OR WAIT 15 SECS
Volume 31, Issue 9
Automation can improve many aspects of bioprocessing, but several hurdles must be overcome before the full range of benefits can be realized.
In less-conservative industries, automation has been widely implemented and is providing many benefits. With so many potential upsides, the risk-averse biopharmaceutical industry is showing increasing interest in automation technologies across upstream and downstream operations. Integration and regulatory challenges and a lack of standardization are some of the hurdles facing manufacturers, equipment suppliers, and software developers. There are, however, high expectations for adoption of automation solutions within the next 10 years.
In the broadest sense, the value of automation comes from streamlining manual systems or processes through the application of technology. Automation is intended to remove any redundant human elements, which are often the sources of the greatest process errors, according to Chad Beesley, global automation manager at Pall Corporation. Automation also reduces or removes unnecessary lag times inherent to manual processes. “The result is repeatable and reliable optimized processes with improved yields, minimization of human errors, optimization of human resources, and more effective collection and analysis of data,” he says.
Implementation of process analytical technology (PAT) reduces the risks of contamination associated with manual sampling, allows for timely identification and correction of process irregularities, and helps to standardize processes, reducing process variations and increasing process robustness, according to Svea Grieb, product manager for PAT and automation with Sartorius Stedim Biotech.
Batch-recipe procedural control and automated monitoring also better enable multi-product flexibility, according to Steve Perry, senior director, drug substance technical operaÂtions at Catalent Biologics. “Automation has the ability to embed quality within electronic batch records by incorporating material genealogy, production recipes, and other relevant enterprise data. Properly constructed batch reports reduce the effort required to release products and allow comprehensive tracking of therapeutic products,” adds Kenneth Clapp, senior manager at GE Healthcare Life Sciences.
Increased accuracy; ensuring quality and consistency; and greater operational efficiency, safety, and compliance are some of the most important benefits of good automation platforms, according to Elvin Vargas, director of automation/IT at Fujifilm Diosynth Biotechnologies Texas. “Overall,” asserts Perry, “these benefits add up to reduced lifecycle/production costs, shortened scale-up and product launch times, and greater ability to pursue continuous improvement.” Automation and the implementation of PAT “enable the production of higher quality and more consistent biologic drugs and regenerative therapies at reduced costs of goods, with higher flexibility and faster time to market,” summarizes Grieb.
In general, all bioprocess operations benefit from automation controls, according to Vargas. Sartorius believes that upstream processes at the development and commercial scales will benefit the most from automation, due to the highly variable nature of biological processes. “A higher degree of automation and standardization of the process steps will lead to improved batch-to-batch consistency, and in turn, product quality,” notes Grieb.
Automated glucose control at low levels (below 1g/L), which is not possible manually, is one example, according to Grieb. Reducing the amount of glucose in a culture reduces the amount of lactate, which has negative impacts on cell growth, and increase product yields. Automated control of pH to maintain the ideal value for the duration of a process is another, according to Vargas. Managing exhaust gas throughput on a single-use bioreactor is a further critical example, according to Clapp.
There are many opportunities to introduce automation in downstream processing operations as well. Optimization of the collection of purified product via automated switching of valving at the right time can directly improve the quality/uniformity of collected fractions, Clapp observes.
Catalent has found that high-speed filling, packaging, and inspection lines benefit the most from automation because humans are unable to perform at such speeds and with such accuracy. “Any processes that require accuracy and a high degree of repeatability, of sequential steps, such as bioreactions, purifications, sterilizations, and cleaning, can benefit from automation,” states Rob Parramore, manager of process technology and automation engineering with Catalent Biologics.
Ethernet/IP-connected smart/intelligent devices used for self-monitoring and calibrating can reduce the need for calibrations and associated process downtime, according to Will Meltzer, senior automation engineer with Catalent Biologics. He also notes that automated serialization and e-pedigree/tracking of finished product packaging are important for patient safety and preventing counterfeiting of products.
Automation of single-use systems is different than automation of stainless-steel processes. “In some ways, automaÂtion is much simpler than with stainless because there are no longer complex clean-in-place and steam-in-place systems that need to be interfaced to equipment. Yet, in some ways it can also be harder, primarily because with stainless systems, the user can take for granted that equipment is connected (because the piping connections are fixed) properly and ready for use,” Beesley explains.
Single-use bioprocessing requires operators to route and secure tubing, make proper connections (e.g., sterile connectors, tube welding), remove tubing clamps, and a host of other actions preÂviously automatable in conventional equipment, adds Clapp. “Automation in single-use becomes a combination of automated process control augmented by guiding the operator with their manual operations,” he says. Single-use production also requires tracking of single-use components as raw materials. Managing the material data for pedigree, traceability, and use within the batch bill of material is well-suited to properly implemented automation, he comments.
The more complicated the system or process, the greater the effort to define, implement, and ensure automation works correctly, according to Clapp. The seamless integration of process equipment and process skids into the automation system, especially when considering flexible manufacturing facilities, is an issue for Grieb. She notes that, in downstream processes, flexible automated skids capable of handling different types of unit operations based on S88-compliant recipes make it possible to run standardized and automated processes using a ‘ballroom’ concept.
There is also the challenge of aligning the process automation concept of a supplier to the facility automaÂtion concept in terms of environmental monitoring, building monitoring, and a certain level of integration into resource planning systems, according to Grieb.
Integrating the right at-line PAT instrumentation with multivariate data analysis (MVDA) for closed loop control is another challenging element of automation, according to Beesley. “While use of PAT is a big shift forward from the traditional bioprocessing paradigm, there is a deep level of impact when users can take models from completed runs and use them as models for future runs,” he says.
There are numerous IT-related issues as well, according to Patrick Asplund, principal process engineer for Catalent Biologics:
Drug manufacturers also tend to be hesitant to adopt new technologies such as cloud computing and wireless commuÂnication of PAT components, according to Grieb. “A comprehensive automation strategy for an entire bioprocess, and potentially an entire production site, requires connectivity of all components and a centralized control unit. However, that would require data sharing and access that implies safety risks,” she explains.
From the supplier perspective, Pall Biotech sees the regulatory aspect as one of the greater challenges, mainly because the biopharmaceutical industry is complex and risk-averse. “When you add in the layers of data and interaction that automation requires, you are also adding a level of complexity that, while common in other industries, is far from the norm in this industry. Fears of data integrity can start to become a concern,” Beesley explains.
In addition, some concepts of modern automation technologies and sensor technologies are not yet covered by regulatory guidelines, particularly MVDA that takes all available data and integrates them into a fingerprint, according to Grieb. The same questions arise for multi-analyte sensors that are based on computational models, as is the case for spectroscopy. Regulatory bodies need to address questions such as: “How do we validate a model for the use of GMP? What are the characteristics of a ‘good and robust’ model?”
Data will continue to grow in importance. “Much of the innovation associated with automation technologies relates to the generation, communication, and use of data,” asserts Clapp. The anaÂlytics, used to convert the data into useful knowledge, will be key to capturing the future benefits from these enabling technologies, he adds.
Batch analysis tools, for instance, allow batches to be compared for similar runs and processes. “Golden batches” (ideal batches) can be identified and used to improve batch efficiency and prediction, according to Vargas. Visualization together with data historians provides internal and possibly external users with immediate feedback of process conditions, he adds.
The application of sophisticated PAT tools in combination with multivariate data analytics has a high impact on commercial processing, agrees Grieb. “Measurements are moved forward in the process to the point of controllability. Using process fingerprints, the state of the process can be assessed at any time, leading to right-time release. Furthermore, the data generated, through real-time univariate and multivariate process monitoring, can be used for simulation and statistical modeling for process design and control and ultimately lead to predictive modeling of product quality,” she says. Sartorius expects in the near future wider adoption of analytics in GMP bioprocessing, such as spectroscopy for metabolite control and bio-capacitance for viable biomass, as well as greater adoption of MVDA and design-of-experiment (DoE) software.
Longer term, data will fuel machine learning and a host of applications enabled by artificial intelligence (AI), ultimately, leading to “smarter” equipment, processes, and production plants, according Clapp. “Analysis of the data is expected to provide more actionable information to drive efficiency and productivity. Modeling of process equipment at a variety of levels means more insights into when equipment should be maintained, how to correct batch deviations before exceptions are generated, and when opportunities for process improvement are available,” he states.
Beesley notes that data collection and processing as currently performed can easily hold back the launch of a product by weeks or months. “If we have better data collection methods and get all that data online, we can potentially reduce the time to market and leverage information more effectively,” he asserts.
There are some issues. “Operations relying on inline sensors with calibration and drift issues are challenging because of the need to perform periodic off-line re-standardization. The lack of probe reliability is a significant barrier to automation,” observes Asplund.
Perhaps one of the biggest challenges to advancing the adoption of automation is the lack of standardization of automation solutions. “IT infrastructure and interoperability challenge the biopharmaceutical industry in old and new ways. The one-off automation approaches, specific to a given biomanufacturing site’s IT environment, originated with custom stainless-steel systems. These environments routinely persist to this day,” observes Clapp. “Without economically viable, industry-wide standards acceptable to drug makers and equipment suppliers alike, potential gaps will remain,” he adds.
Better standardization between equipment suppliers, automation suppliers, and consultative/engineering companies is needed in order for more simplified automation solutions to be possible, according to Vargas. “A global standard will be much more effective than each vendor trying to create a slightly different interpretation of the same approach,” Beesley agrees.
Consistency and standardization will allow automated equipment suppliers and drug makers to address newer, advanced IT topics such as cloud computing and cyber security, according to Clapp. “Protecting and compartmentalizing data, software, and systems represent new and dynamic requirements that will challenge all those involved,” he says. A truly open communication protocol that facilitates plug-and-play implementation of automated unit operations would allow manufactures to be more flexible in their production, according to Beesley.
The pressure to reach standardization and a real interoperability between vendors at the skid-, distributed control system-, and manufacturing execution system-level is being addressed in different working groups of the International Society for Pharmaceutical Engineering, the User Association of Automation Technology in Process Industries, the Central Association of the Electrical and Electronic Industries, and the Biophorum Operations Group (BPOG), according to Grieb.
The goal of everyone in the biopharmaceutical industry is to get high-quality drugs to patients as quickly and efficiently as possible. There have, however, been many different choices and approaches to automation product and service platforms when working toward this goal. Communication between different supplier systems has been the most challenging, requiring collaboration with suppliers and additional equipment installations/configurations, according to Vargas. These tasks are the most time-and cost-consuming with respect to new equipment deployment, and often customers manage the coordination, he adds.
“Advancing automation solutions specifically for bioprocessing absolutely requires coordination as well as cooperation,” asserts Clapp. “Cooperation in the development of standards, the adopÂtion of technologies, and the implementation of operating methods is a responsibility of all stakeholders interested in meaningful progress,” he adds.
This type of collaboration has begun to occur in the industry, according to Beesley. Pall, for instance, has been working with customers, regulatory bodies, BPOG, The Open Group, and other industry associations, and even competitors, to try and address chalÂlenges head on with transparent, effective communication. Sartorius acquired Umetrics and integrated its operations as Sartorius Stedim Data Analytics and is also collaborating with Siemens on the development of its newest automation platform, according to Grieb. Automation companies such as Emerson, Siemens, and Rockwell are also working together, notes Beesley.
“Without each supplier coming to the table and working to identify tarÂgets, as well as carry forward the discussion on standardization, automation can never be fully optimized,” Beesley insists. Adds Grieb: “The task of meeting the requirement of next-generation manufacturing in terms of hardware, software, data analytics, and infrastructure is too demanding and complex to be addressed by just one supplier.”
She observes that equipment suppliers are already taking on a broader role by acting as total solutions providers, looking further than the equipment and also focusing on how their equipÂment fits and can be integrated into customer processes.
Intensified/continuous bioprocessing increases productivity, but these processes are more complex than conventional fed-batch processes and therefore require tighter monitoring and control. PAT and automation provide that control and reduce the complexity for the operator, according to Grieb.
For instance, Beesley points out that advances in at-line measurement along with MVDA models are allowÂing closed-loop control and are helping the industry move toward continuous bioprocessing with real-time release. As another example, automation of multi-column chromatography systems allows the management of the staggered operation of a variable number of columns simultaneously, enabling continuous processes and data collection and tracking, according to Beesley.
In the mid-future, Sartorius expects that modern facilities will apply intensified and continuous processing using state-of-the-art automated process batch management and S88-compliant batch recipe control functionalities, as well as plant-wide visualization and electronic batch records. “Sophisticated analysis tools, such as high-performance liquid chromatography and mass spectrometry, will be automated and integrated in the bioprocess. Together with an increased use of data science, quality-by-design approaches will be applied, allowing real-time release testing of product quality based on batch finger printing,” Grieb says.
Automation is also expected to enable more efficient manufacture of personalized medicines such as cell therapies. “Because these processes are run at small scales with high costs and high risk per batch, reduction of the risk of batch losses using online sensors and automation is highly beneficial,” Grieb says. She adds that the autologous nature of these treatments demands a process that is flexible and can dynamically adjust to wide variations in starting material, which can be accounted for in an automated fashion using PAT.
Going forward, the biopharmaceutical industry will adopt automation technologies that biologic manufacturers are confident will improve their business objectives, including those with either no regulatory impact or those that provide improved regulatory compliance and that are stable, robust, and reduce or eliminate risk, according to Clapp.
Within five years, Catalent expects an increased use of the industrial Internet of things (IIoT) and expanded use of exception-based monitoring. “We also hope for maintenance downtime to be reduced by up to 50% through less need for off-line calibrations and use of process analytical data in feedback and feedforward control,” says Parramore. In 10 years, he envisages a greater use of robotics with less human interaction in processes and more seamless integration of various automation platforms.
Pall Biotech hopes to see more plug-and-play options coming available on the market within five years and the exchange of physical controllers for virtual ones within 10 years. Further in the future, Beesley is hopeful that movement toward real-time development of patient-specific products will be possible. “Once a continuous proÂcessing line is controlled by PAT, learnÂing algorithms can be incorporated so that the machine itself can learn with the operator. And as our society moves toward Industry 4.0 with greater influÂence from artificial intelligence and machine learning, we are going to see that equipment will even be able to ‘talk’ to each other more effectively, furthering the nature of automation,” he comments.
The far-future vision for Grieb is also highly influenced by the industry 4.0 approach and related concepts such as machine learning and the IIoT. “We will see fully automated, continuous bioprocessing pipelines that require no operator interventions. Processes will be monitored and controlled remotely. Every process will have a digital twin that will be used for process simulaÂtion and prediction. More and different data will be gathered and will reside in the cloud, where data analytics can be applied easily to improve processes, regardless of manufacturing location,” she explains.
The challenge to achieving this type of automation in the future is acceptance of new technologies within a risk-averse industry. “Right now, automation has been relegated to lower-stake applications like process development. Users want to see what they can do and how they can leverage automation technologies in an area that they feel is safe. However, we see the need growing for solutions intended for commercial applications. The fact that we are seeing a transition toward more and more complex automation is beyond exciting,” asserts Beesley.
A greater understanding of where automation can provide the most value is needed, according to Parramore. “We need to have a clear strategy about what will and will not be automated, with a view as to where a human decision, verification, or intervention is required. We need to avoid over-automating, so assessment of the levels of automation comparing a fully automated operations to automation at the operational level and manual initiation of operations as directed by batch records will be important. In addition, we need to understand how much flexibility is required and where automation will be beneficial in clinical processes versus commercial processes, for instance, and what the costs and benefits of implementing automation would be for a given application.”
Vol. 31, No. 9
Pages: 10–16, 52
When referring to this article, please cite it as C. Challener, “Evaluating the Rewards vs. the Risks of Automation," BioPharm International 31 (9) 2018.