Embracing the Digital Factory for Bio/Pharma Manufacturing

Published on: 
Pharmaceutical Technology, Pharmaceutical Technology-03-02-2019, Volume 43, Issue 3
Pages: 16–21

New technologies in the digital factory enhance quality, efficiency, and flexibility for bio/pharmaceutical manufacturing.

Modern manufacturing technologies are being adopted by pharma and biopharma companies because of the value that they can provide in improving quality, efficiency, and flexibility, as well as profitability. Replacing manual activities with automated systems can remove error and increase the speed and accuracy of activities. Examples of such technologies range from automatic data capture and electronic batch records, which can improve data integrity, to using robots, which removes the potential for human error and reduces the exposure of operators to ergonomic or safety hazards. Connecting manufacturing systems and individual pieces of equipment using the industrial Internet of things (IIoT) improves data flow, so that decisions can be made more quickly and with more information, and data analytics tools create insights that enable improvements in many areas. Whether these technologies are labeled as advanced manufacturing technologies, Industry 4.0 (1), or the digital plant, they are poised to transform bio/pharma manufacturing.

Bio/pharma companies and equipment manufacturers see the benefits of connecting processes, data, and decision making. “The idea behind the digital plant is to link design, engineering, manufacturing, supply chain, distribution, and services into one intelligent system. The benefit is that the analysis of data from all these disciplines can be used to self-validate, self-improve, and even self-correct production processes within the system,” says Ben Newton, chief digital officer at GE Healthcare Life Sciences. 

“Consistent operations and the use of advanced data analytics allow for continuous improvement in product quality,” agrees Heather Coglaiti, pharma and specialty chemicals market leader for Honeywell Process Solutions. 

The promise of right-first-time, high quality product with improved uptime is a key reason to implement digital technologies, says Pamela Docherty, life-sciences industry manager, USA, at Siemens. She notes that other important benefits include reduced time to market for new product innovations, greater transparency of operations that improves planning, increased employee productivity and health and safety, and better knowledge transfer, which is crucial with an aging workforce. 

“We have to educate ourselves on emerging digital capabilities, so we can see the full set of possibilities,” says Jim Weber, advisor to manufacturing and quality IT digital manufacturing at Eli Lilly and Company. New technologies can provide creative solutions to problems, but he notes that for any project, a business case review is needed to ensure that the focus is the value, not the technology itself. One project at Lilly, for example, enhances predictive maintenance using an in-house engineering productivity tool and information management system, with a payback period of less than a year (2). 

According to analysts at McKinsey, fundamental shifts in the way data are used are leading to “a quantum change in manufacturing and order-of-magnitude improvements in processes” (3). They report that in recent examples of digital plant transformation projects, they have seen changeover times reduced by more than 30%, deviations reduced up to 80%, and increases in overall equipment effectiveness of more than 40% on packaging lines (3).

Digital maturity

Companies can evaluate their progress on the digital transformation journey using various digital maturity models, such as one developed by the BioPhorum Operations Group (4). These evaluations are valuable for setting strategies and goals, as well as for choosing projects. 

“By performing a digital maturity model assessment and associated gap analysis, organizations … can identify which assets and processes need performance improvement and identify specific, measurable KPIs [key performance indicators]. Armed with such data, it is easier to build a business case that can be used to evaluate success,” explains Dennis Belanger, director of operational certainty consulting at Emerson.

“Bio/pharma companies are more advanced than one would expect,” says Ulf Schrader, senior partner at McKinsey and an expert in the application of digital technologies in pharma manufacturing. “Clearly, automotive and electronics are far ahead. But pharma has lots of data, and scientists are interested in obtaining more insights from the data. Compared to some other industries (e.g., consumer goods), pharma is ahead.”

Transformation will require both organizational and operational changes. “Leaders target high business impact, have a clear roadmap to scale this [transformation], and work closely with HR [the human resources group] on the related people topics and with IT [the information technology group] on the tech strategy,” says Schrader.

“Transformation will happen through time, replication, and accumulated learning,” adds Weber. “We believe the most valuable benefits of digital transformation come from the new capabilities that our people develop.”

 

Implementation challenges

Implementing digital transformation can be difficult due to technical aspects (such as acceptance and fast adoption of cloud technology, cybersecurity risks, data integrity risks, and the integration of new and legacy systems), organizational resistance to change, and budgetary constraints, notes Coglaiti. She suggests that initial projects could include introducing on-line tools for workflow, reporting, and documentation-or integrating manufacturing assets, such as sensors or smart instrumentation-with the software for data analytics using the IIoT.

“An organization needs to identify the areas of production that will most benefit from that transformation first and most improve the bottom line; these will help prove ROI [return on investment] for further investments,” says Derrick Tapscott, GTC Lead EMEA at Rockwell Automation.

Shifting to a cloud platform is key, says Docherty. “Data in the cloud provides easier and more affordable access to computing power, the ability to deploy applications that visualize customer-specific insights, and the ability to contextualize data with many data sources. Cloud platforms should be open to allow data collection from any asset and any vendor. Additionally, it is imperative that a comprehensive cybersecurity strategy is put in place that covers all aspects of a facility, ranging from building access to firewalls and preventing phishing attacks.” 

Retrofitting existing facilities

Existing facilities offer a myriad of opportunities for digital transformation, but “not all projects have to be complete plant overhauls,” says Bob Lenich, director of global life sciences at Emerson. “Many of the best digital transformation results come from pilot projects on operational units that underperform. Site-level teams and gap analysis can quickly highlight opportunities for improvement, and the discovered opportunities can jumpstart digital transformations. The benefit of performing pilot projects is that organizations can then scale these pilots up to achieve the full-scale benefits. Therefore, to get the most benefit from digital transformation projects, it is crucial to develop a scale-up strategy as part of pilot implementation.”

A key challenge for transforming existing plants and equipment is changing the control systems.

Facilities typically have equipment from various vendors, and these pieces of equipment may be difficult to connect. “Plants can end up with a mess of various communication protocols and equipment that do not speak to each other,” says Docherty. “In creating the digital plant, it is critical that components are able to communicate with each other to allow potentially large amounts of data to flow seamlessly throughout the plant. Communication protocols that allow this include Profinet, Hart 7, and others.” 

Advertisement

“Many legacy control systems were not built with the ability to connect to IIoT applications and cloud analytics,” agrees Belanger. He notes that integrated process control systems allow new automation controllers to be installed on top of existing equipment.

“Making changes to the control layer in a regulated environment is a challenge, given that any control changes have consequences for validation and qualification,” adds Tapscott. “Another challenge is retrospective data integrity as paper records aren’t linked to the batch system.” 

Advanced manufacturing strategies are useful throughout the industry, whether manufacturing raw materials or finished drugs. For example, GE Healthcare Life Sciences is in the process of transforming its manufacturing facility in Logan, Utah, which produces dry powder media, liquid media, and serum for the company’s cell-culture business, by the middle of 2019. The facility has implemented an electronic batch record system and is implementing a fully automated manufacturing execution system (MES), which will connect, monitor, document, and control the manufacturing processes and data flows. The facility will use GE’s Brilliant Manufacturing software suite, made by GE Digital for use by all types of manufacturing industries; the software combines lean and advanced manufacturing with data analytics.

 

Data analytics

Data analytics-using the “Big Data” collected throughout processes-is a crucial part of digital transformation. Lilly’s digital plant vision-called “Smart Manufacturing”-is “largely driven by analysis of data and the communication of what adjustments need to be made based on the results of the data analysis,” says Wilfred Mascarenhas, advisor for data and analytics, manufacturing and quality IT at Lilly. With “smart machines,” this analysis “may be performed at the ‘edge’ (i.e., at the machine itself) and then communicated to other relevant machines/processes and humans,” he explains. 

Advanced analytics projects at Lilly are being used to identify trends and patterns more quickly and effectively. “Some examples include using natural language processing to find patterns in complaints, applying data models to predict process failures, and using artificial intelligence techniques to identify and fix errors in supply chain data,” says Weber. “We’re doing these things in our existing facilities and leveraging data from current IT and automation systems. At the same time, we’re trying to anticipate equipment and system lifecycle timelines, so that we can build digital plant capabilities into new installations from the start.” 

Data analytics tools are no longer limited to specialists but are available for the average engineer, using domain knowledge to provide context to data, to “translate the findings in Big Data into actionable knowledge,” notes Belanger. “Finding a way to get critical data from the plant floor systems and sensors to edge (computing at or close to the device) and cloud analytics systems with appropriate context for analysis is an essential step for most pharma manufacturers-even the ones that have already begun digital transformation.” One solution is integrated MES and distributed control systems that are designed to work together help eliminate the “islands of automation” that inhibit data access.

Integrating disparate sources of manufacturing data can be a challenge, agrees Tapscott. “Many valuable sources of data are older machines that may have been implemented without data integration in mind. However, an IIoT platform architecture can access that data, integrate it across production, and deliver insights with context to operators and management,” he explains. 

Predictive maintenance is one manufacturing area where machine learning and AI are proving effective. “Due to the predictive and prescriptive nature of these types of tools, we can get very early indications of potential production and reliability problems and take corrective actions to prevent them from occurring,” says Belanger. “Projects with industry clients have been shown to be able to reduce maintenance expenditure by 40–50%.”

Digital twins

One of the technologies enabled by data analytics is the digital twin, which is a virtual model of a manufacturing process or even a complete manufacturing plant. This model can be used to simulate changes in manufacturing before implementing them in the real facility. 

“The digital twin could certainly be a game changer in the pharma industry,” says Docherty. The model could be used for virtual commissioning or for operator training, as personnel could “walk through” a digital, as-is representation of a plant/skid/facility; feedback could be used in new designs. “It is important to realize that the digital twin is not something that is achieved in one simple project; it is a step-wise process that starts with the implementation of a few software tools. From there, it can evolve into different directions depending on the need of the individual business,” says Docherty. “A challenge for the pharma industry will be getting FDA to agree that the digital twin can be used in lieu of online testing, which would enable faster qualification and avoidance of downtime for code changes.” 

Coglaiti points to Honeywell’s Virtual Unit Operations Controller (vUOC) as an example of a digital twin. “The virtual, machine-based UOC is a simulation controller with large memory for use in pilot plants and R&D facilities.”

At GE Healthcare, digital twins are being used to predict the behavior of cell-culture processes so that optimal yield and harvest times can be obtained, says Newton. Data analytics tools can make process development and manufacturing faster and more efficient. “For continuous biomanufacturing, linking experimental design, manufacturing, and the measurement of key outcomes (such as yield and harvest time), we can predict the right conditions for focus and we can also adjust experimental conditions in real time to keep production yields high and costs low,” explains Newton. 

Flexible facilities require traceability and control

Continuous manufacturing of both biopharmaceuticals and solid-dose drugs offers increased flexibility and speed. These processes also require a high level of process control and traceability of materials through the process, which are enabled by digital technologies. 

Smaller patient populations are driving an increasing need for flexibility in volume and the ability to produce multiple products, whether using continuous or batch processing. “With multiple drugs per line, changeover management and operator guidance will be more important to help lower the cost of switchovers in terms of downtime and operator training,” says Tapscott. 

“Digital technologies give facilities greater flexibility through an increased capability to produce multiple products with less changeover time and with a smaller manufacturing footprint,” agrees Coglaiti. “Physical assets can be connected for more efficient processing and data flow. Improved data capture and data analytics allow for data to be more easily accessed by authorized users. Digital plant technologies also enable quick and easy scale up of processes and facilities.” 

Data integration leads to efficiency improvements, such as real-time release and optimized inventory management, in these flexible facilities, adds Lenich. “Alignment between maintenance, operations, quality, and planning allows data-such as equipment maintenance records, batch manufacturing records, and out-of-spec quality results-to be correlated across processes with context.”

Manufacturing of personalized medicines, such as cell therapies, requires the traceability and efficiency enabled by a digital enterprise. “As these targeted therapies become more prevalent, manufacturers will need to adapt to the new challenges created by new processes,” says Bob Lenich, director of global life sciences at Emerson. “To maintain the timeliness, pace, and traceability necessary to deliver life-saving personalized therapies to patients, manufacturers are leveraging electronic scheduling systems with material collection/tracking software and manufacturing execution systems to ensure the patient therapy is properly planned, manufactured, and delivered back to the patient in the required timeline.” 

“In personalized medicine production, the market is challenged with fast production and quick release,” agrees Pamela Docherty, life-sciences industry manager, USA, at Siemens. “The material must be closely tracked through the production to ensure the patient receives their application and, once completed, the release should also be almost immediate in returning the treatment to the patient. The standard release testing is not applicable. A digital enterprise enables proper tracking and data collection to add efficiency.”

 

Don’t be left behind

Increasing pressure from competitors, customers, suppliers, and government regulators is encouraging implementation of digital technologies, concludes Docherty. 

Improved performance obtained by embracing digital transformation and using Big Data analytics capabilities will raise the bar for top quartile productivity performance, predicts Belanger. As a result, he says, “The gap for the adoption laggards will be greater and the pressure for the rest of the pack to get on board will be significant.”

References

1. J. Markarian, “Modernizing Pharma Manufacturing,” Pharm. Tech. 42 (4) 2018.
2. J. Markarian, “Artificial Intelligence Takes Manufacturing Efficiency to the Next Level,” PharmTech.com, accessed Feb. 1, 2019.
3. McKinsey, “How Data is Changing the Pharma Operations World,” McKinsey.com, accessed Feb. 1, 2019.
4. BioPhorum Operations Group, “A Best Practice Guide to Using the BioPhorum Digital Plant Maturity Model and Assessment Tool” (May 2018).  

Article Details

Pharmaceutical Technology
Vol. 43, No. 3
March 2019
Pages: 16–21

Citation

When referring to this article, please cite it as J. Markarian "Embracing the Digital Factory for Bio/Pharma Manufacturing," Pharmaceutical Technology 43 (3) 2019.

About the Author

Jennifer Markarian is manufacturing editor at Pharmaceutical Technology.