The three-year, EPSRC-funded project, named ‘Right First Time Manufacture of Pharmaceuticals’ (RiFTMaP), will begin this September and is a collaborative effort between Sheffield, UCL and Strathclyde University, and Purdue University in the USA.
Professor Jim Litster, vice president and head of the Faculty of Engineering at Sheffield University, explained that the project’s overall mission is to bring a process systems engineering approach to continuous pharmaceutical manufacturing, resulting in benefits such as reduced time-to-market for products, reduced waste, increased resilience and reduced cost.
“Ultimately, we want to support a vibrant and sustainable pharmaceutical manufacturing industry in both the UK and the USA,” he told The Engineer. “Right-first-time comes from the idea that if we put together a continuous manufacturing plant, it has a lot of different advantages in terms of efficiency and product quality, but it’s different from the traditional approach of the pharmaceutical industry which is batch processing, and a lot of what you'd call quality by testing.”
Litster explained that through the continuous manufacturing approach, the pharmaceutical industry can ensure quality by control throughout the process and, combined with real-time release approaches, predict performance of the product. This will allow for higher efficiency and production of multiple products on demand with minimal risk to quality.
Described as a key element of the project’s potential for success are the manufacturing plants at Sheffield and Purdue: the Consigma 25 wet granulation line at Sheffield, and Purdue’s dry granulation line and continuous direct compression line.
The first of its scale in a UK university, the ‘Diamond Pilot Plant’ (DiPP) at Sheffield's Department of Chemical and Biological Engineering will be utilised alongside Purdue’s platforms to allow researchers to validate advanced models and control and optimise procedures for the development of a verified, pilot-tested framework.
“We’re using hybrid model approaches for design and operation, trying to combine the physics knowledge we have of the formulations of the processes with machine learning approaches to give the best of both worlds — the best of machine learning, which is added strength when it has large data sets to interpret, with physics models that allow you to predict performance based on very small amounts of information,” said Litster.
A range of industry project partners on an advisory board will be critical to transferring the ideas seamlessly into practice, Litster added, with frameworks, approaches and tools developed throughout the project expected to begin supporting industry from around year three.
These include pharmaceutical, software and equipment manufacturers such as GlaxoSmithKline (GSK), AstraZeneca, Pfizer, IBM UK and Process Systems Enterprises amongst others.