This research leverages deep generative models to automate and optimize the design of pharmaceutical manufacturing processes. By training on a massive dataset of 385,000 patent-derived procedures, the technology generates plausible, data-driven sequences of manufacturing operations. This approach offers a powerful new tool to accelerate early-stage process development, reduce experimental costs, and support more efficient decision-making in the pharmaceutical industry.
This innovative technology utilizes deep generative models to assist in the complex task of pharmaceutical manufacturing process design. By learning to generate plausible sequences of primary manufacturing operations based on material inputs, this approach aims to reduce development time, minimize costs, and replace reliance on traditional expert-driven heuristics. This capability is particularly valuable for accelerating early-stage process development by offering data-driven suggestions for manufacturing routes.
The core technology features include:
The University of Strathclyde is a public research university in Glasgow, recognised as an international technological university. Industry collaboration is built into the campus through the Technology & Innovation Centre, where companies and researchers work in co‑located labs. System‑scale testing is enabled by the Power Networks Demonstration Centre, while access to the National Manufacturing Institute Scotland provides industrial‑grade facilities and skills development. Anchoring the Glasgow City Innovation District and hosting the Fraunhofer Centre for Applied Photonics, Strathclyde offers direct routes from discovery to deployment; research is backed by UKRI funding, including EPSRC and Innovate UK, with commercialization supported by a technology transfer office and Strathclyde Inspire.