Upgrading Your Bioprocess Supervisory Control with MVDA Is Easier Than You Think
Making the decision about what type of bioprocessing technology to implement is not just about selecting the right bioreactors, but also about achieving a robust control strategy that is scalable and will stay current with the latest data-based quality management practices.
This article is posted on our Science Snippets Blog
In today’s increasingly stringent regulatory world with constant pressure to streamline bioprocesses, reduce the time to market and lower costs, smart manufacturing and process development teams are continuously looking for ways to make their processes more efficient while increasing output using automation and data analytics.
Whether that means upgrading your upstream process with a new suite of bioreactors, developing new bioprocesses, or implementing Process Analytics Technology (PAT), it’s never too early to consider how your data collection and data analytics system affects the bigger picture.
In fact, your existing systems – especially if you have data-based PAT technologies in place – may already be collecting a wealth of data that could help you to optimize and better control your processes. You could see fewer process deviations, less waste, streamlined end-to-end production, greater product output and faster delivery of the final product as a result.
Supervisory process control software such as BioPAT® MFCS SCADA software is an important part of your automation solution to ensure your bioreactor processes stay in control. Whether you have an existing facility, or are creating a new one, combining BioPAT® MFCS software with multivariate data analysis can help solve problems such as:
- Data acquisition from process units and PAT technologies, including spectroscopy
- Native connectivity to SIMCA® MVDA PAT technology
- Process monitoring and control
- Data transformation and scalability
What MVDA Brings to the Table
Adding a Multivariate Data Analysis (MVDA) solution to the equation is a clear way bringing your bioprocessing capabilities to the next level. Are you considering implementing a perfusion process? Would you like to employ advanced data analytics to make that transition faster and more reliable without having dedicated data scientists on staff or having to set up a process historian?
Are you interested in advancing your process control? Have you experienced batch deviations for which the cause of deviation couldn’t be explained? Are you ready to implement a quality control strategy based on advanced PAT models using QbD? MVDA is the way forward in all these situations.
MVDA techniques are increasingly being used for scaling and batch-to-batch comparison investigations to support or derive process understanding and, ultimately, to improve the quality, safety and efficacy of drug products. MVDA techniques are also key elements for successful implementation of spectroscopy-based PAT solutions.
Whether it is historical data that currently resides in your database, time-series or batch process data, tapping into the power of multivariate data analysis will help you gain more effective and actionable insights that lead to huge savings in time, cost and resources.
MVDA Supports Trends In Bioprocessing
Some of the key trends in the bioprocessing and manufacturing today include the move toward data-driven and process analytical technologies. These include:
- Process intensification – which requires strong QbD and process knowledge
- Automation and flexible facilities– which relies on data collection and PAT technologies
- Intensified upstream – supporting the move from fed-batch to continuous processing with real-time monitoring, multivariate analysis and predictive modeling
In addition, many of the common challenges in bioprocessing today can be addressed using data analytics. This includes issues stemming from how the process was initially developed to controlling for upstream issues like raw material variation. It’s no longer enough to have an empirically validated process. In-depth process knowledge and statistical basis for validation has become an important aspect of regulatory approvals.
In fact, the FDA and other regulatory agencies encourage the use of advanced data models, especially for continuous manufacturing processes. MVDA not only helps manage continuous manufacturing, but also is integral in being able to predict where a process is heading in the future.
How Are Control Strategies Different with MVDA?
As a supervisory control platform, BioPAT® MFCS helps keep bioprocesses under control using ISA-88-based standards that integrates online and off-line univariate data. So how is control using MVDA different?
A key difference is that MFCS uses a univariate monitoring and control strategy (one variable at a time) whereas advanced data analytics such as MVDA (using SIMCA® and SIMCA®-online) allows for multivariate monitoring and control. That means you can look at many parameters in parallel and develop control strategies and ways to monitor the process in a more holistic way instead of considering only one function at a time.
For example, if you are monitoring pH in a bioreactor, you might have an alarm set up to alert you if pH starts deviating in order implement a control strategy. BioPAT® MFCS uses a traditional univariate parameter control strategy, whereas SIMCA® evaluates other variables that might be influencing pH, as well as how pH might be impacting other variables. This provides a big picture view of how your overall process is performing. And can (under the right conditions) even predict where it will end up.
BioPAT® MFCS supervisory control looks at individual variable trendlines, whereas MVDA (using SIMCA®-online) combines all variables into a single overview.
BioPAT® MFCS supports native integration using the SimAPI to SIMCA® and SIMCA®-online, providing you with multivariate analysis and real-time process intelligence.
In addition, integration with remote alarm management systems allow process supervisors to receive timely alerts via email or phone about any process alarms.
The GMP module of BioPAT® MFCS ensures that a full audit trail related to each batch is recorded and available for review and approval in compliance with FDA 21 CFR Part 11 and EU Annex 11.
Best Working Practices: Combining Supervisory and Multivariate Control
Having BioPAT® MFCS and SIMCA®-online running in parallel can help translate data-driven insights into tangible actions. For example, let’s say you see a process drifting in SIMCA®-online when BioPAT® MFCS is not alarming because univariate alarms are in their normal ranges. SIMCA®-online not only can help to identify this earlier than univariate signals, but it also informs operators how to bring the process back into the normal control limits before any devastating process deviations occur.
You may be wondering, how does BioPAT® MFCS fit into the picture? Well, when SIMCA®-online raises a multivariate warning or alarm, operators would just need to execute those changes in BioPAT® MFCS, and the process can be easily brought back into spec.
See an example: Find out how Janssen uses MVDA to spot process drifts before their normal control systems can.
It’s Easier Than You Think to Implement
Sometimes process engineers think that implementing a data analytics control strategy will be too complicated, or requires a full-on digital transformation strategy. The truth is, if you’ve been using BioPAT® MFCS, you have the data you need to unlock the potential of data analytics. You don’t need an additional historian or complex data lake to make multivariate data analysis work for you. SIMCA® is easy to connect and set up with native integrations between the Sartorius systems.
Or, if you’re outfitting a new facility, setting both systems up from the start using platform designed to scale as you grow just makes sense. You can start small and add more bioreactors to the system when you need them. Integrated data analytics makes the entire scale-up process that much easier.
(Read more on bioreactor scaling).
Combining BIOPAT® MFCS SCADA with SIMCA® MVDA software provides out-of-the-box process insights that can help you predict the trajectory of bioprocesses and achieve better control.
BioPAT® MFCS offers a module for collecting offline sample data, the data can be incorporated into your data models to monitor the impact on your product throughout the production process. Offline samples provide an additional source of data that enriches the multivariate model and can help with quality tracking. And, for process development, that means you can implement Quality by Design (QbD).
As a scalable platform that natively integrates with advanced data analytics, a combined BioPAT® MFCS and SIMCA® MVDA SCADA solution can help upgrade your bioprocesses. Multivariate data analytics is the proven tool to help you achieve consistent quality and develop robust processes that stand up to scale-up (or scale-down) and regulatory scrutiny.
