Post-Doctoral Research Visit F/M Hybrid Modelling and Optimal Control of Microalgae

Inria Sophia Antipolis Sophia Antipolis, France

About the position

Context

Within the framework of a collaborative partnership between the Inria GREENOWL project-team and the DeepTech startup DareWin Evolution (EPIX project)
.
The objective is to develop and integrate a hybrid model
(a digital twin combining mechanistic models and neural networks) as well as optimal control (MPC) algorithms implemented within the ODIN+ supervision software, specifically dedicated to the accelerated Darwinian selection of EPA-overproducing (eicosapentaenoic acid) microalgal strains
.

Assignment

Assignments:
With the support of Olivier Bernard, Francesca Casagli, and Walid Djema (GREENOWL project-team), as well as the R&D team of DareWin Evolution (Hubert Bonnefond and Charlotte Gaviard)
, the recruited postdoctoral researcher will develop and validate a digital twin of microalgal growth (mixotrophic or autotrophic) and lipid/EPA production (combining mechanistic models and neural networks under stoichiometric constraints), design optimal control strategies (PMP, MPC, and online learning) to accelerate minimal-time Darwinian selection in a selectiostat, and implement these algorithms on an experimental platform controlling a photobioreactor via the ODIN+ supervision software
.
For a better knowledge of the proposed research subject:
A state of the art, bibliography, and scientific references are available on demand.
Collaboration:
The recruited person will work in close interaction with researchers from the GREENOWL team, the Laboratory of Oceanography of Villefranche (LOV), and the R&D team of the startup DareWin Evolution
, who conduct the dynamic selection experimental campaigns and generate the biological datasets used to calibrate the hybrid models, experimentally validate the control laws (accounting for 30% of the time), and favor the emergence of EPA-rich phenotypes
.

Responsibilities:

The recruited person will be in charge of the mathematical modelling and optimal control of the EPIX project and will take initiatives to propose innovative hybrid learning architectures, compute optimal dynamic environmental stress trajectories, implement them on the experimental system, and publish the methodological results in top-tier international journals (automatic control, machine learning, and biotechnology)
.
Steering/Management:
The recruited person will be responsible for leading the operational integration of their predictive control and online adaptation algorithms as plugins within the ODIN+ software, as well as coordinating the modelling/experimentation follow-up between Inria (Sophia Antipolis), LOV (Villefranche-sur-Mer), and DareWin Evolution
.

Main activities

Examples of activities:
Propose hybrid modelling solutions and calibrate them using DareWin Evolution's experimental datasets
.
Design optimal control strategies (PMP, MPC, and online learning) to accelerate minimal-time Darwinian selection of overproducing microalgae in a selectiostat
.
Drive experimental campaigns of dynamic selection in photobioreactors using the predictive control algorithms developed and implemented within the ODIN+ supervision software
.
Test, improve, and validate the digital twins and control laws by participating in experimental follow-ups with DareWin Evolution
.
Write methodological scientific articles in top-tier international journals (automatic control, machine learning, biotechnology) and present work progress to the partners
.

Skills

Technical skills and level required :
Languages :
Relational skills :
Other valued appreciated :

This listing was collected from a public source and is reproduced here for information only. Always confirm the details on the original posting before applying.
View the original posting

Similar positions