Overview of the Inria Bordeaux - Sud-Ouest research centre
2 likes1,531 views
The document summarizes research activities at Inria Bordeaux – Sud-Ouest Research Centre in Bordeaux, France. It discusses the centre's 173 project teams, 1,300 doctoral students, 120 startups created, and research focus areas including modeling, high-performance computing, uncertainty management and optimization, modeling and simulation for health and biology, and human and computing interaction and visualization. It also outlines the centre's partnerships with academic and industrial organizations and its role in transferring research to companies.
Overview of the Inria Bordeaux - Sud-Ouest research centre
- 1. Bordeaux _ Sud-Ouest Research Centre
- 2. 173project teams 1 300 doctorants 120start-ups created 229,5 M€ 2015 initial budget 2 700 members of taff 4500 scientific publications 1 200 PhD students different nationalities87 Science and people at Inria Inria, the French National Institute for computer science and applied mathematics, promotes “scientific excellence for technology transfer and society”. Inria’s 2,700 employees, graduates from the world’s top universities, rise to the challenges of digital sciences. Research at Inria is organised in “project teams” which bring together researchers with complementary skills to focus onspecificscientificprojects.Withthisopen,agilemodel,Inriaisabletoexplore original approaches with its partners in industry and academia and provide an efficient response to the multidisciplinary and application challenges of the digitaltransformation.Inria,thesourceofmanyinnovationsthataddvalueand create jobs, transfers expertise and research results to companies (start-ups, SMEs and major groups) in fields as diverse as healthcare, transport, energy, communications, security and privacy protection, smart cities and the factory of the future.
- 3. Inria in the greater south-west region The Inria Bordeaux – Sud-Ouest Research Centre was created in 2008. It is established on the university campuses of Bordeaux and Pau. Toge- ther with its academic and industrial partners, it carries out research activity in the digital sciences and technologies (computer sciences, mathematics). The Centre’s teams focus their research on four areas: • Modeling, High-Performance Computing and Parallel Architectures • Uncertainty management and optimization • Modeling and simulation for health and biology • Human and computing: interaction and visualization Our research teams help to solve fundamental questions posed by computer science and mathematical modeling, the programming of complex distributed systems and the interactions between different actors - both human and artificial. Beyond high-profile scientific advances, which are recognised at the highest inter- national level, the Inria Bordeaux-Sud-Ouest teams play a role in the innovations which, tomorrow, will be at the heart of everyone’s daily lives. By way of example: secure programming of machines regrouping thousands of computing cores, design of applications and software for petroleum engineering, 3D or 4D vizualisation and manipulation of complex objects, design and program- ming of robots, forecasting the evolution of diseases and taking part in advances in the field of bioinformatics... «Our teams - which are recognized and involved at the highest international level - are producing high-profile scientific results and quality applications...Inria Bordeaux-Sud-Ouest wants to help make the Aquitaine region a highly visible scientific cluster in the field of digital sciences.» Monique Thonnat, director of the Inria Bordeaux _ Sud-Ouest Research Centre
- 4. Links with prominent partners The Centre’s scientific development strategy seeks to further the emer- gence of new, ambitious subjects. These are organized in partnership with academics local or international actors at the cutting edge of other disciplines, in order to maximize the impact of the research being car- ried out. > EPSTs (French public scientific and technical Establishement) CNRS, INRA, INSERM > Universities University of Bordeaux, University of Pau and the Adour region > Schools Bordeaux INP, Ensta Paris-Tech, Institut d’Optique Graduate School > Laboratories IMB (UMR 5251), IMN (UMR 5293), LaBRI (UMR 5800), LMAP (UMR 5142), LP2N (UMR 5298) > Institutes Institut Bergonié (U916), ISPED (U897), LYRIC IHU (university hospital institute) > Hospital CHU - Hôpitaux de Bordeaux (Bordeaux university hospital) > Themed research groups Bordeaux Neurocampus, Labex BRAIN and CPU (Idex University of Bordeaux excellence cluster) > International bilateral cooperation (associate teams) Stanford University, University of Colorado Boulder, Universidade de São Paulo, Université Tunis El-Manar, University Mohammed V Agdal Maroc, University of Colorado Denver, Columbia University New York, Indian Institute of Technology Roorkee
- 5. Furthering innovation partnerships and projects A major part of the research that we carry out at the Centre aims to meet to the technological requirements of businesses.Thanks to the support of the Aquitaine region, we maintain an ongoing dialogue with the socio-economic actors of the greater south-west region. To date, the Centre has formed in excess of ten partnerships with SMEs and major industrial groups. In addition, the Centre has a close relationship with the French Alternative Energies and Atomic Commission in CESTA site within the framework of research projects carried out on the Megajoule Laser, the most powerful in the world. The Centre is also forging close links with a network of regional partners, including competitiveness clusters, innovation agencies and themed associations.
- 6. Modeling, High-Performance Computing and Parallel Architectures The distinctive feature of this field is without doubt the continuumthatexistsbetweentheworkthatiscarriedout-which goes from applied mathematics to computer science research for data-intensive computing. In practical terms, the teams design and develop original modeling and simulation or high-perfor- mance computing methods, and produce software that can meet industrial requirements. 8 8 7 4 2 7 7 4 9 < < < 6
- 7. t e a m COMPUTATIONAL AGILITY FOR INTERNAL FLOW SIMULATION -COMPARISON WITH EXPERIMENT This project aims at developing appropriate tools for studying the turbulent flows that are encountered in internal aerodynamic such in jet engine combustion chambers or in the cooling loop of a power station. The originality of this project stems from the simultaneous and in-situ strong coupling between experiment, physical modeling and simulation. Fluid Dynamics • Direct Numerical Simulation • Finite Elements • Turbulence Modeling • Experiments • Internal Aerodynamic • Numerical Methods • Parallel Solvers • High performance computing MAVERIC: Test bench for the study of turbulent flows with mass transfer at the wall
- 8. t e a m CERTIFIED ADAPTIVE DISCRETE MODELS FOR ROBUST SIMULATIONS OF COMPLEX FLOWS WITH MOVING FRONTS From partial differential equations to certified computational models, this team aims at providing a new methodology to develop and certify adaptive numerical models for compressible and incompressible flows with moving fronts. Numerical Methods • Mesh Adaptation • Uncertainty Quantification Fluid Mechanics • High Performance Computing Simulation of Tsunami Runup onto a ComplexThree-dimensional Beach
- 9. t e a m HIGH-END PARALLEL ALGORITHMS FOR CHALLENGING NUMERICAL SIMULATIONS The goal is to contribute to the design and the implementation of numerical methods and associated algorithmic tools to solve complex large scale scienti- fic problems on future extreme scale machines. This requires real scalability to efficiently use a huge number of possibly heterogeneous computing cores. High-End Computing for Frontier Simulations • Algorithmics • Linear Algebra Fast Multipole Method • Code Coupling • Materials Physics Modeling, simulation and high performance computing
- 10. t e a m LITHE AND FAST ALGORITHMIC NUMBER THEORY Algorithmic number theory, dating back to the dawn of mathematics, is being boosted by the arrival of more and more powerful computers. The team covers all its aspects, from complexity theory over optimized implementations up to cryp- tologic applications. Algorithmic number theory • Number fields • Function fields • Algebraic curves Cryptology Researchers of Lfant research team
- 11. t e a m ADVANCED 3D NUMERICAL MODELING IN GEOPHYSICS The team was created to apply recent advances in three-dimensional scien- tific computing to various areas in geophysics, and particular seismic wave propagation. First, advanced models must be developed in order to take the complexity of underlying physical phenomena into account. Second, these models can be applied to realistic cases that require the solution of large systems, which in turn implies that numerical methods must to be optimized. Waves • Numerical Methods • Inverse Problem • Geophysics High Performance Computing • Finite Elements • Multiscale Models Simulation of seismic wave propagation underground
- 12. t e a m MODELING ENABLERS FOR MULTI-PHYSICS AND INTERACTIONS The project aims at a new leap in numerical modeling in order to answer current industrial needs. The goal is to implement these new models in efficient codes on HPC infrastructures and to make them available to respond to societal needs. We do this by developing two fundamental catalyzers: reduced-order models and Cartesian grid methods. Thanks to these catalyzers it will be possible to transfer complexity handling from engineers to computers, providing fast, on-line numerical models for design and control. Numerical modeling • Scientific Computation Fluid Dynamics • Fluid-structure Interaction Numerical model of an inertial system for conversion of wave energy: air-water-floating body interaction with feedback to the inertial system
- 13. t e a m STATIC OPTIMIZATIONS, RUNTIME METHODS The increase of parallelism and of heterogeneity presents many challenges in terms of parallel programming, code optimization and runtime systems. Among these challenges, some are particularly difficult to solve with only static methods (at compile time) or dynamic methods (at runtime). These topics are at the heart of the team’s work. This includes issues in terms of programmer productivity, sca- lability and efficiency of parallel codes. Parallel programming languages • Compilers and runtimes • Heterogeneous programming • High Performance Computing • Adaptability • Scalability Researcher from Storm team
- 14. t e a m TOPOLOGY AWARE SYSTEM SCALE DATA MANAGEMENT FOR HIGH PERFORMANCE COMPUTING The goal of the project is to design and build a stateful system-wide optimization service layer for HPC systems. The challenge is to gather the needs of the applications (memory, computing, storage, network) in order to optimize the execution of all the running applications in a coordinated fashion and at system scale. High Performance Computing • Scalability • Topology • Data • Locality • Systems scale • Data Management Mesh Communication matrix for high performance computing placement
- 15. Uncertainty management and optimization The development of stochastic and deterministic methods for the modeling, optimization and management of uncertainty is at the heart of scientific activity in this area of research. Applicative requirements cover a wide variety of fields: analysis of cardiac signals, path control, scheduling of jobs, satellite imaging, reliability of industrial systems, analysis of rare events... The teams provide novel solutions in order to understand, predict and manage such systems as part of industrial or academic par- tnerships. 8 8 7 4 2 7 7 9 < < < 6
- 16. t e a m QUALITY CONTROL AND DYNAMIC RELIABILITY The core component of our scientific agenda focuses on the development of statistical and probabilistic methods for the modeling, evaluation and the optimization of complex systems. Markov chain • Piecewise Deterministic Markov Processes • Markov Decision Processes • Network • Queeing system • Dimension Reduction Models • Stochastic control • Semi parametric and non parametric modeling • Multivariate data analysis Optimal control of a submarine
- 17. t e a m GEOMETRY AND STATISTICS IN ACQUISITION DATA The team develops methods to study complex signals with multiscale properties, such as heartbeat signals, satellite and space observation data, speech. Derived from the study of disordered systems in Statistical Physics, these methods open novel approaches in the analysis of complexity in natural signals. Signal processing • Non-linear methods • Comple systems • Turbulence Multiscale methods • Complexity • Scale invariance • Complex signals Speech processing • Adaptative optics Complex signals studies : ocean movements and heart
- 18. t e a m REFORMULATIONS BASED ALGORITHMS FOR COMBINATORIAL OPTIMIZATION Our team develops efficient solvers for complex combinatorial optimization problems by virtue of tight formulations. We are specialized in extended formula- tion and decomposition approaches in mathematical optimization. Through indus- trial partnerships, the team targets large scale problems such as those arising in logistics (routing problems), in planning and scheduling, in network design and control, and in placement problems (cutting stock problems). Operations research • Combinatorial optimization • Graph • Decomposition Branch-and-price • Branch-and-cut • Primal heuristics Researchers from Realopt team
- 19. Modeling and simulation for health and biology The teams involved in this area carry out research in applied mathematics or computer science for biological or medical modeling. Thanks to solid partnerships with other research actors from clinics and education, we are already witnessing major results in a variety of subjects, such as the understanding of neurodegenerative diseases, epidemiology, cardiac electro- physiology and oncology. 8 8 7 4 2 7 7 9 < < < 6
- 20. t e a m MODELING AND NUMERICAL METHODS FOR CARDIAC ELECTROPHYSIOLOGY The team develops innovative models and numerical methods to simulate the propagation of the cardiac action potential, from the scale of the cell to the scale of the body. It aims at improving the knowledge and the treatment of electrical cardiac pathologies as well as the exploitation of all available electrical signals. Multiscale Models • Numerical Methods • Inverse Problems • Scientific Computation High Performance Computing • Biomedical Engineering Researchers of the CARMEN research team
- 21. t e a m SYSTEMIC APPROACH TO BRAIN MODELING At the frontier between integrative and computational neuroscience, the project- team proposes to model the brain as a system of active memories, in synergy and in interaction with the internal and external world, and to simulate it both as a whole and in situation. Computational neuroscience • Machine learning • Autonomous robotics Decision making • Memory • Adaptative systems Researchers of MNEMOSYNE research team
- 22. t e a m MODELING IN ONCOLOGY Thepurposeofthisteamistodevelopnumericalmodelsbasedondatainoncology in order to provide decision support to clinicians and biologists. The problem of modeling of tumor growth from data is addressed in three ways: custom models derived from medical images in clinical applications; biophysical modeling for therapies; and finally models for pre-clinical studies. Scientific Computation • Modeling • Multiscale Models • Computational Biology Oncology modeling
- 23. t e a m PATTERNS OF DIVERSITY AND NETWORKS OF FUNCTION The study of biology associates pattern recognition of diversity with modeling of functional and evolutionary processes. Pleiade addresses the double challenge of measuring dissimilarity between biological objects quickly and precisely, and exploring the relations between diversity in traits and diversity in function at multiple scales. We develop algorithms, models, and software frameworks for applications in ecology, evolution, and biotechnology. Pattern recognition • Modeling • E-Science • NGS • Biodiversity • Biotechnology Researchers of the Pleiade team
- 24. t e a m STATISTICS IN SYSTEMS BIOLOGY AND TRANSLATIONAL MEDICINE The team is devoted to the development of statistical methods for the integrative analysis of data in medicine and biology. Thanks to technological improvements, clinical and biological research is generating massive amounts of data. The challenge is to analyze these big data to answer clinical and biological questions in immunology by using appropriate statistical methods. Statistics • Systems Biology • Stochastic Modeling • Statistical Learning Epidemiology • Big Data • Immunology • Infectious diseases Transmission of HIV-1 from cell to cell
- 25. Human and computing : interaction and visualization The distinctive feature of this area is the consideration of hu- mans and their interactions with digital technology. From re- search on image synthesis to human-machine interaction, the teams are working - among other things - on robotics, hybrid optical and digital instrumentation and intelligent interfaces to help people. These subjects meet a strong societal demand and research findings notably come to life through start-ups or in partnership with SMEs. 8 8 7 4 2 7 7 9 < < < 6
- 26. t e a m FLOWING EPIGENETIC ROBOTS AND SYSTEMS The project studies mechanisms that can allow robots and humans to acquire, autonomously and cumulatively, repertoires of novel skills over extended periods of time. This includes mechanisms for learning by self-exploration, as well as learning through interaction with peers, for the acquisition of both sensorimotor and social skills. The project also studies how these models and technologies can be used in new educational tools. Developmental and social robotics • Situated and embodied cognition • Exploration Learning • Intrinsic motivation • Natural human-robot interaction • Education Learning to walk by Poppy, bio-inspired humanoid robot
- 27. t e a m MELTING THE FRONTIERS BETWEEN LIGHT, SHAPE AND MATTER The project studies the interactions between light, form and matter, in order to achieve new representations of appearance. It also takes in consideration exchanges from the real world to the virtual world (acquisition) and from the digital world to the end viewer (restitution). Representations, algorithms and systems combining optics and computer science will allow direct control of appearance. Through effective communication between the real world and the 3D digital world, new uses (scientific, artistic, industrial) will emerge. Computer Graphics • 3d Modeling • Geometry Modeling • Rendering Augmented Reality • Virtual Reality 3D reconstruction of the statue of a giant from beacon of Alexandria
- 28. t e a m PROGRAMMING LANGUAGE TECHNOLOGY FOR COMMUNICATION SERVICES More and more, communicating objects equip our living spaces and render them digital. The scientific challenges consist in designing languages and tools that assist the development of these spaces by ensuring reliability and security. The project has developed applications especially for cognitive support for people undergoing loss of autonomy. Programming languages • Object oriented programming • Verification Software engineering • Pervasive computing • Assistive technology Human-computer interaction • Cognitive science An experimental apartment and an assisted living platform
- 29. t e a m POPULAR INTERACTION The overall goal of Potioc is to explore new approaches that foster rich interaction with the digital world through appealing and motivating interfaces. The ultimate goal is to stimulate creation, learning, or entertainment. To this end, we focus on the design, development, and evaluation of new methods for 3D interaction by non specialists. Human-Computer Interaction • 3D User interfaces • Virtual Reality • Augmented Reality • Tangible interaction • Brain-computer interfaces • Physiological computing Cognitive sciences Collaborative visualization of mental states using spatial augmented reality and tangible interaction
- 30. t e a m MODELS FOR FUNCTIONAL PROGRAMMING OF SPACE AND TIME The interaction between art, science and technology is a constant in the history of mankind. The development of digital technologies has catalyzed that interaction. The project aims to develop concepts, methods and practices that simplify the synchronization, coordination and orchestration of real-time analysis and synthesis tools dedicated to the design of complex interactive multimedia environments. Computational music • Algebraic modeling • Functional programming Geometric representation of polyrhythmic sequences