Umetrics Suite Blog

A case for using predictive analytics to optimize and control bioprocesses

June 1, 2018

In bioprocessing today, a shift is happening that takes the ability to monitor, optimize and control processes to the next level. Whereas in the past manufacturers aspired to measure data in order to find out why a bioprocess action happened (using descriptive and diagnostic analytics), today we are able to use predictive analytics to determine what will happen in a bioprocess based on specific process data measured in real-time. This migration “up the food chain” to a higher level of data analytics requires automation, ongoing process monitoring and the ability to make adjustments in real-time.



Biopharma manufacturing today has moved from striving to analyze past actions to predicting future reactions using multivariate data analysis for process control.

In this context, advanced data analytics can be used to accomplish different types of modelling objectives. Monitoring models are used to detect and diagnose process deviations as they happen. Predictive monitoring models (‘forecasting models’) are used to detect and diagnose process deviations before they happen. Predictive control models (or ‘advised future models’) are used to detect and mitigate process deviations before they happen.

How can we turn from a process monitoring model to a model predictive control (MPC) one?  

Predictive monitoring in process control

One approach that can make this a reality in biologics manufacturing is the use of multivariate modeling to calibrate process models as they are being run. But what about predicting the future?

In a recent article for BioProcess International “Model Predictive Control for Bioprocess Forecasting and Optimization,” Sartorius Stedim Data Scientist Chris McCready explains:

“We turn such models into control charts and track a process as it is running. But a better approach (related to MPC) is to ask, “Based on how cells are responding and the process is running, can we predict where the process is going? If we can forecast and we have good models, can we then optimize?
Can adjustment that we can make so that the process follows the path that we want it to run? How do we then predict the future?” Right now, the turnkey strategy is to use imputation methods. In the future, biomanufacturing industry would benefit from using a “hybrid” approach — using both data-driven aspects and mechanistic understanding.

In a follow-up article, Characterizing a Bioprocess with Advanced Data Analytics, McCready discusses how modeling at various stages of the data analytics continuum aids in the scale comparison of a bioreactor process.

Here, we can see how advanced data analytics tools are used to find the “golden nuggets” in historical data, to aid in process development, to fine-tune production, and to achieve long-term improvements in product quality and throughput. Applying the right data analytics strategy to each stage of the manufacturing process can provide manufacturers with a competitive edge.

For example: “Along the data analytics continuum (described in detail in the illustration below), the most advanced challenge is being able to predict what will happen in the future and, in the event of an undesirable outcome, prescribe certain activities or interventions to prevent it from happening. Although looking ahead into the future is of greatest commercial interest, value is created at every stage of data analytics; it all depends on the specific need and the tools and approach to the analytics process.”

types of analytics from descriptive to predictive

(Figure 1: Illustration from BioPharma International) Type of of analytics from descriptive to predictive. 

To find out more about how data analytics can be used in a predictive way for process control, and for a case study for a development project for a fed-batch cell-culture biological process encompassed 75 batches, read the article in Biopharma International: Characterizing a Bioprocess with Advanced Data Analytics.

Predicting the future with SIMCA-online

You can also find out more about how real-time data analytics can be used to predict the future of processes. If you missed it, read our blog post on this topic: How to predict the future and optimize your production process.

See how it works

To find out more about how Control Advisor in SIMCA-online can help with predictive optimization and to get a free demo, click below.

Book a free demo


Have a question or comment?

Leave your thoughts below.


Topics: Real Time Process Monitoring, Statistical Process Control, Manufacturing Quality Control, Manufacturing Processes

Lennart Eriksson

Written by Lennart Eriksson

Sr Lecturer and Principal Data Scientist at Sartorius Stedim Data Analytics

Search the Blog

    Subscribe to the Blog

    View the:

    Data Analytics Glossary of Terms

    List of Webinars

    Get a free trial