OR WAIT 15 SECS
Lisa Graham leads the analytics engineering team at Seeq Corporation. She holds a Ph.D. in chemical engineering and is a registered professional chemical engineer. With 20 years of experience across many industries, her technical strengths include chemical engineering, product development and process model development. She has a well-developed acumen in crafting process analytical technology and data analytics solutions to help drive innovation and growth while reducing costs.
Advanced analytics streamlines continuous manufacturing by providing improved insights to data.
Not too many years ago, the timing of our communication and media options was largely outside our control. The phone rang at someone else’s command, while TV and radio shows arrived on their own schedules in continuous barrage. Within the past two decades, technology has dramatically altered these interactions. Now we can email each other and stream media at convenience and according to our own schedules. We have moved from a continuous to a batch mode of operation for communication, improving efficiency and enjoyment for most of us.
While batch mode may be optimal for communications, continuous mode is a much better and more efficient way to make products, as a growing number of manufacturing industries have learned. Batch operations typically introduce inefficiencies due to starting and stopping processes, and often result in notable variation in product quality. Making operations fully or partially continuous implies improved efficiency, offering ongoing opportunities to fine-tune production.
Some process industries, such as refining, have successfully adopted continuous manufacturing techniques to improve quality and reduce costs, mainly by increasing throughput. These industries were driven to make this change, mainly by process engineers with extensive training and experience in manufacturing science, at a time when there was an economic incentive to produce high volumes of goods quickly.
Today’s pharmaceutical manufacturing industry presents a different picture, and its slowness to adopt innovation is typically blamed on different issues, e.g., strict or confusing government regulations, the high costs and risk of drug development, and the overwhelming amounts of data required for drug development and manufacturing, coupled with lack of the analytics that are needed to interpret these data. What can be done to overcome these obstacles and drive the necessary advancements?
Once we understand the advantages of continuous manufacturing, we can begin the journey to improvement and optimization by helping pharma workers put their knowledge into action. One practical first step is incorporating key elements into existing processes to capture time-series process data. One powerful approach is by implementing and using streamlined process analytical technology (PAT) and improved data management and storage options. Once those are in place, companies can become more agile and transparent as industry professionals leverage advanced analytics applications and collaborative tools that allow them to share results. Results include positive culture shifts and operational improvements.
Considering these factors, the importance of extracting value from time-series data becomes strikingly apparent. Overcoming current analytics challenges such as connecting, displaying, interpolating, cleansing and contextualizing data can eliminate days or weeks of fruitless searching for insights via cumbersome spreadsheet gymnastics.
The continuous manufacturing journey requires an intuitive, visual, innovative, fast, flexible and collaborative approach. The following elements are required for success:
Advanced analytics: In continuous manufacturing, time-series data are foundational to developing process understanding and creating models. Advanced analytics enable a complete lifecycle approach, from connecting the data required to generate process insights, to producing process improvements. Key to success is putting advanced analytics applications directly into the hands of the scientists and engineers who utilize the data every day.
Data and results sharing: Employees across every pharma organization should be able to leverage insights, annotate, comment and share feedback with colleagues. These capabilities extract more value from data that are already being collected, and allow transparency so that insights can be translated into rapid improvements.
Culture: Willingness to change across the organization is crucial to the continuous manufacturing journey. Individual scientists and engineers should take ownership of advanced analytics, instead of waiting for data science experts to deliver results.
The continuous manufacturing journey ultimately benefits patients by enabling manufacturers to achieve reliable product quality and quantity of medicines at reduced cost. Predictive analytics, machine learning, and data science are great tools to help us the industry reach these goals. However, they will only be hype if the industry fails to take the concrete steps required to benefit from these tools (i.e., tapping into the expertise of current employees; supporting collaboration and knowledge capture to foster sharing and reuse of analytics; distributing insights, rapidly, to the people who need them so that they can improve outcomes, quickly.
The benefits of these advances have already been proven in other industries, and pharma can build on existing successes. Significant challenges remain, but clear methods are available today to allow the pharmaceutical industry to implement advanced analytics and improve manufacturing.