Tenure-track faculty: Machine-learning for Dynamical Systems

The aim of this faculty position is to set up a research program around the theme of explainable and physically-informed AI for modeling dynamic systems, and to steer a strong and original vision for the place of AI in the SUPMICROTECH curriculum. This research program aims to develop the next generation of machine-learning tools for controlling complex physical or physiological systems, by using a data-driven approach to discover physically interpretable models of dynamic systems based on temporal data.

For example, the project could fit in with the recent emergence of data-driven methods for identifying systems, such as reverse-correlation, from the automatic control and biological systems modeling communities (Daube et al. Patterns, 2021), symbolic regression methods (SINDY; Brunton et al. PNAS 2016) or the learning of physically constrained representations (DMD; Schmid Annual Review of Fluid Mechanics, 2022). This work could be applied to the modeling of biological biological systems, to provide new diagnostic or prognostic tools in e.g. neuroscience (Durstewitz, Koppe & Thurm, Nature Reviews Neuroscience 2023), cancerology, or new micromanipulation techniques for cellular cell characterization or surgery.

The recruited faculty is expected to join the Dept of Automation and Robotics (Automatique et Systèmes Micromécatroniques, AS2M) of the FEMTO-ST Institute, of which SUPMICROTECH is one of the host operating institutions. Research conducted within the AS2M department is based on a multi-disciplinary foundation combining mechatronics, automation and data science. Machine learning for data-driven estimation of dynamic systems is a strongly emerging field in machine-learning (see, for example, the recent AI Institute in Dynamic Systems at the University of Washington in Seattle, the L4DC cycle of international conferences, or the Deep Learning for Physical Processes research chair at Sorbonne Université). This research area is central to the FEMTO-ST Institute’s recent work, whether in the control of innovative robotic architectures robotic architectures (e.g. flexible and dexterous robotics), modeling unconventional physical systems at micro/nanometric scales (e.g. machine-cell interaction) or the prediction of complex dynamic systems in the fields of environment and health (e.g. identifying dynamical systems from neurobiological data). By bridging the gap between data science and dynamical systems theory, the project will be able to give rise to new cross-disciplinary projects combining, for example, modeling and control control or control and prediction. Finally, the project will reinforce the department’s strong recent momentum in the field of micro/healthcare technologies, with applications at the level of the cell (cancer cell characterization), organ (sensors for oncology diagnostics), or organism (neurosciences). (neuroscience). If concerned with neuroscience, the research project may benefit in particular from the Dept’s human EEG experimental platform (https://neuro-team-femto.github.io).

The position being funded at a national level, applications should be submitted solely via the Galaxie portal at : https://www.galaxie.enseignementsup-recherche.gouv.fr/ensup/cand_CPJ.htm. Please note the Galaxie portal is mostly in French and, depending on your language level, filling in the information and various required files may be error-prone. Please plan accordingly, and seek advice from the contact persons below.

Job reference: 4094
Job description: https://www.galaxie.enseignementsup-recherche.gouv.fr/ensup/ListesPostesPublies/FIDIS/0250082D/FOPC_0250082D_4094.pdf
Deadline: 15th May 2024.

A pre-selection of applications will be made by the search committee. Successful candidates will be invited for an interview, which will include a presentation of their research and teaching project, as well as a mock teaching session, the details of which will be specified in the invitation.

Before submitting their application, we strongly encourage applicants to make contact with the FEMTO AS2M and SUPMICROTECH research teams in order to adapt their research and teaching statements to the strategic directions of the two institutions:

• General contact for research:

Prof. Yann Le Gorrec, Department Chair, Dept. of Automation and Robotics, FEMTO-ST Institute. yann.le.gorrec@ens2m.fr

• Contact for neuroscience applications if relevant:

Dr Jean-Julien Aucouturier, PI Neuro group, Dept. of Automation and Robotics, FEMTO-ST Institute. aucouturier@gmail.com

• Contact for teaching statement:

Prof. Christophe Varnier, Head of Studies, SUPMICROTECH/ENSMM. christophe.varnier@ens2m.fr