• Recherche

ERC Advanced 2019

De nouveaux succès à la Faculté des Sciences et Ingénierie de Sorbonne Université. Quatre des 21 chercheuses et chercheurs récipiendaires français font leur recherche dans des unités mixtes de recherche (UMR)  associées à la Faculté des Sciences et Ingénierie de Sorbonne Université. Félicitations aux quatre récipiendaires !

L'appel ERC Advanced s'adresse à des chercheuses et chercheurs expérimentés et reconnus qui s’engagent sur un projet de recherche novateur.

Conseil européen de la recherche

 

Gérard Assayag, Chercheur IRCAM (STMS1)

(1) Laboratoire Sciences et Technologies de la Musique et du Son

Gérard Assayag, directeur de recherche Ircam, a fondé et dirige actuellement l'équipe Représentation musicale (RepMus) du laboratoire STMS qu’il a dirigé de 2011 à 2017. Il est co-porteur de l'Institut Collegium Musicae de Sorbonne Université et contribue aux activités du Sorbonne Center for Artificial Intelligence (SCAI). Il est co-fondateur de la Société Française d'Informatique Musicale (AFIM), la Society for Mathematics and Computation in Music (SMCM) et son organe, le Journal of Mathematics and Music (JMM). Il a aussi été responsable du master ATIAM, une formation proposée par Sorbonne Université en partenariat avec l'IRCAM. Ses recherches portent sur la modélisation des interactions co-créatives entre humains et machines dans le domaine musical et ont produit des outils informatiques largement utilisés dans le monde par artistes et chercheurs. 

Projet : Raising co-creativity in cyberhuman musicianship (REACH)

Résumé (en anglais) :
Digital cultures are increasingly pushing forward a deep interweaving between human creativity and autonomous computation capabilities of surrounding environments, modeling joint human-machine action into new forms of shared reality involving "symbiotic interactions”. In the artistic, cultural or educative fields, co-creativity between humans and machines will bring about the emergence of distributed information structures, creating new performative situations with mixed artificial and human agents. This will disrupt known cultural orders and significantly impact human development. Thanks to the computation of semantic structures from physical and human signals, combined with generative learning of symbolic representations, we are beginning to comprehend the dynamics of cooperation (or conflicts) inherent to cyber-human bundles. To this end the REACH project aims at understanding, modeling, and developing musical co-creativity between humans and machines through improvised interactions, allowing musicians of any level of training to develop their skills and expand their individual and social creative potential. Indeed, improvisation is at the very heart of all human interactions, and music is a fertile ground for developing models and tools of creativity that can be generalized to other activities, as in music the constraints are among the strongest to conduct cooperative behaviors that come together into highly integrated courses of actions. REACH will study shared musicianship occurring at the intersection of the physical, human and digital spheres as an archetype of distributed (natural / artificial)  intelligence, and will produce models and tools as vehicles to better understand and foster human creativity in a context where it becomes more and more intertwined with computation.

 

 

Jacques Laskar, Directeur de Recherche CNRS (IMCCE2)

(2) Institut de Mécanique Céleste et de Calcul des Éphémérides

Jacques Laskar est directeur de recherche au CNRS, à l'Institut de Mécanique Céleste et de Calcul des Éphémérides (IMCCE) qu’il dirige au sein de l'Observatoire de Paris. Il travaille sur la dynamique des systèmes planétaires. Il a mis en évidence le mouvement chaotique des planètes du Système Solaire, avec un horizon de prédictibilité de seulement  60 millions d'années. Il a aussi montré que l'axe de rotation de Mars est fortement chaotique et que l’axe de la Terre ne doit sa stabilité qu'à la présence de la Lune. Il est membre de l'Académie des Sciences et du Bureau des Longitudes. 

Projet : Astronomical Solutions over Geological Time (AstroGeo FR)

Abstract
According to Milankovitch's theory (Milankovitch, 1941), part of the great climatic changes of the past is due to the variations of the insolation on the surface of the Earth resulting from the deformation of its orbit resulting from the gravitational disturbances of the other planets. These variations can be found in the stratigraphic records accumulated over several million years (Ma). The correlation between the geological data and the calculations of celestial mechanics is sufficiently established so that since 2004, the geological time scales of the most recent periods are established from the astronomical solutions developed at IMCCE (Laskar et al, 2004) (http://vo.imcce.fr/insola/earth/online/earth/earth.html). Since then, a major effort has been devoted to extend this astronomical calibration over ever longer periods covering the entire Cenozoic, up to 66 Ma. In these works, the astronomical solution, given by calculation, is used to calibrate the geological data.

However, extending this work is difficult because celestial mechanics does not allow us to retrace the planetary orbits beyond 60 Ma due to the chaotic nature of the movement of the planets, highlighted by Jacques Laskar   thirty years ago. The AstroGeo project aims to overcome this predictability horizon, imposed by the laws of gravitation by using ancient geological data as an additional constraint in obtaining astronomical solutions. In this ambitious project which makes it possible to reconstruct the orbits of the planets of the solar system, the limit is no longer imposed by the exponential propagation of uncertainties, with an almost impassable horizon at 60 Ma, but by the quality of the sedimentary data which could be used until very remote times, going as far as the Archean, 3 billion years ago.

The feasibility of the AstroGeo project was demonstrated with the publication in PNAS (Olsen et al, 2019) by an international team including Jacques Laskar, of a study showing how the analysis of sediment data from the Newark basin had made it possible to find the state of the solar system 200 Ma ago (https://www.imcce.fr/news/mouvements-planetaires-geologie).

AstroGeo will more generally benefit from the very numerous collaborations established for three decades by Jacques Laskar with geologists around the world, with in particular in France, the ISTEP laboratory (Institute of Earth Sciences in Paris, UMR 7193), partner of IMCCE as part of the National Agency for Research AstroMeso project started in October 2019. AstroMeso is part of the wider AstroGeo project, and allowed to recruit a post-doctoral researcher to study sediment data from the Paris basin and the Vocontian basin in the South-East of France.

Since the first long-term solution of the movement of the planets established by Urbain Le Verrier in 1840, and used by Milutin Milankovitch a century later to establish the astronomical theory of climates, the Paris Observatory has provided most of the planetary solutions which were used to correlate changes in insolation on the Earth's surface with glacial or sedimentary records. AstroGeo aims to continue this tradition of excellence by transforming the way solutions will be distributed to the paleoclimate community. Instead of producing a single reference solution which serves as the basis for establishing the geological time scale, which is no longer possible beyond 60 Ma, AstroGeo will provide a set of solutions, all compatible with the best observations from the moment, also giving users the possibility to search among these solutions, those which will best correspond to the newly collected data. The solution could then be constrained by these new stratigraphic data.

 

 

Frédérique Le Roux, Directrice de Recherche Ifremer (LBI2M3)

(3) Laboratoire de Biologie Intégrative des Modèles Marins

Frédérique Le Roux s’intéresse à l'étude des mécanismes moléculaires impliqués dans l'émergence de la pathogenèse Vibrio chez les invertébrés marins. Ces mécanismes comprennent l'épidémiologie moléculaire, en particulier la biodiversité, la virulence et la plasticité des génomes, la génomique environnementale et les interactions hôte/pathogènes. Elle est membre de la rédaction de la revue Environmental Microbiology.

Projet : A mechanistic approach to understand microbiome-viriome dynamics in nature (DYNAMIC)

Résumé (en anglais) :
Facing the therapeutic impasse of antibiotics, farming systems, among which aquaculture, should consider the extraordinary resource of phages, the natural bacterial predators, for environmental friendly practices. It is, however, crucial to understand how phages can control pathogens in a sustainable and safety manner. The goal of this project is to shed light on key ecological and evolutionary processes underlying phage-bacteria dynamics in the marine environment. An oyster bacterial pathogen, Vibrio crassostreae and its infecting phages will be used as model system to investigate the molecular bases and evolution of phage infections in nature. Based on a field approach, we will determine whether phages influence V. crassostreae dynamics in the wild by reducing bacterial density via predation and if co-evolution applies in this natural system. Combining comparative and functional genomics we will identify genes involved in the phage host range, host resistance, and phage–host coevolution. Exploring phage-vibrio interactions in culture, we will analyze whether fitness costs can constrain evolution of resistance in oyster hemolymph. We will identify vibrio virulence genes that are negatively selected by phages. In addition we will study whether phages in combination act synergistically to control V. crassostreae. Focusing on a T4-giant phage as a model, we will assess the molecular mechanisms underpinning its broad host range and decipher its potential to spread bacterial genes by horizontal gene transfer. We will finally revisit the phylogenomics of T4-related phages, and reconstruct the T4-giant phage ancestral genome to determine how the ability to infect multiple hosts has evolved in this group. This project has significant potential to make truly ground breaking discoveries on phage-bacteria coevolution providing new and major knowledge for the future generation of phage therapy in aquaculture.

 

 

Stéphane Zaleski, Professeur Sorbonne Université (JLRA4)

(4) Institut Jean Le Rond d'Alembert

Stéphane Zaleski est professeur des universités à Sorbonne Université où il mène ses recherches au sein de l’Institut Jean Le Rond d'Alembert. Il y développe des méthodes numériques adaptées à l'étude des milieux multiphasiques, que ce soit sous l’angle des phénomènes d’écoulement, des problèmes de transfert de masse ou de changement de phases. Ses recherches les plus récentes concernent l’apprentissage automatique de méthodes numériques ainsi que des méthodes implémentées dans des codes libres en collaboration avec des sociétés d'ingénierie. Il est membre honoraire de l’Institut Universitaire de France (promotion 1995).

Projet : TRansfers at tiny scales in tUrbulent multiphase FLOW (TRUFLOW)

Résumé (en anglais) :
The prediction of heat and mass transfer across fluctuating fluid interfaces is a considerable challenge. It is however not only an ubiquitous part of industrial processes, but also a critical component of the global climate system through ocean-atmosphere interactions. Sustainable development and greenhouse gas emission containment will require an overhaul of already knowledge-intensive processes. TRUFLOW thus aims at enabling the quantitative prediction of the heat and mass transfer in fluid flow using simulation, high performance computation and multiphysics, multiscale methods. Using presently available, cutting edge interface tracking and subgrid scale methods TRUFLOW will investigate a range of critical processes, allowing for example industry to plan for improved carbon capture processes such as rotating packed beds, new processes such as hydrogenbased metallurgy to replace carbon based metallurgy, heat and mass transfer in hydrogen fuel cells, boiling and cavitation simulation and CO2 transfer across the wavy ocean surface. The key limiting factor in the success of simulation in this domain is the considerable range of scales expected, with slowly diffusing chemicals creating boundary layers that are orders of magnitude smaller than the typical fluid structures, bubbles or droplets. Critical heat fluxes in boiling and interface motion at the microscale are determined by contact line motion, which involves tiny molecular scales. TRUFLOW will bridge these various extreme length scale gaps using state of the art methods. It will result in direct high performance simulations of heat and mass transfer, coupled simulation and analysis of existing experimental data, an analysis of the performance of reduced models of flows with tiny scale transfers, and a systematic use of these models in industrial or natural configurations.