From recent research, the number one challenge pharmaceutical manufacturers raise about data-driven optimization is that no-one has the overview. Process engineering have one system, the QA teams another, manufacturing looks at different data streams and the service technicians work in yet another system.
Without an overview, it’s next to impossible to correlate data from different sources and fully understand how the different parts of a complex system – like an industrial robot or a production line – perform and interact and how that affects product quality.
Drawing from an academic research project, this presentation will explain how this has potentially been solved with new software technologies that are able to interface with existing machine devices and software to track events, parameters, and users to deliver detailed insights into the production process without affecting the existing validation. The presentation will provide insights into experiences and results from the development of a digital twin that can predict the quality of insulin pens based on real-time data collection.
Once fully implemented, this type of technology will help meet the demands of health authorities and regulators, who are increasingly looking for much higher levels of product and process traceability during the manufacturing process