Postgraduate research opportunities Hybrid data-driven and physics-based models for medicines manufacturing
ApplyKey facts
- Opens: Monday 7 October 2024
- Deadline: Thursday 31 October 2024
- Number of places: 1
- Duration: 3.5 years
- Funding: Home fee, Stipend
Overview
The project will explore hybrid modelling of crystallisation processes in pharmaceutical manufacturing. Supported by extensive datasets produced by autonomous robotic platforms, you will combine the advantages of data-driven and physics-based first principles models to enhance the performance of predictive tools for crystal nucleation and growth.Eligibility
Students applying should have (or expect to achieve) a minimum 2.1 undergraduate degree in a relevant engineering/science discipline, and be highly motivated to undertake multidisciplinary research.
Knowledge and/or experience in process modelling, particle technology, material characterisation, data analytics, machine learning and programming skills (e.g. Python, Matlab, PyTorch/TensorFlow) are desirable.
Project Details
The application of Artificial Intelligence (AI) in manufacturing is quickly developing in many fields. In chemical and pharmaceutical manufacturing, AI is seen as an opportunity to correct the significant differential in terms of process efficiency, agility and reliability with respect to other manufacturing sectors such as aerospace and automotive. AI is also a key technology driving the development of more sustainable processes for future medicines manufacturing. Nevertheless, one of the main limitations for the uptake of Machine Learning and Deep Learning approaches in this field is still the significant amount data required in processes characterised by high variability and complex non-linear interactions between multiple variables.
At present, purely data-driven approaches are limited to material design, process operation and fault detection and diagnosis. However, hybrid models can extend the applicability of these approaches by incorporating first principles knowledge to constrain the search space, therefore compensating for the lack of data. Furthermore, Machine Learning methods are useful in finding hidden correlations in complex datasets from which unknown physical meaning can be derived to enhance the predictive ability of first principle models.
The project aims to develop data-driven and first principles hybrid models to enhance the current understanding of nucleation and growth in crystallisation processes and inform robust scale-up strategies. Extensive datasets captured using state-of-the-art autonomous robotic platforms and Process Analytical Technologies (PAT) available at the Centre for Continuous Manufacturing and Crystallisation (CMAC) will inform the design of these hybrid models to incorporate currently unaccounted phenomena such as unexpected phase transitions into population balance modelling. The project will expand the physical understanding of nucleation and growth phenomena to increase process efficiency, reliability and sustainability in pharmaceutical manufacturing, while ensuring the relevance of the research through continuous engagement with CMAC industrial partners.
In addition to undertaking cutting edge research, students are also registered for the Postgraduate Certificate in Researcher Development (PGCert), which is a supplementary qualification that develops a student’s skills, networks and career prospects.
Information about the host department can be found by visiting the Department of Chemical & Process Engineering or our PhD in Chemical & process engineering page.
Funding details
The funding package (including both fees and stipend) is available for UK students.
Supervisors
Dr Javier Cardona Amengual
Chancellor’S Fellow - Senior Lecturer
Chemical and Process Engineering
Apply
Number of places: 1
To read how we process personal data, applicants can review our 'Privacy Notice for Student Applicants and Potential Applicants' on our Privacy notices' web page.
Chemical and Process Engineering
Programme: Chemical and Process Engineering