$ads={1}
Biopharma process optimization using machine learning (OPPORTUNISTE)
- Sciences de l’ingénieur
- Génie des procédés
- Robotique
Description du sujet
The objective of this project is to explore the potential of machine learning and deep learning methods in the field of biotechnology, particularly in a context where process analytical technologies (PAT) emphasize the importance of mathematical modeling, supervision and control of bioprocesses as essential tools to improve and ensure product quality and consistency. Several problems will be studied, in the context of the culture of mammalian cells, of the CHO type for example, in fed-batch, continuous or discontinuous bioreactors producing monoclonal antibodies, including: a) The use of machine learning and deep learning in bioprocess monitoring. New online probes and new analytical equipments have appeared on the market, which make it possible to follow online the time evolution of the concentrations of certain substrates and culture products. Among them, the Raman probes, which produce a spectrum that can be correlated to the concentrations of the components of interest, have been given great attention. b) The development of hybrid models of bioprocesses, combining a mechanistic structure based on macroscopic reaction diagrams and the mass balances which can be deduced therefrom, on the one hand, and neural networks to represent parts of the system that are more difficult to model by classical structures, on the other hand. c) The use of Real Time Optimization (RTO) to control, in a robust way, the bioprocesses, by overcoming the problems of parametric uncertainties in the models. The latter are indeed difficult to establish and, more particularly, to validate, on a new strain leading to a new product.
Prise de fonction :
Nature du financement
Précisions sur le financement
Présentation établissement et labo d'accueil
Founded in 1837, the Faculty of Engineering of Mons (FPMs) is the oldest Engineering Faculty in Belgium, with 25 research and education departments, approximately 1000 students, and a large panel of bilateral, Erasmus-Socrates and TIME agreements. Since 2010, FPMs is now the Faculty of Engineering of the University of Mons (UMONS).
The SECO Department of the Engineering Faculty of UMONS has a staff of about 20 collaborators.
It ensures undergraduate and graduate education in automatic control. Research has developed in several directions of mathematical modeling, parameter and state estimation, and process control. The group participates actively in two research institutes of UMONS, namely Energy and Biosciences.
SECO has established collaborations with academic and industrial institutions in different parts of the world.
Site web :
Intitulé du doctorat
Pays d'obtention du doctorat
Etablissement délivrant le doctorat
Ecole doctorale
Profil du candidat
The candidate should:
- Be electrical engineer, control engineer or bioengineer/chemical engineer;
- Be fluent in English (being fluent in French is an asset);
- Have a good background education in machine learning is an asset;
- Have good programming skills, for instance, in Matlab, Python and C++.
