Virtual Experimentation Helps Implement Online Batch Analytics and MPC

by | Jan 31, 2007 | Industry, Life Sciences & Medical, Services, Consulting & Training, Simulation

Jim Cahill

Jim Cahill

Chief Blogger, Social Marketing Leader

Pharmaceutical Technology Europe has a recent article entitled, Artificial intelligence the key to process understanding. It discusses the opportunity to enhance the FDA’s Process Analytical Technologies (PAT) initiative using artificial intelligence based tools like neural networks, fuzzy logic and genetic algorithms. I shared this article with Greg McMillan who has been quite immersed with advanced control as it applies to bioprocesses.

I received this response which I’ll share in total (I’ve inserted some context-sensitive hyperlinks to his work on Process Control Insights):

There are opportunities to improve plant performance in the front end of the process where most of the product qualities are set by the use of online process models, batch analytics, and Model Predictive Control (MPC). Online process models based on first principals offer a significant source of knowledge discovery for both the process and the control system. The models are part of a virtual plant that enables virtual experimentation for the exploration of “what if scenarios”.

This is important for the next steps of implementing online batch analytics and MPC. Since fermentation batches take days to weeks to complete and the cost of wasted batches is considerable, the virtual plant can provide data on various degrees of adverse operating conditions that would be infeasible to obtain from the actual plant in terms of time and cost.

The virtual plant facilitates the development of techniques for the proper unfolding and alignment of batch data and more advanced analysis techniques such as super model based Principal Component Analysis. Neural networks can be employed to provide reaction rates when information on the kinetics is insufficient.

Fuzzy logic rules can be formularized and tested for a wide variety of scenarios. Inferential measurements can be developed for viable mass growth rates and product formation rates to fill in the blanks between lab measurements for MPC applications to improve batch consistency and yield and to reduce batch cycle time.

In summary, the virtual plant offers a synergistic environment for the application of online batch analytics, artificial intelligence, and advanced control. These opportunities and others are discussed in the book New Directions in Bioprocess Modeling and Control published and in the lectures on the Process Control Insights website.

Popular Posts

Comments

Follow Us

We invite you to follow us on Facebook, LinkedIn, Twitter and YouTube to stay up to date on the latest news, events and innovations that will help you face and solve your toughest challenges.

Do you want to reuse or translate content?

Just post a link to the entry and send us a quick note so we can share your work. Thank you very much.

Our Global Community

Emerson Exchange 365

The opinions expressed here are the personal opinions of the authors. Content published here is not read or approved by Emerson before it is posted and does not necessarily represent the views and opinions of Emerson.

PHP Code Snippets Powered By : XYZScripts.com