Integrating Analytics Across Process Development Lifecycle

For batch processes found in industries such as pharmaceutical & biopharmaceutical, food & beverage, specialty chemicals, etc., analytics technology advancements are helping to improve quality, consistency and reliability.

Emerson's Zuwei Jin


I caught up with Emerson’s Zuwei Jin for an update on these advancements and how they are being applied. He put together a chart of the process analytics landscape. The orange lines indicate the first wave of analytics solutions developed for batch control based upon the DeltaV distributed control system.

The blue lines indicate services that have recently been developed by the Emerson consulting team to address additional needs from some of the industries—the Life Sciences in particular. The green lines represent how specific solutions from Emerson address some of these key interests.

Zuwei explained how the process analytics pieces are connecting to help improve operational performance. The on-line multivariable data analysis (MVDA) platform, DeltaV Batch Analytics, helps with the requirements in the research & development, technology transfer, chemometrician, operations, regulatory compliance and manufacturing—from operational support to business process support. Comprehensive modeling service teams can work with manufacturers to build the MVDA models drawing from the available and relevant data sources.

He noted that DeltaV Batch Analytics platform is expanding to allow it to monitor any process running in a plant through a method called Virtual Batches which is essentially OPC connections from third-party controlled processes to the DeltaV system. Other advancements include the expansion the DeltaV Batch Analytics network architecture to provide secure & comprehensive access from business-level applications. Zuwei shared how this is being brought together by project service consultants who provide modelling building and analytics platform implementation services with DeltaV engineering projects.

With some of these advancements, “golden batches” can now be readily implemented. Model-building in the on-line MVDA platform can be based on virtually any kind of data sources, such as first principle dynamic models, statistical model like design space from design of experiments (DOE), or other theoretical insight. Key technologies supporting this include dynamic time warping in real-time batch alignment. A simulation method based on the Monte Carlo method, makes it possible to implement the MVDA model without having to have multiple batches of data.

The DeltaV Batch Analytics application was designed with the U.S Food & Drug Administration (FDA)’s Stage 3 Continued Process Verification (CPV) in Process Validation general principles and practices. The fault detection and quality prediction windows, introduced years ago, are the automated form of continuous process verification.

DeltaV Batch Analytics

This application constantly measures the real-time process against the design space, allowing DeltaV Batch Analytics to be a key component of the CPV program as recommended by the FDA. The Normalized r^2 and Q are statistically more robust way of identifying abnormality and assuring the process is running under the state of control than the traditional biplots in principal component analysis (PCA).

DeltaV Virtual Batches, a collection of virtual units and receipts based on OPC connections to third party processes allows this analytics solution to be applicable for the entire plant with non-DeltaV third party processes. An analytics server connects at level 3 and interconnects with the control system historian to provide a common interface for other business intelligent applications to use the process analytics data for regulatory compliance or manufacturing intelligence.

You can connect and interact with other data analytics and pharmaceutical & biotech experts in the DeltaV and Life Sciences groups of the Emerson Exchange 365 community.

Leave a Reply

Be the First to Comment!

Notify of
avatar
wpDiscuz