By Gareth Macdonald
Digital technologies are forcing drug companies to rethink how they manage information on the factory floor. After a slow start, drug industry use of digital systems—bioprocessing technologies that monitor parameters in real-time and transmit data for modeling and process control—is increasing.
A 2021 survey by Deloitte suggested a growing number of drug firms are using digital twins—computer models or production processes—and “cloud” data storage technologies—as part of a wider drive for efficiency and supply chain security.
And these observations are echoed in a new study by researchers at the University of Delaware, Rutgers, and Purdue University that looked at how drug manufacturing in particular is evolving in the digital age.
They wrote: “Together with the digitalization trend of the pharmaceutical industry, the shift of manufacturing paradigms from batch to continuous manufacturing has transformed the focus in process development phase. Companies are now looking at an agile and efficient timeline while maintaining high quality.”
But digitization comes with challenges, according to lead author Marianthi Ierapetritou, PhD, from the department of chemical and biomolecular engineering, University of Delaware, who points to integration as a major hurdle.
“One facet of the problem is that different control software systems do not use the same formats to store and retrieve data from processes,” Ierapetritou says.
Another issue is the volume of data generated on the factory floor, particularly for firms running continuous manufacturing processes.
“Because the processes are operated continuously, manufacturing conditions, material properties, and quality attributes of intermediates and final products need to be strictly monitored and controlled, preferably in real-time.
“Therefore, the processes are considered data-rich, leading to the demand for appropriate process analytical technology (PAT) and other digitalization tools to collect, store, manage, and analyze the data.”
The need for more effective ways of managing and using disparate manufacturing data has not gone unnoticed. The International Society of Automation (ISA), a non-profit technical society working to promote industrial automation, has developed two standards: ISA-88 and ISA-95, which cover process control and automation, respectively.
However, while they are of use, as the adoption of digital technologies accelerates the scale of the data management problem is only likely to increase, Ierapetritou says.
“ISA-88 and ISA-95 are excellent steps and have been adopted but as more and better data are available from different sources, new platforms and integration tools will be needed.”
And process data generation will be a major focus for the industry going forward, particularly for those working at commercial scale, according to Ierapetritou.
“At large scales we are data limited and further development is needed especially on the process modeling front so that we can make better predictions.”