Progressing Toward Full Automation in Biomanufacturing

Funtap – Stock.Adobe.com

Advances in manufacturing for biologics and new biotherapeutics are often based on current trends and/or changes or upgrades in equipment and processes. As biotherapeutic modalities grow more complex, equipment and process trends must keep pace. An exploration of current trends in equipment and processes can shed light on how these changes are supporting the way biomanufacturing is advancing going forward.

Developing bioprocesses

Starting early in the manufacturing lifecycle, the challenges associated with bioprocess development for a new biotherapeutic can fall into a few major categories. For one, in bioprocess development today, more than ever, there is a need for bioprocesses to be well aligned with the manufacturing supply chain and distribution strategy that is intended for that product right from the get-go, says Smriti Khera, PhD, global head of life sciences strategy and marketing at Rockwell Automation.

“If you look at the industry today, there are two parallel but opposing mega trends. One is that the manufacturing scales needed today are very different across various drug classes,” Khera says.

The bioprocess development for follow-on drug candidates, Khera emphasizes, “… willoptimize process to boost yield and productivity of the manufacturing process, and that might require increased use of automation and evaluating new digital tools and technologies such as digital twins, PAT [process analytical technology] solutions and beyond. An increased use of automation will produce efficiencies across the process, plant, and personnel needed to support the large-scale manufacture of blockbuster drugs,” Khera says.

On the other hand, precision medicines such as cell and gene therapies (CGTs) have a diametrically opposite challenge. The challenges in precision medicines is that these drugs are custom made, either to a specific patient or to a very small patient population. “In this case, the scale of GMP [good manufacturing practice] production itself is much smaller but follows a complex protocol, and the end product is re-engineered immune cells or virally encapsidated DNA that is intended to modify the disease mechanism,” Khera says.

There is also a challenge with precision medicines of addressing global demand and delivering them to the patient—who may be anywhere on the globe—within a very short shelf life of the product. “That may necessitate a completely different decentralized production. Ultimately its important for bioprocess development to bein sync with the intended manufacturing supply chain and distribution strategy for the specific drug,” Khera remarks.

The more traditional challenges in bioprocess development are around determining and monitoring set points for critical process parameters (CPPs) and critical quality attributes (CQAs). “To do this, [bioprocess developers] need to perturb and evaluate the design space around a process and get a better understanding of the process overall, and that will help them ultimately put a better process and control strategy together,” Khera observes.

CPPs can be dramatically different at different scales, so maintaining conditions correctly, such as homogeneity, temperature, and dissolved oxygen, can be very different at different scales and must be monitored. “Determining the right conditions is [typically] a very labor-intensive process within bioprocess development,” notes Khera. “If you put that into context of today’s extremely diverse and massive therapeutic pipelines, it creates a resource constriction and massive demand for the process development experts’ time. That’s where a big opportunity exists today to leverage available data and science together through mechanistic and data-driven machine learning (ML) models and be able to run process simulations to help cut down on the manual experiments and expensive pilot runs needed.”

A look at key equipment

In terms of key equipment, there are a variety of critical equipment applications in biotherapeutic manufacturing, including single-use bioreactors, cell culture equipment, viral vector production systems, purification systems, fill/finish equipment, and real-time analytics gathering, says Anthony Christopher, commercial manager of bioprocessing at Kent Elastomer Products (KEP).

Christopher emphasizes that one of the most prominent recent developments in the biomanufacturing space is the increased adoption of single-use technologies (SUTs). “Advances in disposable components and equipment in a range of applications can help to reduce contamination risks, streamline cleaning requirements, and enable faster changeovers,” he states.

Christopher points to tubing, specifically, which presents one of the critical applications for single-use systems. However, alternatives to traditional silicone tubing offer potential quality and cost optimization benefits that Christopher says can make a true impact on biopharmaceutical manufacturing operations. “Bioprocessors have the opportunity to evaluate and source high-performing elastomeric tubing options that can help bring overall costs down without sacrificing quality or accuracy issues,” he states.

“Elsewhere,” Christopher adds, “we’re monitoring other evolutions, including continuous manufacturing systems, modular production units, integrated chromatography systems, advanced sensor and monitoring systems, and more.”

Meanwhile, automation has been playing a big role in advancing the equipment supporting bioprocessing. Christopher highlights digital twin technology, integrated control systems, automated sampling and analysis, and smart bioreactors as some of the automation developments being witnessed in the biomanufacturing space. “As a company that is actively seeking to help build greater efficiency into the bioprocess applications without compromising safety or quality, KEP is continuously monitoring these trends to see how and where our product solutions have potential to further improve these processes,” he remarks.

Automation is also being increasingly used at a commercial level. “From what our team has seen, most commercial biomanufacturing processes are partially automated, with automation implemented in upstream (e.g., bioreactors, feeding systems) and downstream (e.g., chromatography, filtration) steps,” says Christopher. He interjects that manual intervention is needed for setup, troubleshooting, and quality control, but that many facilities now use automated systems for data collection, process monitoring, and quality assurance. “But,” he qualifies, “full end-to-end automation is rare.”

Establishing automation for scale up

When it comes to developing a manufacturing process for new biotherapeutics, process engineers and, in general, drug developers can benefit from implementing automated technologies into the manufacturing process.

“As an automation engineer,” says John Hatzis, global life sciences industry consultant at Rockwell Automation, “I’m always thinking about tech transfer from process development into GMP manufacturing, and it’s important to use automation technologies that can make that transition easy.”

“It’s all about leveraging industry standards out-of-the-box so as to avoid the need to reprogram the system or retrain personnel when your therapy gets approved,” Hatzis explains. “If you adopt an automation platform that applies standards [such as] ISA-101 and ISA-88 earlier in your manufacturing, you can seamlessly transition your operations into GMP without costly retraining of personnel or re-architecting your automated systems. Having the features built into the platform that can be easily configured to comply with the regulations for data integrity when you need them is really important.”

For example, Hatzis points out, a developer that builds up a repertoire of automation software for a piece of equipment that does not have “off-the-shelf” features may need to undergo significant reworking when the time comes for tech transfer into a clinical manufacturing, where data integrity becomes crucially important. “All of these things are going to be really important to make tech transfer easy and efficient [going] into clinical or GMP manufacturing,” he states.

Nicolas Pivet, vice-president and general manager, Manufacturing Capacity & Digital Solutions, Cytiva, shares the company’s experience in working with partners to develop bioprocesses for new biotherapeutics. “What we’ve been experiencing is that many of our customers actually use a platform approach for process development, which is great for speed but sometimes is not optimized for everything,” he explains, noting that it was this gap between speed and optimization that prompted Cytiva to develop a software simulation tool (GoSilico). “[This tool] helps to characterize and optimize large sets of chromatography resins, for instance, and saves a lot of time,” says Pivet.

Pivet also brings up the automation layer, which is a layer that allows equipment information to be viewed together. “Over the past seven years,” he explains, “we’ve been developing [an] automation offering (Figurate), which is a wide range of automation solution and libraries—from historian to control systems with Unicorn [control software] to SCADA [supervisory conrol and data acquisition] with Wonderware [automation platform] and to distributed control systems [e.g., DeltaV, PlantPax].”

Developing an automation library for cell culture and purification steps and improving control over CPPs such as pH, oxygen, and temperature, enables parallel processing, which can ease scale-up from lab to production, Pivet explains. Meanwhile, applying automation to downstream chromatography and filtration steps helps optimize product purification, improving purity and yield.

The improvement in process PAT has also been key in the implementation of automation in bioprocessing, Pivet notes. “Today, it’s really about outlier technologies that are used in the industry—[Raman] spectroscopy with chromatography and mass spectrometry [for example]—to monitor product quality, but those solutions have been quite cumbersome on the [manufacturing] floor, so we are working on new and breakthrough technologies [as well as] simulation[s] to avoid the burden of physical experiments,” Pivet says.

Pivet also notes that in-silico process development tools are a game changer. For example, he discusses the development of a bioreactor scaler using computational fluid dynamics capabilities, which helps with scaling up or scaling down between processes. “We’ve shown that we can save up to 50% of the time to run characterization optimization studies, and we can also achieve up to five points higher yield. Truly, I think that automation benefits the whole workflow, the operations—from seed production … to [cell] harvesting and downstream processing, even fill/finish,” he explains.

Furthermore, PAT enhances monitoring at “moments of truth,” where critical control points impact drug quality, safety, and efficacy, Pivet also explains. Integration with manufacturing execution systems and laboratory information management systems ensures full process visibility.

Moving forward

Automation technologies, including PAT, are important now and will be crucial to the industry moving forward. “I think, in the area of process analytical technology, the industry needs to move faster. [Regarding] PAT sensors [for example], there’s the emergence of improved sensing technology. And measuring CQAs, for instance, in line or online is really the holy grail of the biopharma industry right now, so the industry needs to get there in being able to measure CQAs in line and helping with real-time release, or right-time release, of the drug,” Pivet states.

“At the end, we need a software orchestration to take all those sensor data, manage them in real-time, and, using artificial intelligence [AI]/ML modeling, drive a real closed-loop optimization of the batch to end up with a batch that’s prescriptive, where you can really intervene on the batch before it’s too late and maximize the outcome of the batch. So, that’s where the industry is going,” Pivet concludes.

Meanwhile, Christopher observes that, from the research his company’s team has performed and from what is being seen among the customer base, “fully automated commercial-scale biologics manufacturing is not yet common, but the industry is rapidly moving in that direction.”

He points to advances in robotics, AI, and ML accelerating progress. “But, plenty of challenges stand in the way, including cost of implementation, regulatory compliance, and more.We have heard from experts who anticipate that fully automated, ‘lights-out’ biomanufacturing facilities could become a reality within the next five to 10 years,” Christopher adds.

About the author

Feliza Mirasol is science editor at BioPharm International®.

Article details

BioPharm International®
Vol. 38, No. 2
March 2025
Pages: 6–8

Citation

When referring to this article, please cite it as Mirasol, F. Progressing Toward Full Automation in Biomanufacturing. BioPharm International 2025, 38 (2) 6–8.