With digital innovations revolutionizing consumer-facing products such as medical devices, questions are arising about whether the biopharmaceutical and broader pharmaceutical industries are embracing digital transformation to drive process improvements and meet changing product demands. Below, Fausto Artico (global head and product director of innovation and data science at GSK) shares his insights about digitalization among pharmaceutical companies that are developing protein-based biologics, vaccines, and advanced therapies.
Artico has driven several of GSK’s digitalization initiatives, including work with artificial intelligence (AI) and machine learning (ML), cybersecurity, and “hyperautomation.” Before joining GSK, he worked on key digitalization and supercomputing projects at the NVIDIA Corporation in Santa Clara, CA, and at the IBM Thomas J. Watson Center in Yorktown Heights, NY. He holds doctoral degrees in information science and computer science from the University of California at Irvine as well as several professional certifications in data science, leadership, strategy, business development and management, and finance. The remarks herein come in advance of presentations that Artico will deliver at the May 2023 BPI Europe Conference in Amsterdam, The Netherlands, and the June 2023 Connect in Pharma Conference in Geneva, Switzerland.
Honing Industry Efforts
Where are the biggest opportunities for new technologies in the pharmaceutical supply chain? Opportunities lie across the entire supply chain. For example, pharmaceutical companies have multiple products characterized by different manufacturing processes — e.g., for production of inhalers, tablet compounds, biologics, and vaccines. “Big data,” AI, and ML can help companies to optimize many such processes. Doing so would be impossible without advanced infrastructure systems to execute analytical computations; the data volume is too large to handle otherwise.
Optimize everything you can. Key opportunities for packaging lie in robotics. From folding materials to creating packages, multiple activities should be automated as much as possible. In terms of distribution, companies are focusing more on how to plan, coordinate, and schedule deliveries leveraging hyperautomation. Companies even have adopted software that derisks processes thanks to its ability to explore “what-if” scenarios. No part of the global supply chain remains untouched by automation, and every digitalization technology is an important enabler.
What role can data science play in pharmaceutical development? How do we know which data to collect, and what barriers hinder broad adoption of a data-driven approach? Data science is going to become the centerpiece of all other pharmaceutical activities. It is the foundation that enables us to retrieve, contextualize, and harmonize all of the data we have collected from both new and legacy systems. However, you never have a completely clear picture of what to collect a priori. You need to create use cases for the obstacles you will face and talk to subject-matter experts (SMEs) to get their opinions about what information would be most valuable to have. And you must prioritize the most important data sets; otherwise you risk getting lost in a process of never-ending ingestion, cleaning, linking, contextualization, and harmonization that would yield only limited benefits. Ask yourself, What information will add the most value to my research in the shortest time?
Where it concerns barriers, enormous complexities come from operating in highly regulated environments. It is difficult for pharmaceutical companies to share and handle private data and to try emerging technologies in ways that technology developers can.
My experience is that people can be the biggest barriers to or the greatest enablers of digital transformation. Many initiatives should focus more on people. People run businesses, not technology. Sometimes people don’t want to try new practices if they don’t understand what benefits could come. Some people simply fear losing their jobs once a new process is in place because they do not know where they will fit in. We must continue to educate such people because technology will allow them not only to spend less time on tedious tasks but also to focus on what matters. I have had success with showing workers — not just telling them — that digitalization is for the greater good, theirs included, and will benefit them in manifold ways.
What processes and products have you helped to improve at GSK? I’ve been involved in several exciting initiatives. One was on identifying optimal bioprocess-equipment settings. It was fascinating to discover how to create optimal environments, first to make cells grow and then to produce high-quality drugs. I also helped to establish big-data capabilities that ingest, clean, link, contextualize and harmonize many data sources to generate one, holistic version. That work was applied to research and development, human resources, finance, security, technology, manufacturing, supply chain, and other functions. My teams also created AI/ML algorithms for many activities. We designed, developed, and deployed large-scale software infrastructure that can share sensitive data securely across companies and partners. We researched, created, and published about minimal viable products (MVPs), leveraging technologies such as quantum computing and blockchain.
Thinking Globally
What role might “big pharma” companies play in the digital transformation of the sector? Such companies would do well to consider biotechnology start-ups. Early stage companies have freedom to explore or create innovative products and processes in flexible ways, but they often lack funding. By interacting with large pharmaceutical companies, start-ups can better understand the difficulties of the sector, find the correct product-market fit, and secure the needed funding.
A benefit of working with a large pharmaceutical company is that it has both capital and distribution muscle. I see big-pharma companies as enablers. For instance, they may fund fledgling companies’ clinical trials, which can cost hundreds of millions of dollars, and then help to file regulatory submissions once an early stage start-up has established proof of concept. Such a model enables exciting products to reach commercial markets in ways that are not workable for small companies on their own.
How do global pharmaceutical companies manage differences across regulations for digitalized manufacturing systems? That is a major difficulty. For example, GSK has employees in more than 100 countries. Regulations differ across locations. Our software systems must satisfy distinct requirements in different geographical regions. Such divergence puts pressure on us. We need to make the systems flexible enough to satisfy each jurisdiction’s regulations, and doing so with digital technologies is a major focus of our work.
Sometimes it is easier to buy a product that requires minimal extension or a company that already operates in a new market instead of building capabilities. Partnering can be the answer, too. Regulations constantly evolve, and global businesses are realizing that they cannot do everything in house. Rather, they need the advice and skills of experts who specialize in the needs of particular areas.
What regional “hot spots” are leading digital transformation in pharmaceutical industry? Europe has several hubs. Paris-Saclay University in France is likely to become one of the biggest healthcare poles in Europe. Likewise, some regions in France could become digitalization havens because of programs such as BioFIT. The strongest push I see at the moment is in the Luxembourg area. Other potential hot spots include Switzerland, where important pharmaceutical companies are situated, and Germany, which has several excellent universities that are keen to support and incentivize biotech spin-outs. And of course, there are Cambridge and, to a lesser degree, Oxford in the United Kingdom.
Considering Broad Implications
Can digitalization and data science make the pharmaceutical supply chain more environmentally sustainable? Data science is critical in that regard. It is useful to log and monitor everything. In doing so, you understand how much waste is produced and where it occurs. That gives a better sense of where to focus efforts, making it easier to create opportunities and eco-friendly solutions.
Rarely can you buy products that work optimally. Often you need to fine-tune processes after creating baselines that show that certain activities are consuming too many resources. Data science is fundamental for such fine-tuning, and as you automate more activities, it also becomes easier to deliver targets for environmental, social, and governance objectives.
Is there a concept from science fiction that you believe will soon become science fact? Hyperautomation — automation systems soon will assess their own components and adapt to changes in their operating environments, helping to ensure greater product yields of the right quality. Such systems also will execute big-data activities seamlessly and in ways that are too error-prone and time-consuming for humans to do.
But unlike in sci-fi movies, machines will never be fully autonomous. There always will be need for some degree of human interaction. We won’t have fully sentient AI systems, and we probably will need to put boundaries in place to ensure that machines will never take key decisions without human involvement. That might seem to be a futuristic scenario, but several of the components needed to realize such systems are already mature and available on the market (e.g., think of the logging, tracking, and monitoring capabilities we discussed). However, machines will do only the tasks that humans don’t want to or cannot do. They will make our lives easier, and we will become free to focus on the more creative parts of our work.
Dani Edmunds is SEO and editorial executive at Sciad Communications; [email protected]; 44-0-2034-05-7892. For more information about Artico’s presentations at BPI Europe 2023 (produced by Informa Connect) and Connect in Pharma 2023 (produced by Easyfairs), visit https://informaconnect.com/bpieurope and https://www.connectinpharma.com/visit/programme, respectively.
You May Also Like
