About

I am interested in everything that moves in flows, whether passively (bubbles, particles) or actively (bacteria, animals).

My current research lies at the crossroads of fluid mechanics, artificial intelligence and biophysics. I aim at deciphering the algorithms that living entities use to navigate their complex environment, from immune cells in human tissues to insects in turbulent flows. My goal is also to find ways to outsmart them, with applications to the design of artificial microswimmers and microdrones.

I am the main developer of OTTO, a software dedicated to olfactory navigation and its solutions. It implements an original reinforcement learning algorithm that trains a deep neural network at finding an odor source in turbulence faster than any other known strategy.

Job openings

A fully-funded 3-year postdoc position on cooperation strategies for bio-inspired navigation in complex flows is currently available.

Spontaneous applications for internships, PhDs or postdocs are also welcome.

Highlights

In 2022, we trained a deep neural network to locate an odor source, a task relevant to insects looking for food but also sniffer robots tracking down gas leaks [09]:

• Centrale Marseille Actualités: Entraîner une intelligence artificielle pour faire aussi bien que l’Humain dans la recherche olfactive

On the same year, we showed that planktonic organisms can exploit turbulent fluctuations to migrate much more efficiently in the ocean by ‘surfing’ the flow [11]:

• Le Monde Science & Médecine: Comment le copépode ‘surfe’ pour monter et descendre plus vite dans l’océan, by David Larousserie
• Physics Magazine: Turbulence-Surfing Plankton Can Double Their Speed, by Katherine Wright
• CNRS INSIS Actualités: Pour aller deux fois plus vite, le plancton peut surfer sur les turbulences

In 2018, we demonstrated that mechanical energy can be extracted from an ‘active’ fluid such as a bacterial suspension [05]. This implies that bacteria can be harnessed to power a (tiny) motor:

• Quanta Magazine: Swarming Bacteria Create an ‘Impossible’ Superfluid, by Charlie Wood

Publications

[15] A critical assessment of reinforcement learning methods for microswimmer navigation in complex flows
S. Mecanna, A. Loisy and C. Eloy
under review (2025)
arXiv

[14] Applying Reinforcement Learning to Navigation In Partially Observable Flows
S. Mecanna, A. Loisy and C. Eloy
17th European Workshop on Reinforcement Learning (EWRL) (2024)
PDF DOI

[13] Deep reinforcement learning for the olfactory search POMDP: a quantitative benchmark
A. Loisy and R. A. Heinonen
The European Physical Journal E 46, 17 (2023)
PDF DOI arXiv

[12] The nonlinear motion of cells subject to external forces
A. Ioratim-Uba, A. Loisy, S. Henkes, and T. B. Liverpool
Soft Matter 18, 9008–9016 (2022)
PDF DOI arXiv

[11] Surfing on Turbulence: A Strategy for Planktonic Navigation
R. Monthiller, A. Loisy, M. A. R. Koehl, B. Favier, and C. Eloy
Physical Review Letters 129, 064502 (2022)
PDF DOI arXiv

[10] OTTO: A Python package to simulate, solve and visualize the source-tracking POMDP
A. Loisy and C. Eloy
Journal of Open Source Software 7, 4266 (2022)
PDF DOI

[09] Searching for a source without gradients: how good is infotaxis and how to beat it
A. Loisy and C. Eloy
Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences 478, 20220118 (2022)
PDF DOI arXiv

[08] How many ways a cell can move: the modes of self-propulsion of an active drop
A. Loisy, J. Eggers, and T. B. Liverpool
Soft Matter 16, 3106–3124 (2020)
PDF DOI arXiv

[07] Tractionless self-propulsion of active drops
A. Loisy, J. Eggers, and T. B. Liverpool
Physical Review Letters 123, 248006 (2019)
PDF DOI arXiv

[06] Exact results for sheared polar active suspensions with variable liquid crystalline order
A. Loisy, A. P. Thompson, J. Eggers, and T. B. Liverpool
Journal of Chemical Physics 150, 104902 (2019)
PDF DOI arXiv

[05] Active suspensions have nonmonotonic flow curves and multiple mechanical equilibria
A. Loisy, J. Eggers, and T. B. Liverpool
Physical Review Letters 121, 018001 (2018)
PDF DOI arXiv

[04] The effective diffusivity of ordered and freely evolving bubbly suspensions
A. Loisy, A. Naso, and P. D. M. Spelt
Journal of Fluid Mechanics 840, 215–237 (2018)
PDF DOI HAL

[03] Buoyancy-driven bubbly flows: ordered and free rise at small and intermediate volume fraction
A. Loisy, A. Naso, and P. D. M. Spelt
Journal of Fluid Mechanics 816, 94–141 (2017)
PDF DOI HAL

[02] Interaction between a large buoyant bubble and turbulence
A. Loisy and A. Naso
Physical Review Fluids 2, 014606 (2017)
PDF DOI HAL

[01] Role of meteorological processes in ozone responses to emission controls in California’s San Joaquin Valley L. Jin, A. Loisy, and N. J. Brown
Journal of Geophysical Research: Atmospheres 118, 8010–8022 (2013)
PDF DOI

[PhD] Direct numerical simulation of bubbly flows: coupling with scalar transport and turbulence
A. Loisy
Université de Lyon (2016)
PDF HAL