(596b) Continuous Production in Pharmaceutical Manufacturing: The Informatics View | AIChE

(596b) Continuous Production in Pharmaceutical Manufacturing: The Informatics View


Giridhar, A. V. - Presenter, Purdue University
Joglekar, G. - Presenter, Purdue University
Reklaitis, G. V. - Presenter, Purdue University

Pharmaceutical manufacturing has traditionally favored batch production paradigms. As demand and competition increase, continuous manufacturing has become more feasible from a business standpoint. From a production viewpoint, continuous production in pharmaceutical manufacturing requires the development and management of new sophisticated sensors, process models, and control models. During production, a large amount of process history data is generated. Further, a large amount of knowledge is generated and needs to be managed as well: equipment parameters, material properties, mathematical models, troubleshooting faults that arise during production, and so on. There is a huge benefit in managing all the disparate data and knowledge. Advantages include better decision support during production, improved ability to relate product formulations to material properties, improved control strategies, and easier validation of the data and knowledge stored in the system. Validation is particularly important in pharmaceutical industry, both for compliance and for liability reasons.

In this work, we demonstrate the continuous operation of a pilot plant to manufacture tablets from powders through a process of dry granulation. The key informatics components are the DeltaV distributed control system, developed by Emerson, and The Ontologies for Particulate Systems (TOPS), an ontological informatics framework that was developed by the Engineering Research Center for Structured Organic Particulate Systems (ERC-SOPS) at Purdue University. Our approach includes a rich variety of tools and technologies: an ontology system developed in the Web Ontology Language (OWL); a low-level control system written within DeltaV; a high-level control system and a fault detection/diagnosis system written in Matlab; operator interfaces that use Visual Basic; and miscellaneous off-line utility programs in Java and other languages. We also show the vital importance of interoperability and open standards for file formats and communication protocols within a production environment. In this work we summarize our key findings and offer some general conclusions.