Industry 4.0 connects machines and data
Today, 3 billion people worldwide are connected to the consumer internet via digital technologies, which has drastically changed how we live, work, and communicate. By 2020, it is expected that 30 billion machines will be connected to the internet, transforming the way we can use data and analytics to drive efficiency, accelerate productivity, and attain operational excellence.
As this evolution occurs, the biopharma industry is also experiencing dramatic changes. Increasingly complex molecules, biosimilar competition, and personalized medicine are just a few of the latest trends driving new ideas and strategies about how we can efficiently bring high-quality products to the market. The convergence of the tangible physical world (machines, actuators, sensors) and the digital world (connectivity, algorithms, data analytics) opens up opportunities to access valuable insights about our development processes and manufacturing cycles for these drugs, as well as an ability to support real-time decision making.
Fig 1. The combination of data, statistics, and expertise can provide insights to improve biologic development and manufacturing.
We look to the voices of our customers to paint a picture of what they need to optimize performance in their daily roles. From lab managers who want to be able to remotely monitor equipment to engineers seeking highly reliable preventative maintenance tools, GE is creating digital solutions that can achieve these goals and more across the bioprocessing industry. By connecting machines, collecting data, and then using that data to improve the safety and quality of your products, we can minimize process variability and enable continuous improvement.
Asset Performance Management in cell therapy manufacturing
Asset Performance Management (APM) is a suite of software and service solutions designed to help collect, analyze, and visualize the data that is crucial to your assets’ health. Asset Performance Management gives you visibility into what your asset availability and reliability are and allows you to identify risks before they happen. This can be incredibly valuable in any facet of bioprocessing, but it can be particularly beneficial in cell therapy manufacturing where only one patient’s cells make up one batch. Cells extracted from a patient’s blood are purified, genetically modified, and grown in a bioreactor. They are concentrated, checked for quality, and injected back into the patient for blood-based cancer therapies.
This process creates specific challenges that make cell therapy manufacturing especially difficult, as cell and gene therapy companies must achieve a zero failure rate during the manufacturing process. Should one of these bioreactors fail overnight, for example, when no one is able to notice and there is no alarm or remote access capability, the batch fails, preventing a patient from receiving their vital treatment.
By connecting various cell processing devices and digitizing data gathered from its instruments, GE was able to develop a solution for a partner in Toronto manufacturing cell therapies. The solution required a tight collaboration between Life Science Digital, GE Digital, and GE’s bioprocess service and automation queues, which helped create a physical and digital connection between the operational and informational systems. Using Asset Performance Management, GE retrieved critical data during a cell therapy manufacturing process in order to generate a real-time monitoring dashboard. This dashboard is able to track the process health of multiple cell therapy manufacturing lines, so operators receive an alert when something goes wrong, such as a temperature control fail, and avoid the costly and disastrous situation of lost batches.
Fig 2. An individual’s T cells are multiplied in a Xuri Cell Expansion System W25, shown here as part of GE’s end-to-end solution for T cell processing.
Creating a digital twin for ultrafiltration
The evolution of Industry 4.0 has taught us that the possibilities of what we can achieve with machines become much more feasible when they are digitally enabled, especially in biopharma where we could use artificial intelligence to create simulations of our manufacturing processes. Such a “digital twin” is created when the physical, biological, and chemical properties of an asset or process are transferred to a digital format to enable a complete statistical analysis of product quality. The resulting “transfer functions” connect the asset or process quality output variables to the process and raw material variables at each level of the value stream.
For the biopharma industry, raw material variability is one of the biggest risks we face when it comes to ensuring asset or process quality. This is because the available data usually comes from supplier-provided certificates of analysis of individual raw materials based on tiny samples. Yet, these samples might not adequately characterize the specific portions of the raw material lots being consumed by the process. If we can successfully collect, analyze, and digitize representative samples of the raw materials being consumed by the process, we can feed this information to the process’ digital twin to analyze the effect of the raw material variability on the process output variables. This, in turn, will allow us to gain greater control over process outcomes.
In one case study, a GE Healthcare Life Sciences customer wants to maximize the yield of a specific protein molecule in their bioprocess, which requires a deeper understanding of how this molecule passes through GE ultrafiltration membranes. The GE team is using designed experiments at the process level and multivariate analysis at the raw material level to build a digital twin of the filter manufacturing process. The digital twin can then be used to learn how to set up the manufacturing process to produce filters capable of optimizing the yield of the customer’s specific molecule.
While it is not always easy understanding and accounting for material variation, discerning sources of process and product variation is critical for reliable manufacturing and supply. It is also a focus of regulatory bodies, as they intensify surveillance and increase requirements for traceability. If we heighten supply chain visibility, we can proactively detect and mitigate supply risks. One solution is to connect digital data platforms between suppliers and manufacturers for seamless data access.
At GE’s Cell Culture Center of Excellence in Logan, Utah, this journey has already begun with what GE calls its “brilliant factory.” A GE brilliant factory merges lean manufacturing and advanced manufacturing with advanced software and IT infrastructure to enhance productivity and enable continuous improvement. It is built on a common systems strategy leveraging GE’s cloud-based software development ecosystem, to enable digital analytics in production processes.
This ecosystem will enable GE to collect data from its suppliers and input specific information into the data platform. This platform can then be shared with customers, who also add related information. Data sharing will allow GE and its customers to look at data simultaneously rather than having to provide it manually. Ultimately, the vision is to become fully digitized with optimized processors and real-time release, where it is not just the factory itself connected but also the ecosystem around it.
Value of Industry 4.0 for manufacturing biologics and cell therapies
The digital solutions created by GE illustrate the capabilities of Industry 4.0 and what we can accomplish with them in the new era of bioprocessing. By applying remote monitoring solutions, advanced modeling, and optimization of manufacturing processes, Industry 4.0 can offer improved process reliability and product quality through the unlimited possibilities of connecting machines and data.