Structuring research upstream with industry players
Inria is establishing close collaborations over the long term with large industrial groups. These partnerships combine research objectives and industrial interests. They allow Inria's researchers to face the major industrial issues of the moment and thus create long-lasting relations with a partner from the economic world.
The industrial partnerships are major advantages for the institute. A research program is developed jointly by Inria and its industrial partner based on the industrial player’s priority topics. Research projects, part of this program, are then built by Inria’s teams and the partner.
Alcatel-Lucent: The future communication networks
After a decade of collaboration, Inria created a joint virtual laboratory in 2008 with Alcatel-Lucent to design the communication networks of the future, capable of supporting the Internet of the future. They work together to decide on the topics, and the research efforts are split between them. Three research projects have been defined: optimized network management, self-organized wireless networks, and semantic networks. Fifteen original teams from five Inria centers Inria have been mobilized on these three projects. These successful relationships also allow the partners to respond jointly to competitive bidding for projects such as Univerself, the project of the 7th European PCRD, or the ANR Ecocell project.
Interview with Jean-Luc Beylat
What are the stakes in the collaboration between Inria and Alcatel Lucent Bell Labs France? Meet Jean-Luc Beylat, president of Alcatel Lucent Bell Labs France.
Can you tell us about the genesis of the collaboration between Alcatel-Lucent and Inria?
Jean-Luc Beylat: It was my predecessor who formalized the partnership in July 2008 (and for four years), but the meeting between Inria and Alcatel Lucent ultimately followed naturally. Because of the closeness of our activities, of course, but also because of the history of our past collaborations. In fact, the research teams and particularly Olivier Audoin, the joint laboratory's operations director, and Albert Benveniste, the scientific director, had already worked on optical propagation topics in the late 1990s. The need for a greater synergy emerged when Alcatel Lucent wanted to have a more long-term view of its prospects possibly provided by the Public Research work.
What are the various intersecting interests of Alcatel-Lucent and Inria?
Jean-Luc Beylat: The joint laboratory is an important base that feeds an entire ecosystem of other initiatives that we have together: our joint presence on the executive board of the System@tic cluster, our joint investment in ICT Labs, the Green Touch consortium initiated by Alcatel Lucent (which Inria joined immediately), the Mobile Services Initiative brought by Inria and supported by Alcatel Lucent, and even the laboratory in the making called LINCS, which will consist of leading researchers and major industrial partners to construct the Internet of tomorrow. All these things form an ecosystem that provides direction and fuels our partnership.
What do you feel is the distinctiveness that Inria contributes?
Jean-Luc Beylat: I see several. First of all, the focus of the clearly outlined research subjects; the scope of the project is discernible, which is essential in the face of our sector's economic and industrial issues. Then there is the excellence in the research of Inria and Bell Labs. In our partnerships, we have favored the quality of the teams over quantity. Lastly, the support of Inria's management: the drive and will shown by Michel Cosnard to move our projects forward, to lead to something concrete, as well as the motivation of the teams make the collaboration more effective; we have converging objectives and interests, which is exciting for both parties.
Can you share the results of the joint laboratory?
Jean-Luc Beylat: There is successful joint production in terms of research results, publications, and patent filings. The scientific performance is therefore very satisfactory. There is also the impact on Alcatel-Lucent's innovation, since this joint laboratory played an important role in the definition of new mechanisms for autonomy of the optical networks that we launched this year.
What are your expectations now? Your objectives?
Jean-Luc Beylat: With telecom networks growing faster and faster, we need to contribute more and more capacity and intelligence to them; in this case, this intelligence will focus on self-testing and network automation, which can grow thanks to the algorithms developed by Inria. Our long-term vision focuses on the growth of mobile networks. Lastly, in a more forward-looking manner, we will work on the security dimension, which will have increasingly significant issues.
What are the essential topics for the future?
Jean-Luc Beylat: With the Green Touch program, Alcatel-Lucent has taken a strong initiative by uniting a large number of major players in our sector in order to divide network consumption by one thousand. We are convinced that the technologies need to be virtuous, as their impact will only increase in the coming years. This objective is vital for Alcatel-Lucent. It is part of an approach to take into account our corporate social responsibility.
EDF R&D: Simulating energy production and consumption
Fifteen Inria project teams, totaling 25 to 30 researchers, are conducting research on supercomputing and high-performance energy simulation, which are priorities for EDF R&D. Three Inria spin-off start-ups (Distene, Graal Systems, and Caps entreprise) are associated with the project. They will be in a favorable position to offer enriched solutions at the end of the projects. This strategic partnership, entered into in early 2010, is the natural result of a long-standing collaboration. The projects focus on the optimization of electricity production and energy management systems, which are strategic topics for EDF.
Interview with Jean-Yves Berthou
What are the stakes in the collaboration between Inria and EDF R&D? Meet Jean-Yves Berthou, information technology program manager at EDF R&D.
What do modeling and simulation represent for the field of energy?
Jean-Yves Berthou: They are quite simply essential tools. We have used numerical simulation for many years. It allows us to provide long-term support for the performance of the 58 nuclear reactors and hydroelectric and thermal power plants that we operate. Simulation helps us to resolve operational problems but also allows us to optimize the balance between the production and consumption of energy.
What do you expect from this collaboration with Inria?
Jean-Yves Berthou: We have been working with Inria for many years. The institute is part of the major international organizations of reference in the field of applied mathematics and scientific computing. This strategic partnership allows us to reinforce simulation and supercomputing at EDF R&D. Today, we are collaborating on fifteen projects that mobilize approximately one hundred researchers from 17 Inria research teams, three start-ups, and six departments of EDF R&D. They revolve around the major computing codes in which EDF capitalizes its know-how. This involves, for example, developing innovative tools for the multi-resolution viewing of large volumes of data or programming models of hybrid multicore architectures.
To what do you attribute the success of this collaboration?
Jean-Yves Berthou: A shared strategic view on many topics. Our new industrial issues are also new major scientific challenges for Inria's researchers.
STMicroelectronics: Embedded systems of the future
In November 2008, Inria signed a strategic partnership agreement with the global manufacturer of electronic chips STMicroelectronics. Focusing on embedded software, two major projects have been launched since then, with the participation of some of Inria's project teams. As part of the first project, eight teams are participating in seven of the Nano 2012 projects, a program intended to improve the design process for embedded systems in systems on a chip (complete electronic system integrated into a chip). Three topics are explored in Nano 2012: optimization of program execution on an embedded multicore processor, acceleration of multimedia applications, and improvement of the design process for embedded systems. Six teams are also involved in Plateforme 2012, a project for experimentation on multicore architectures with more than 100 cores. Other joint research projects could be launched.
Bull and Inria have established a roadmap to define the priorities for the supercomputer manufacturer. The first framework agreement was just signed this summer. A dozen project teams are involved in working on the architecture, software environment, and services of supercomputers. These calculation beasts are used in many applications, from climate simulation to aerodynamic simulations as well as molecular modeling. The research focuses on parallel programming environments, middleware, computer security and reliability, as well as energy consumption optimization.
Andra: Modeling the storage of radioactive waste
Six Inria project teams are currently involved in the seven projects finalized as part of the strategic partnership with Andra (the French national agency for radioactive waste management). Focusing on radioactive waste management, this type of application is relatively new for Inria. The research focuses on the modeling and simulation of physical or chemical processes during storage, i.e. the transport of radioactive elements. They also involve the geometric representation of infrastructures and their environment (meshing of the site, certain geological zones, storage cells). In addition, the computing codes and resources used by Andra will be adapted in order to be applied to supercomputers or computing resources pooled to create a computing grid.
Interview with Patrick Landais
What are the stakes in the collaboration between Inria and Andra? Meet Patrick Landais, scientific director of Andra.
How did this partnership start?
Patrick Landais: We have been collaborating with Inria for more than ten years, especially with the Estime project team. Inria also participates in the MOMAS research group - of which we are a partner - on the modeling and simulation of radioactive waste management. It gradually became clear to us that it was appropriate to work together more closely. Our scientific needs and the demands of our collaborations justified a true partnership.
How did it materialize?
Patrick Landais: We submitted our topics of interest to Inria. Several teams responded by proposing projects. Together, we selected seven of them. Most of the research will begin in September 2010. It will help to refine the meshings for modeling, optimize the use of our increasingly large computing codes, and manage the constantly growing amount of data.
In concrete terms, what difference does this make?
Patrick Landais: It provides a long-term vision. Andra has always been a player of reference in numerical simulation in the field of radioactive waste management. This partnership will strengthen our skills. Our issues also contribute new research topics to Inria. So, the benefit is reciprocal. For example, Inria can participate in the European technological platform for the geological storage of radioactive waste (IGDTP), launched in November 2009.
Inria-Microsoft Research joint research center: Research for e-Sciences
Launched on January 11, 2007, the Inria-Microsoft Research joint research center is the continuation of a collaboration that started several years ago. Based in Orsay, it is organized around two research areas: formal methods and computing tools for the Sciences. The purpose of the research center is long-term research in the fields of formal methods, software security, and computing methods and tools for the sciences. The establishment of this joint research center demonstrates the international stature of French research in computing. It is proof of the determination that these two players have to serve a common ambition: to make fundamental advances that will benefit both Science and Society.