Laboratories jointly run with companies

In the same way as other joint research structures like CNRS/Companies Joint Research Units (UMR) or International Research Laboratories (IRL), associated research laboratories take traditional partnerships further in the form of collaboration contracts. They go beyond classic short-term research projects with objectives that are regularly updated through ad hoc governance. Associated research laboratories operate on the basis of collaboration contracts that provide a framework to structure projects with high scientific and industrial stakes. These generally involve one or more research team(s) and an industrial concern working together on a shared medium- or long-term research programme.

Arès, Machine Learning And Computer Vision For Intelligent Robots

ANR LabCom. Partners: LIRIS / Siléane

The increasing automation of dangerous, tedious and repetitive tasks is a symbol of technological progress and essential for effective productivity levels and competitiveness. The sustainability of an industrial site partly depends on its flexibility and the scalability of its production tools in a globalized market context with strong competition and rapidly changing requirements.

The objective of this Labcom project is to develop new machine learning and computer vision methods to create flexible, adaptable, autonomous robotized tools able to guarantee high productivity and manage ultra-flexibility as required in many industrial applications.

The aim is to contribute to improving the performance of the task of sorting and placing products in a context of high diversity and high flexibility. This task is often very tedious and repetitive and is found across all industrial manufacturing processes (foundry, chassis, machining, stamping, sheet metal, assembly)

To achieve these aims, the Labcom ARES project combines two complementary partners who will pool their research, innovation and development capacities and make their already strong cooperation on R&D official:

Image removed.
© Siléane

Behaviors.ai, an artificial intelligence engine which enhances the verbal and non­verbal interactions of

ANR LabCom. Partners: LIRIS / HOOMANO

The mission of the joint laboratory Behaviors.ai (Behaviors.ai is an Engine enHancing verbAl and non­Verbal InteractiOns of RobotS, based on Artificial Intelligence) is to develop an intelligent interaction engine capable of generating human-robot interactions and making them more empathic, intuitive and natural.

This project brings together artificial intelligence researchers from the Computer Science Laboratory for Image Processing and Information Systems (LIRIS - CNRS/ Lyon INSA /Université de Lyon/École Centrale de Lyon) and HOOMANO, a start-up company which specializes in developing software for social robots. The objective of Behaviors.ai is to succeed in transferring fundamental research results, notably in developmental learning, to the business-to-business markets HOOMANO works in.

 

Image removed.
© Stéphane Rambaud

 

Cherchons pour Voir, des solutions interactives et inclusives au service des déficiences visuelles

Laboratoire commun (LCR). Partenaires : IRIT / CESDV Institut des jeunes aveugles

Le laboratoire commun Cherchons pour Voir se situe à l’interface entre le monde professionnel de la déficience visuelle et celui de la recherche. Il vise notamment à développer des technologies d’assistance pour améliorer l’autonomie et la qualité de vie des personnes non voyantes et malvoyantes. Les projets se consacrent, entre autres, à l’élaboration de cartes interactives d’aide à la localisation, des dispositifs numériques pour faciliter l’accès à l’information, l’éducation ou encore la culture.

Sites internet des partenaires :

FiT, lutter contre la criminalité financière

Joint laboratory. Partners: LIP6 / Bleckwen

Over the last years, financial crime became a very expensive plague slowing down the advent of the digital society. In 2019, the cost of financial crime was estimated at 1.45 trillions dollards.

Within this context, fighting against frauds, identity theft, or money laundering is a major issue. This fight relies on the analysis of massive sequences of financial transactions aiming at detecting traces of illegal activities such that proper counter measures can be undertaken. Nonetheless, advances in this fight are hindered by a ruthless lack of methods and tools to model and analyze the transactions in a suitable way.

The FiT labcom's primary objective is to lead the scientific and technological advances required to significant progresses in anomaly detection in financial transactions.

Bleckwen and the team ComplexNetworks wish to join their expertise to model and analyze financial transactions as link streams (see image below). They are currently developping and implementing the formalisms and algorithms enabling the proper exploitation of these data.

FiT
© FiT

 

IA LUSIS, des solutions innovantes en intelligence artificielle pour le domaine bancaire

Chaire industrielle (CHIR). Partenaires : LISN / LUSIS

La chaire IA LUSIS entre l’entreprise Lusis et CentraleSupélec se consacre à l’application de méthodes d’l’intelligence artificielle au domaine bancaire. Elle développe des algorithmes pour la détection de fraude par carte bancaire, soit moins de 0,5 % de ces transactions mais une perte de plusieurs milliards d’euros par an. La chaire cible également l’application d’une méthode d’apprentissage par renforcement afin de réaliser du trading automatisé consistant à acheter une action pour la revendre à meilleur prix.

Sites web des partenaires :

Image4US : une imagerie de pointe pour la recherche clinique

Labcom ANR. Partenaires : CREATIS / DBSAS

Bien que l’imagerie par ultrasons 2D soit largement répandue à travers le monde, l’échographie 3D peine, quant à elle, à fournir des images de qualité à haute cadence. Le laboratoire commun Image4US s’est ainsi donné pour objectif de développer un dispositif d’imagerie médicale ultrasonore 3D de précision et à grand nombre de voies (1024). Le système proposé se distingue par ses caractères ouvert, flexible et transportable À terme, un tel outil se destinera à accompagner les chercheurs dans le cadre de recherches cliniques pour une meilleure identification des structures anatomiques ou encore une estimation plus précise du flux sanguin.

Sites web des partenaires :

PosiLab, Precision Parallel Positioners and Redundancy

ANR LabCom. Partners: LIRMM / Symétrie

The PosiLab LabCom aims to demonstrate the potential of redundant parallel positioner robots for precise positioning applications which are demanding in terms of performance (high rigidity, large workspace, high stability) in the fields of optics and space. The LabCom works in a tense economic and industrial context with a high level of competition and where competitiveness is vital. This LabCom brings together researchers and engineers from the Laboratory of Computer Science, Robotics and Microelectronics of Montpellier (a University of Montpellier/CNRS joint research unit) and Symétrie (an SME based in Nîmes). PosiLab has the ViaMéca competitiveness cluster label.

The PosiLab LabCom finds answers to the questions of "how to design optimally, then control robustly positioner robots capable of high precision (below micron and arcsecond) while (i) remaining insensitive to external disturbance and (ii) providing a large workspace." These are the identified requirements in the fields of optics, space, medical and microelectronics. The use of redundancy - or rather redundancies – will be implemented to achieve this.

 

Image removed.
© Symétrie

Redchain-Lab, la technologie blockchain pour un cloud solide, décentralisé et respectueux de la confidentialité

Joint laboratory. Partners: LIRIS / iExec Blockchain Tech

The RedChain-Lab joint laboratory project aims at developing protocols and operational blocks to set up a decentralized cloud infrastructure that offers scalability, resilience, trust and confidentiality properties. This infrastructure will be based on blockchain technology whose unique intrinsic properties of decentralization, robustness and transparency make it the most qualified technology to achieve functional objectives.
 

Partners' Websites : 

Rob4Fam, Robots For the Future of Aircraft Manufacturing

Joint laboratory. Partners: LAAS-CNRS / Airbus

One of Rob4Fam's (Robots For the Future of Aircraft Manufacturing) scientific challenges is to provide robots with the capacity to be reactive or, in other words, to ensure they can detect a change in their work environment and make a real-time decision to adapt to it. For example, a "reactive" robot needs to stop if it perceives a human operator intervening and then resume its task taking into account any modifications made by that operator.

Rob4Fam is also working on a project to develop technologies that can be adapted to different generations of robots.

 

Rob4Fam démonstration
© LAAS-CNRS

SIBIL-Lab, des solutions logicielles pour contrer les fraudes dans le B2B

Labcom ANR. Partenaires : LIRIS / Attestation Légale (OFA)

Le SIBIL-Lab s’attaque à la problématique de la gestion de l'identité numérique et à la détection des tentatives de fraudes. Le labcom développe des solutions dans le cadre du réseau social B2B créé par l’entreprise Attestation Légale qui permet de gérer des relations clients/fournisseurs interentreprises. L’objectif : soutenir la mise en place d’une plateforme capable de dimensionner, détecter et contrer les tentatives de fraude et ainsi assurer la sécurité et confidentialité de données sensibles des acteurs économiques.

Sites web des partenaires :

SIVALab, perception and localization systems for autonomous vehicles

Joint laboratory. Partners: Heudiasyc / Renault

 

Renault and the Heuristics and Diagnostics For Complex Systems Laboratory (Heudiasyc - CNRS/ University of Technology of Compiègne) have set up a new joint research laboratory called SIVALab (Integrated systems for Autonomous Vehicles Laboratory) which will be based in Compiègne.

This scientific and technological partnership is the fruit of a relationship of trust between Renault and Heudiasyc that has lasted over ten years. The SIVALab laboratory has shared governance and resources and provides a structure adapted to the sustainability of scientific developments and the implementation of large-scale projects. It is working on a four-year research programme.

This collaboration project's main focus is on the reliability, integrity and accuracy of perception and localization systems used for the navigation of autonomous communicating vehicles. The data analysed comes from the vehicle's sensors, pre-established navigation maps and from the exterior, via dynamic links to other road users and infrastructure. Thanks to an intelligent management of the combination of this data from multiple sources, the vehicle has a greater capacity to locate itself in its environment than it could alone with its sensors.

 

Image removed.
© UTC - C. Schryve

 

SurferLab, Distributed Intelligence for Transportation Systems Laboratory

Joint laboratory. Partners: LAMIH / Bombardier Transport / Prosyst

SurferLab is built around a core theme involving embedding artificial intelligence based processing capabilities in complex mobile systems (trains, cars, planes, trucks, AGVs, etc.) to implement and optimize various control, monitoring and diagnosis (health monitoring) functions.

SurferLab is a joint research laboratory formed of Bombardier Transport, Prosyst and the Laboratory of Industrial and Human Automation control, Mechanical engineering and Computer Science (LAMIH – CNRS/ Université Polytechnique Hauts-de-France. LAMIH is renowned for its work on transport and security, mobility and disability.

SurferLab's aim is to make these systems more intelligent, more autonomous, and capable of adapting to the unexpected. These systems are immersed in a fleet interacting with a control or maintenance centre and a human operator. This is a cyber-physical type (CPS) approach and in principle concerns all modes of transport.

 

Image removed.
© SurferLab

Telecom4Health (T4H), des technologies et solutions dédiées à la santé

Laboratoire commun (LCR). Partenaires : LIG / Orange

L’objectif du laboratoire commun T4H est de mener des travaux de recherche autour de l’intelligence artificielle et des nouvelles technologies appliquées à la médecine dite “4 P” : prédictive, préventive, personnalisée et participative. Les activités du labcom portent sur la gestion de consentement grâce à l’apport de la blockchain ; les sons santé afin d’atteindre des transmissions médicales de haute qualité ; le suivi à distance et l’IA raisonnée.

Sites internet des partenaires :