Research

Information science is a scientific field with constantly evolving experimental methods and theoretical foundations which permeates all areas of research. CNRS Informatics occupies a central place in the emergence of new fields of study that are at the crossroads of several disciplines.

Thematics

Research in information science covers a vast thematic field that is constantly expanding through its contacts with other disciplines and the socio-economic world.

Areas of research are evolving and diversifying both at the core of the discipline and at the interfaces.

Research areas at the heart of information science...

  • Fundamental computing and algorithms;
  • Data and knowledge sciences;
  • Security and information protection;
  • Networks and systems (communicating objects, distributed computing);
  • Signal, image and language processing;
  • Autonomous and interactive systems (robotics, human-machine interaction, modelling, control and observation of systems);
  • Systems-on-chips and embedded systems.

... and at interfaces with other disciplines

Interdisciplinarity is at the core of CNRS Informatics's activities. New research directions are emerging at the interfaces with other disciplines and can actually become sub-disciplines in their own right with their own research objects as in the case of bioinformatics. The following areas are already being worked on using a multidisciplinary approach:

  • Bioinformatics;
  • Computing for astronomy;
  • Quantum computation models;
  • Brain modelling;
  • Cognition and interaction;
  • Factories of the future and industrial process modelling;
  • Cities of the future and support and support systems for complex decision-making.

 

CNRS Informatics Thematic reports

          IA                Quantum

Transversal projects

The Institute puts the strengths of its laboratories to work on important transversal issues corresponding to national research priorities, including cybersecurity, AI and quantum computing among others.

 

High performance computing

High-performance computing for digital simulation and the processing of large amounts of data (also known intensive computing or high-performance data analysis) represents a strategic challenge for the production of new scientific knowledge. The scientific competitiveness and attractiveness of the CNRS and of the French academic community in general depend how this field is structured.

In all disciplines, intensive computing has become essential to produce new knowledge. The considerable resources it requires means many research actors are needed to work together.

The Institute for Development and Resources in Intensive Scientific Computing (Idris)

Idris is located on the Université Paris-Sud Orsay campus and is the CNRS's main centre for high-performance intensive numerical computing. This computing centre under the authority of CNRS Informatics makes its exceptional IT resources available to the public sector scientific community. Since 2013, Idris has particularly been able to benefit from using the computing power of the Ada and Turing supercomputers. The 2019 launch of the Jean Zay supercomputer partly dedicated to Artificial Intelligence at Idris confirms the importance of developing powerful computing tools for all scientific disciplines.

The Idris website

The Computing–Data Mission (MiCaDo)

CNRS Informatics plays a leading role in the Computing–Data Mission (MiCaDo) which implements CNRS policy on high-performance computing, grids, clouds and massive data infrastructures. MiCaDo particularly focuses on building a joint strategy with CNRS partners and ensuring this is coherent with national and international projects.

The MiCaDo website

Download the Livre Blanc (White Paper) on data

The Big National Equipment for Intensive Computing (Genci)

CNRS Informatics works on developing its HPC strategy in collaboration with the French Ministry of Higher Education, Research and Innovation (MESRI), CEA, Inria and the Conference of University Presidents (CPU) in the framework of the Big National Equipment for Intensive Computing. This civil company ensures the project management of national computing resources, a Very Large Research Infrastructure (TGIR) located in three computing centres in Bruyères-le-Châtel (TGCC), Montpellier (CINES) and Orsay (Idris).

The Genci website

Artificial intelligence: knowledge is in the data and intelligence in the algorithms

Artificial intelligence has benefited from the development of information technology and recently experienced spectacular growth thanks to the development of new learning algorithms which use very large amounts of data and intensive computing architectures. Beyond learning which crystallizes hopes and questions, AI research covers a wide range of thematics with new fields of application such as robotics and human-machine interaction, autonomous vehicles, support for decision-making, control of and commanding complex industrial systems, personalized health, language processing and text analysis, etc.

CNRS Informatics can draw on the support of a broad internationally visible community and is one of the main partners in national initiatives launched around AI, both in terms of scientific management, structuring and the development of AI-specific computing infrastructures.

The Institute has created a research network (GdR) on the fundamental aspects of artificial intelligence to drive the scientific community made up of its laboratories. This research network also participates in scientific monitoring and national debates on artificial intelligence.

The AI research network website

Cybersecurity: an essential requirement to preserve freedom and democracy

The development of the digital society contributes to citizens' well-being and the fluidity of their exchanges with public authorities and the economic sector but the multiplication of communication channels (the Internet of things, connected homes, self-driving cars, domestic robots, environmental sensors, etc.) has also introduced vulnerabilities which can threaten people's private lives and endanger both the industrial sector and public power. CNRS Informatics is a major player in research and innovation on data protection, network and transaction security, cryptography and the fight against intrusions into sensitive computer systems.

The French "security" community is currently made up of different segmented sub-communities each with their own different facilitation tools. In this context, the Computer Security research network is a unified scientific community working on all aspects of computer security, such as cryptology and encryption, verification, privacy issues, vulnerability and protection mechanisms, physical security, etc.

The Computer Security research network website

Platforms

The Institute runs long-term projects to structure the national research community in information science, in particular through research platforms.

CNRS Informatics guarantees access to high-performance research resources involving both hardware and software through the development of shared research platforms with support staff.

Examples of platforms supported by the Institute

Image removed.

Digiscope

is a unique infrastructure consisting of ten interactive and collaborative visualization platforms interconnected by a telepresence network. These platforms are installed and operational on the campus of the Université Paris-Saclay and on the premises of the project's partners - two universities (Paris-Sud and Versailles Saint-Quentin), three schools (CentraleSupelec, ENS Paris-Saclay and Telecom Paris) and three research organizations (CEA, CNRS, Inria). There are three types of platforms: immersive CAVE-type equipment, "powerwalls" allowing the visualization of 3D data in stereoscopy and very high resolution "hyperwalls".

Digiscope is an "equipment of excellence" financed in the framework of the investments for the future programme.

"Science Heads for the Big Screen(s)" ("Viewing things in a large format", a CNRS film in its Journal date 22/02/2018))

The Digiscope platform website

 

 

Image removed.

SILECS builds on the experience accumulated by the FIT and GRID5K platforms.

This platform aims to provide the scientific community with a tool to explore and test various design hypotheses on both fundamental protocols and technologies (radio networks, 5G/6G, IoT, next generation Internet, Cloud, HPC, Big Data) and on verticals (Industry4.0, Transport, Smart City, Energy, etc.). SILECS aims to position itself at the highest world level and cooperates closely with the international community working on this subject (NSF_PAWR/USA, CENI/China, Fed4Fire+/EU).

The SILECS platform website

 

Image removed.

Robotex is a national network of experimental platforms which provides robotics researchers with a working environment that meets the most demanding international quality criteria. Its specific purpose is to promote synergies between research teams and enable the development of new methodological approaches based on multidisciplinary cooperation. It also has to promote synergies between public research and the industrial, economic and social world.

Robotex is an "equipment of excellence" financed in the framework of the investments for the future programme.

The Robotex platform website

Thematic years

Thematic years are another of CNRS Informatics's instruments for scientific steering.

Since 2015, each year CNRS Informatics has focused on a major theme related to major societal and digital issues. This approach particularly involves the recruitment of researchers, the launch of a call for proposals for Initial Support for Exploratory Projects (PEPS) related to the chosen theme as well as communication with the general public via articles, open forums or popular science writings.

2024 : Optimisation

Cette discipline, omniprésente mais souvent cachée, joue un rôle essentiel dans les disciplines scientifiques, dans de nombreuses applications industrielles et dans notre vie quotidienne. Pour calculer les trajets les plus courts, organiser une production industrielle, concevoir les horaires des transports en commun, ou encore formuler des recommandations personnalisées… il faut des méthodes d'optimisation efficaces et flexibles. L’optimisation se retrouve aussi au cœur de grandes questions sociétales, par exemple pour aider la prise de décisions plus équitables et plus robustes face aux incertitudes d’un monde en pleine transformation. Des algorithmes d'optimisation se retrouvent aussi au cœur des sciences informatiques : en intelligence artificielle, en robotique, en traitement du signal, etc. CNRS Sciences informatiques compte une communauté de recherche forte dans les différentes sous-disciplines de l’optimisation.

Ce focus thématique permettra de mettre en lumière le rôle de l’optimisation, de communiquer sur ses résultats passés et présents, mais aussi de renforcer les interactions entre ces sous-disciplines, d'échanger sur les avancées récentes et de discuter des opportunités et des défis du domaine, par exemple dans le contexte de la disponibilité croissante des données.

2023 : Les sciences informatiques écoresponsables

L’écoresponsabilité est présente dans toutes les communautés scientifiques, et les sciences informatiques ne font bien évidemment pas exception. Il existe cependant une tension inhérente à cette discipline : car face aux enjeux environnementaux, le numérique fait partie à la fois de la solution et du problème.

Il fait partie de la solution puisqu’il est très largement utilisé par de nombreux secteurs pour optimiser la consommation d’énergie ou la gestion des différentes ressources. Mais le numérique fait également partie du problème car les outils ont une consommation énergétique préoccupante et en augmentation constante. 

Les champs scientifiques concernés ici sont très variés, puisqu’ils regroupent l’essentiel des domaines couverts par CNRS Sciences informatiques :

  • Le matériel tels que les capteurs, les centres de données, les antennes, les drones, etc. ;
  • Le logiciel tels que les services numériques, les outils de simulations, les interfaces utilisateurs, etc. ;
  • Les fondements tels que les algorithmes, les techniques d’apprentissage machine, les techniques d’analyse de données, etc.

Ce focus thématique a compté une série d’articles autour des sciences informatiques écoresponsables et s’est clôt par une conférence intitulée « Vers une informatique plus durable », qui rassemblé la communauté scientifique, des industriels et la presse.

2022 : Sciences du logiciel

Les sciences du logiciel recouvrent plusieurs communautés de recherche qui s’intéressent au processus de développement de logiciels et aux outils associés, logiciels qui vont de quelques lignes de code utilisés une seule fois à des milliers de lignes qui vont évoluer, être réutilisées et perdurer pendant plusieurs dizaines d’années. L’objectif est de construire des logiciels corrects, fiables et performants qui évolueront dans le temps et dans l’espace.

Le logiciel touche toutes les facettes de la société : télécommunications, aérospatial, transport terrestre, finance, santé, administration, éducation.

Les recherches en sciences du logiciel visent à faciliter le développement et la maintenance de logiciels informatiques complexes, fiables et durables, tout en prenant en compte les enjeux sociétaux actuels tels que la maîtrise de l’énergie, la globalisation des développements et l’intégration de l’intelligence artificielle. Elles s’étendent des méthodes formelles aux études empiriques en intégrant la construction d’outils, la fouille de codes, la vérification et la validation de logiciels.

Ce focus thématique a donné lieu à une conférence intitulée « Sciences du logiciel : de l’idée au binaire », et à un livret thématique sur les sciences du logiciel.

2020 - 2021 : FutuRobot

Les robots ne cessent de susciter fascination et questionnement. Ces alliés grandissants de notre quotidien se trouvent déjà dans les domaines de la santé, de l'agriculture, de l'automobile, du transport, ou encore de l'exploration spatiale… Leur développement constitue pour beaucoup une promesse de réindustrialisation pour de nombreux pays, comme la France.

Ce focus thématique a donné lieu à plusieurs événements grand public partout en France.

2019 - Information and health sciences

In our digital world, information science plays a leading role in many areas and particularly in the field of health. Here are just a few examples:

  • tomorrow's medicine will use data science to analyze heterogeneous masses of data on diseases, patients, genomics, etc. to propose a personalized, preventive, predictive and participatory approach;
  • traditional medical imaging (EEG, MRI, ultrasound) is regularly being enriched by new, more precise and faster techniques moving towards 3D or 4D imaging;
  • robotics is increasingly providing valuable assistance for surgeons.

Obviously, the development of these new approaches requires multidisciplinary research involving researchers in information science and biological and medical sciences.

The thematic year and some of its different facets will be illustrated by two conferences open to the public and three scientific conferences in 2019 and early 2020.

2018 - Information science interacting with humans

The objective of this thematic year was to emphasize the importance of integrating human factors into the development of digital systems. How is the human factor taken into account in the design, control, safety, operation and evolution of these systems? How should the distribution of tasks and interactions between humans and autonomous systems be modelled and organized? These interdisciplinary questions are central to CNRS Informatics and particularly involve the human and social sciences and life sciences.

The thematic year involved a conference open to the public on the subject of the interaction between humans and the digital sphere.

2017 - Communicating objects

Communicating objects are becoming increasingly common in our daily lives and include smartphones, cars, physical activity monitoring devices, etc. The data they collect, produce and exchange represent multiple technological and scientific challenges as do their specific performance, energy consumption and security limits. Research in this field naturally draws on the full range of topics studied at CNRS Informatics.

2016 - Computer security

Security is present in many forms within the scientific scope of information science - cryptology, verification, communication protocols, vulnerabilities, protection of personal data, searching for weak signals, physical security especially in embedded systems, operational reliability, digital watermarking, etc.

The thematic year involved a conference open to the public on the subject of Computer security: myths and reality. You can find the videos of the conference speeches and presentations on the Sécurité research network website.

2015 - Data Science

Technological advances in data storage and indexing, access to heterogeneous and massively distributed data sources, optimization of such access using parallel architectures and high network throughput make it possible to search for and analyze data within acceptable time scales. This thematic year provided the opportunity to initiate an editorial project which led to the publication in March 2017 of the collaborative book "Les Big Data à découvert" ("Big Data exposed", CNRS Éditions).

Interdisciplinary projects

CNRS Informatics has set up interdisciplinary projects in collaboration with the other CNRS institutes, through bilateral initiatives or within the Mission for transversal and interdisciplinary initiatives (MITI).

CNRS Informatics notably led projects responding to two major challenges, the first involving big data and data science (the Mastodons challenge) and the second image processing (the Imag'In challenge).

The MITI website

The challenge of managing large masses of scientific data (Mastodons)

Mastodons is an emblematic interdisciplinary project launched in 2012 which focuses on the processing and subsequent use of multi-source data. It has resulted in 22 research projects and more than 100 publications. Significant results have been recorded in high throughput genomic sequencing and phenotyping, social network analysis and query optimization on astronomical data.

Find out more about Mastodons

The challenge of finding new imaging methods (Imag'In)

The Imag'In challenge began in 2015 and is aimed at specialists in sensors and complex imaging systems, information processing experts and experts in related application fields. They collaborate to help new imaging methods emerge in multiple fields namely physics, chemistry, biology, medicine, earth and universe sciences, geography, archaeology, etc.

Find out more about Imag’In

Prospective scientifique de CNRS sciences informatiques

CNRS Sciences informatiques a lancé depuis le printemps 2022 un travail de prospective scientifique pour alimenter la réflexion sur l’évolution et les enjeux scientifiques de ses disciplines.

CNRS Sciences informatiques souhaite associer à cette prospective scientifique tout au long du processus :

  • les unités qui lui sont rattachées,
  • les équipes des GDR pilotés par CNRS Sciences informatiques,
  • les responsables scientifiques de PEPR (co)pilotés par CNRS Sciences informatiques,

et plus généralement toutes celles et ceux dont les recherches se situent dans le domaine des sciences informatiques.

Synthèse des éléments de prospective 2022 des GDR

Dans le cadre de l’auto-évaluation du CNRS en 2022, CNRS Sciences informatiques a réalisé une synthèse des contributions des GDR pilotés par l’institut et des retours de son conseil scientifique. Ce document a vocation à être enrichi et complété en 2024 par le rapport de prospective du conseil scientifique d’institut (CSI) et à partir d’une consultation large des unités de recherche rattachées à l’institut. Il s’agit donc du point de départ d’une seconde étape de travail prospectif de plus grande ampleur qui va être mené jusqu’à la fin 2024.

Prospective du conseil scientifique de CNRS Sciences informatiques

Entre 2021 et 2023, le conseil scientifique d’institut a procédé à différents séminaires thématiques pour élaborer un document de prospective qui sera disponible début 2024.

Prospective des GDR

Les GDR réalisent régulièrement des travaux de prospective sur leur champ disciplinaires.