Sites Inria

Version française

Inria Joint Labs

Promoting upstream research with industry

Laboratoires Communs Chercheurs © INRIA / Photo C. Dupont

By engaging in prestigious partnerships with the most important digital players, Inria cultivates individual talent and partnership research. The institute supports the transfer of results and skills from its researchers to companies, from start-ups to major groups. In order to promote upstream research with industry, the institute creates framework agreements or joint laboratories with long-standing partners. The main purpose of the latter is to lift the technological locks on major and targeted subjects.

Joint Laboratory with Qwant

The first contacts between Qwant and Inria began within the scope of the work conducted by the Wimmics team (Inria), on the themes of the Data Web and education.

The first objective for Qwant is to continuously improve the algorithms and technologies relating to its core business: to collect, index, search, sort, interact, secure, distribute and protect data. It is also about developing new functionalities such as the detection of emotions or the analysis of videos, or new products like Qwant Junior, each time keeping in mind the notion of 'privacy' and the ability to give control of their data back to the users.

For Inria, the objective is to develop a technical collaboration with a fast-growing European Web player, by working on and with the algorithms, models and large volumes of data from a real search engine with millions of users and very heterogeneous content. In particular, this collaboration aims to work on the extremely sensitive subject of citizens' privacy.

In order to develop their collaboration, Qwant and Inria wanted to set up a joint research laboratory on the theme: "Smart Search and Privacy". This partnership consists of four research areas:

  • Crawling, indexing, execution platform searching, privacy by design, security
  • Mapping, geography, maps, globe, GPS, navigation, itineraries, traffic
  • Augmented interactions, connected objects, online assistants, personal assistants
  • Education (EdTech) and users of Qwant Junior

All of the Inria project teams relevant to the above four themes are concerned (around 20 research teams from the eight Inria centres have already been identified).

This joint laboratory will be led by Fabien Gandon (Inria, Wimmics, Inria-UNS/UCA-CNRS joint project team and Inria's W3C representative), with Sylvain Peyronnet as deputy director (Qwant, scientific director).

Joint Laboratory with Nokia Bell Labs

Launched in 2008 and renewed in 2017, the joint lab associates permanent researchers from the two partners with PhD students and postdoctoral researchers, sharing the same strategic vision to solve the key scientific challenges linked to the evolution of networks and network applications. The aim of this joint research is to offer users the benefits from required network and cloud resources for a contextual and personalized experience of the digital connected world.

The future networks will have to manage a multitude of connected objects, to host and interconnect massively distributed functions, to feature an unprecedented agility to support differentiated and demanding use cases like the connected car, industry 4.0, smart city and e-health. This will require strong guarantees in terms of security and privacy, while hiding the complexity through a high level of automation. To achieve this aim, this new phase of the common lab will associate advanced research in information theory, machine learning, graph theory, game theory, cybersecurity, network virtualization and advanced control software.

The scope of the collaboration covers several research actions dedicated to:

  • Information theory and algorithms to solve the challenges of IoT
  • Analytics and machine learning to dynamically and automatically optimize the virtualized network
  • Scalable distributed learning and control for augmented intelligence to operate complex IoT networks of dynamic elements
  • Cybersecurity to provide privacy, data integrity and resilience against intrusions

Framework agreement with EDF

This strategic partnership concluded at the beginning of 2010 and renewed in October 2016 for a further five years, is the natural continuation of a long-standing collaboration. The projects concern the optimisation of electricity generation and energy management systems, which are strategic subjects for EDF.

One could mention the following economic and scientific issues identified by the partner: the supply of clean energy at lower cost, the use of all renewable energy sources and optimisation of the complex systems which lie at the heart of electricity production.

These issues led to the identification of a certain number of avenues for research : Stochastic modelling and uncertainties, interoperable schemes, solvers and software platforms for intensive digital simulation, Visualisation and collaborative work, Big Data, Smart Grids.

In recent years, the joint research carried out has enabled a certain number of results to be achieved : prioritisation of sources of uncertainty in reliability calculations – application to fast fracture risk assessment for reactor pressure vessels, estimation of marginal costs in daily electricity generation, functional approach for time-series forecasting, as well as data mining preserving confidentiality in a P2P environment.