccuracies or incomplete understanding of the system.”
To improve process control, Samsung Biologics has been applying process analytical technology, online monitoring systems, and process modeling and simulation in the virtual environment. “These technologies,” Yi asserts, “can provide online data and capture the dynamics of bioprocesses in a more comprehensive manner, ultimately improving clients’ process control, efficiency, and product quality.”
Technology drivers
Yi expects that advanced technologies will be used more frequently as they become easier to integrate on the factory floor. “AI and machine learning certainly have a role,” he adds, “especially in areas such as root cause analysis and recurrence check, which are data analysis tools used to forecast errors and optimize the overall drug development manufacturing process.”
Amr van den Adel, PhD, a senior lecturer at the Eastern Institute of Technology and Zhejiang Yuexiu University, also predicts that biopharma digitalization will accelerate as technology connectivity increases. “IoT sensors and devices can be integrated into manufacturing equipment, laboratories, and supply chains to collect real-time data on parameters such as temperature, humidity, pressure, and quality metrics,” he elaborates. “This data can be utilized for process optimization, quality control, and predictive maintenance.
“Also, by applying AI algorithms and advanced analytics techniques to the collected data, biopharmaceutical companies can gain valuable insights into process performance, identify patterns, optimize resource allocation, and enable predictive maintenance.”
But for van del Adel, the biopharma industry’s next step on its digitalization journey is as much about how people work in a digital environment—which he calls “Industry 5.0”—as it is about technologies themselves. “Adopting a 5.0 approach in the biopharma industry would require a combination of technologies to enable seamless collaboration between humans and machines,” he explains. “For example, by applying AI algorithms and advanced analytics techniques to the collected data, biopharma companies can gain valuable insights into process performance, identify patterns, optimize resource allocation, and enable predictive maintenance.”
Genentech’s Cenk Ündey also thinks that the extent of biopharma digitalization depends on the quality of the interactions between people and technology. He suggests that people may not make the most of the Wi-Fi-enabled, smart facilities of the future unless they have a “digital mindset.”
“A digital mindset refers to a way of thinking and an approach to work that leverages digital technologies and data to improve operations, decision making, and overall performance,” he explains. “How successfully we digitally transform our development and manufacturing will be contingent on how well we adapt and operationalize this digital mindset.”
References
1. Sanofi “all in” on artificial intelligence and data science to speed breakthroughs for patients. Sanofi. Press release. June 13, 2023.
2. Digital twin: using advanced technology to accelerate vaccine development. GSK. Behind the Science. May 30, 2022.
3. How Pfizer Leveraged Digital Twins to Create a Process Scale-Up Roadmap. M-Star. Case study. December 6, 2021. https://mstarcfd.com/case-study/pfizer/.
4. Delivering value through digital manufacturing—The global impact of Diagnostics Operations. Roche Stories. March 10, 2023.
