In the biomanufacturing industry where a single manufacturing disruption can cost millions, leading biotech and pharma companies are achieving 90% faster development times and 70% reduction in solvent usage through digital transformation. Just ask Massimo Morbidelli, whose pioneering work in continuous bioprocessing has transformed the industry—advancing from groundbreaking continuous chromatography to recombinant protein manufacturing driven by machine learning.
Here's why your next strategic decision should focus on integrated continuous biomanufacturing, including automation, and digitalization—and why waiting could put you years behind your competitors.
In this article, we will explore the journey of Massimo Morbidelli, Professor Emeritus at ETH Zurich and Politecnico di Milano, as he shares his insights into the future of bioprocessing, particularly focusing on integrated continuous biomanufacturing, chromatography, and emerging technologies like machine learning and AI.
This concept is discussed in greater detail in an episode of the Smart Biotech Scientist Podcast, hosted by David Brühlmann, founder of Brühlmann Consulting.
Massimo Morbidelli's Bioprocess Development Journey
Massimo Morbidelli's career in bioprocessing began over two decades ago when many of the concepts now integral to modern biotech were considered too complex or impractical. Continuous processing, upstream and downstream integration, and modeling were revolutionary ideas.
Despite initial resistance in the industry, these concepts have gained acceptance and led to significant advancements in biomanufacturing.
Early Challenges and Breakthroughs in Bioprocessing
Massimo Morbidelli's early work in bioprocess development focused on chromatography and separation processes. One of the most notable challenges he faced was the separation of chiral components using chromatography, a task requiring high selectivity.
Over time, he transitioned to applying continuous processes in biomanufacturing, particularly for biomolecules, which presented different challenges compared to small molecules.
Through his work, Massimo Morbidelli and his team sought to improve productivity, yield, and purity in bioprocessing by developing continuous processes, such as those used in simulated moving beds for chiral separations.
Their success in these areas paved the way for further innovations, including continuous chromatography and the implementation of machine learning for process optimization.
The Evolution of Continuous Bioprocessing
Massimo Morbidelli's work on continuous chromatography for biomolecules began in the early 2000s, focusing on enhancing the efficiency of purifying biomolecules while maintaining high yield and purity.
Initially met with skepticism by the industry, this process eventually established a new standard for biomanufacturing. In 2019, the concept was so widely accepted that YMC acquired ChromaCon, the company Massimo Morbidelli co-founded, further solidifying the importance of continuous bioprocessing.
The Role of Digitalization in Modern Bioprocessing
Machine Learning and Automation: The Future of Biomanufacturing
As bioprocessing continues to evolve, Massimo Morbidelli's career transitioned to integrating digitalization into biomanufacturing. Machine learning and automation are now essential tools in optimizing continuous processes. These technologies allow bioprocesses to become self-regulating and self-optimizing, offering significant productivity and cost-efficiency improvements.
Massimo Morbidelli's collaboration with DataHow, a company he co-founded, has played a pivotal role in this transformation. DataHow focuses on digitalization in biomanufacturing, leveraging machine learning to create digital twins of production processes. These digital twins enable real-time process adjustments, helping to optimize the manufacturing process while reducing waste and time-to-market.
Digital Twin Technology: A Game-Changer for Biotech
The digital twin concept is central to Massimo Morbidelli's ongoing projects. A digital twin is a virtual replica of a physical system, such as a biomanufacturing process. In Massimo Morbidelli's vision, digital twins can monitor and optimize bioprocesses' development and manufacturing stages.
These virtual models continuously learn from the real-world processes they replicate, allowing them to predict and mitigate potential disruptions, ultimately improving process efficiency and consistency.
From Concept to Commercialization
Massimo Morbidelli's work has also led to developing commercial solutions for continuous bioprocessing, including collaborations with companies like ChromaCon and DataHow.
The integration of digitalization with continuous chromatography represents a significant advancement in the field, enabling biopharmaceutical companies to produce high-quality products more efficiently. These advancements are poised to revolutionize the industry, enabling more cost-effective and scalable manufacturing processes.
Current Projects: Digitalizing Bioprocesses for E. Coli Production
The Challenge of Continuous Digitalization in E. Coli Production
After retiring from ETH Zurich in 2019, Massimo Morbidelli continued his work through collaborations with academic institutions and companies.
One of his most exciting current projects, funded by the European Research Council, focuses on the continuous digitalization of bioprocesses using E. coli as the production host. This project aims to develop a fully integrated, automated system for producing plasmid DNA.
The main challenge of this project lies in the continuous operation and digitalization of cell lysis, a critical step in extracting plasmid DNA from E. coli. This operation has no commercially available continuous units, making it a key research focus.
By developing these technologies, Massimo Morbidelli and his team aim to establish a more efficient and scalable process for producing plasmid DNA, which could have broad applications in gene therapy and vaccine production.
Collaboration and Innovation in Continuous Bioprocessing
Massimo Morbidelli's project on continuous digitalization of E. coli production brings together multiple collaborators, including Professor Alexandros Kiparisides from Thessaloniki University and his company, DataHow, for the digital aspects of the project.
Additionally, Massimo Morbidelli is working with YMC ChromaCon for the continuous chromatography purification process. This collaborative approach combines expertise in various fields, creating a powerful combination of innovation and practical application.
The Benefits of Digitalized Continuous Biomanufacturing
Speeding Up Process Development
Digitalization, machine learning, and digital twins can significantly speed up the process development phase.
By leveraging historical data and optimizing experimental conditions through techniques like transfer learning, the number of experiments needed to develop a process can be reduced. This leads to a faster time-to-market for new biopharmaceuticals, which is crucial in the rapidly evolving biotech industry.
Reducing Manufacturing Costs
Companies can reduce manufacturing costs using digital twins to optimize real-time manufacturing conditions.
The ability to predict and mitigate issues before they arise means less downtime and fewer resources wasted. Additionally, the continuous nature of these processes reduces the need for batch processing, further lowering operational costs.
We see a lot of continuous units, unit operations in the biopharma industry, even more than never before. And I was thinking about the bioreactors, for example, perfusion had a brilliant career in the past years. Now there is a renaissance of perfusion bioreactor as well. Not to talk about downstream. Also a lot of continuous operation downstream. And we have the blooming of course of machine learning, mathematical modeling.
Enhanced Process Control and Optimization
Digital twins and machine learning allow for precise control of bioprocesses, adjusting parameters such as titer and chromatography settings based on real-time data. This optimization level ensures biomanufacturing processes operate at peak efficiency, producing high-quality products with minimal waste.
The ability to continuously learn and adapt is a significant advantage in biopharmaceutical manufacturing.
Challenges in Transfer Learning for Bioprocessing
In the bioprocessing industry, there is a growing interest in leveraging knowledge gained from existing technologies to reduce experimental efforts and improve the efficiency of new product development. Massimo Morbidelli discusses the challenges in transferring learning from one bioprocess to another, especially as new modalities emerge.
Transfer of Knowledge in Bioprocessing
Transfer learning has proven effective in bioprocessing, particularly when product variations, such as vaccines, are being developed.
The power of transfer learning lies in its ability to accelerate process development. When adapting vaccines for new virus variants, biotech companies can leverage existing knowledge to eliminate up to 90% of traditionally required experiments—dramatically reducing time-to-market and development costs."
However, challenges arise when transitioning between different cell lines or product types. For instance, shifting from CHO cells (used in producing monoclonal antibodies) to another cell line or product may not yield substantial knowledge transfer. Despite this, Massimo Morbidelli's experiences suggest that even in these scenarios, the number of experiments can still be reduced by approximately 50%.
Although knowledge transfer has limitations, experts with experience in cell culture can still offer valuable insights, even if they are unfamiliar with a specific process. This expertise in general cultivation methods plays a significant role in achieving successful outcomes in new modalities.
New Modalities and Continuous Technologies in Chromatography
Bioprocessing innovations, particularly in chromatography, are evolving rapidly, focusing on continuous operations. These advancements are expected to enhance process efficiency and sustainability.
Trends in Continuous Chromatography
A significant area of focus is the development of stationary phases optimized for continuous operation.
Traditional stationary phases have been designed with batch processing in mind. However, continuous chromatography requires higher flow rates and requires larger particles to minimize pressure drop.
Smaller particles typically used for improved mass transfer in batch processes are less suitable for continuous operations. New approaches, such as monoliths and chromatographic stationary phases resembling membranes, are being explored to address these challenges.
Benefits Beyond Process Efficiency
Massimo Morbidelli notes that continuous chromatography's advantages extend beyond improving process efficiency.
- For example, continuous systems can work around the clock, allowing for higher productivity and a smaller environmental footprint. The ability to operate continuously reduces the need for waste treatment facilities and minimizes the personnel required for the process.
In particular, continuous operations have shown a tenfold increase in productivity and a reduction in solvent usage by up to 70%, even in peptide production. Leaving a constant process running without human intervention offers substantial cost savings and operational efficiencies.
The Role of Process Analytical Technology (PAT)
One of the key enabling factors for continuous operations is the advancement of Process Analytical Technology (PAT).
Sensors based on spectroscopy, such as Raman and infrared spectroscopy, provide real-time data that can be used for continuous monitoring. These sensors offer valuable insights into important parameters, such as viable cell count, glucose levels, and ammonia concentration, all from a single spectrum.
Impact of PAT on Continuous Operations
Integrating PAT systems in bioprocessing has revolutionized the ability to monitor and control continuous operations. This is crucial because continuous processes require real-time feedback to ensure consistent product quality.
In contrast to traditional batch processes, where samples are taken periodically and analyzed offline, continuous operations demand constant monitoring. PAT has made this possible by allowing for near-instantaneous data collection and analysis, which enables more efficient, uninterrupted production cycles.
Machine Learning and AI in Bioprocessing
Massimo Morbidelli emphasizes the transformative potential of machine learning and AI in bioprocessing. As the industry moves from lab-scale to commercial-scale production, digital twins and machine-learning models will be essential in optimizing processes at each stage.
The Role of Digital Twins in Process Scaling
One key concept in modern bioprocessing is developing and using digital twins—virtual models replicating physical processes. These models enable process optimization from the early stages of development, allowing smoother transitions from laboratory-scale reactors to pilot and commercial-scale operations. Machine learning techniques can be used to adapt digital twins and optimize the processes based on real-time data.
This concept also aligns with transfer learning, where knowledge gained from a previous process can be applied to new projects, whether they involve scaling up or switching to a new product variant. This approach allows for seamless knowledge transfer across processes and scales, reducing the need for extensive trial-and-error experiments.
Biopharma has been the last industrial sector in using mathematical modeling. And now they are willing to do it, they're very happy to do it, they're doing a lot. And immediately now they're faced to use machine learning techniques. So hybrid modeling, combining statistical tools with mechanistic models. And they have to do it because things are happening in that area very, very fast. And those who don't do it will remain behind and pay much more in terms of time and cost.
Overcoming Challenges in Machine Learning Integration
Integrating machine learning into bioprocessing presents its own set of challenges. Unlike other fields where vast amounts of data are available for machine learning algorithms to process, bioprocessing typically deals with limited data. Therefore, the focus is not on big data but on how to extract the most meaningful insights from the available data to achieve specific goals.
The Need for Domain Expertise
Another challenge in implementing machine learning in bioprocessing is domain expertise. While machine learning techniques can be applied to many areas, understanding the underlying biological and chemical processes is essential for success in bioprocessing.
Integrating computer scientists and bioprocess engineers is necessary to ensure that machine learning models are grounded in biological principles. Companies like DataHow, which combine expertise in both fields, are better positioned to tackle these challenges.
Moreover, bioprocessing data is expensive, necessitating a careful approach to minimize the number of experiments required. Reducing costs and time through more efficient data collection and analysis can provide a significant advantage in the industry.
Strategies for Staying Ahead of the Curve
Massimo Morbidelli advises academia and industry to stay ahead of the curve in light of the rapid advancements in bioprocessing technologies.
Industry Advice
Massimo Morbidelli stresses the importance of embracing new technologies, particularly machine learning, for industries involved in bioprocessing. Companies must integrate computer science expertise into their teams to remain competitive.
Biopharma companies, in particular, need to combine traditional statistical tools with machine learning techniques to stay competitive in an evolving market. The rapid pace of change means that those who adopt new technologies early will reap the benefits of cost and time savings.
Academia Advice
In academia, Massimo Morbidelli encourages young scientists to stay informed about the industry's evolving needs. By gaining expertise in bioprocessing and the technologies surrounding it, such as machine learning and process control, they can make significant contributions to the field. Understanding the industry's challenges and identifying ways to solve them through interdisciplinary work is key to success.
Final Remarks
We explored various aspects of bioprocessing advancements, from the challenges of transferring learning across processes and scales to the impact of continuous chromatography and machine learning. We discussed the importance of interdisciplinary knowledge in bioprocessing and the potential of digital twins and PAT technologies to optimize processes.
Integrating machine learning and AI into bioprocessing promises to significantly enhance efficiency and reduce costs, but it requires a deep understanding of both the science and the technology.
For the next generation of scientists and engineers, staying ahead of the curve involves mastering bioprocessing fundamentals, embracing emerging technologies, and engaging in interdisciplinary collaboration. As the industry continues to evolve, the insights shared in these discussions will help shape the future of bioprocessing.
About Massimo Morbidelli
Professor Emeritus at ETH Zurich (retired in 2019) and Politecnico di Milano (retired in 2023) is also a co-founder of ChromaCon, now YMC, and DataHow. He is currently the recipient of the ERC Advanced Grant for the project "Continuous Digitalisation of Bioprocesses" (CoDiBio) at the Aristotle University of Thessaloniki.
Connect with Massimo Morbidelli by email: massimo.morbidelli@polimi.it
David Brühlmann is a strategic advisor who helps C-level biotech leaders reduce development and manufacturing costs to make life-saving therapies accessible to more patients worldwide.
He is also a biotech technology innovation coach, technology transfer leader, and host of the Smart Biotech Scientist podcast—the go-to podcast for biotech scientists who want to master biopharma CMC development and biomanufacturing.
Hear It From The Horse’s Mouth
Want to listen to the full interview? Go to Smart Biotech Scientist Podcast.
Want to hear more? Do visit the podcast page and check out other episodes.
Do you wish to simplify your biologics drug development project? Contact Us
