Inria - 2011 annual report


Published on

Digital technology now forms a central part of our everyday lives. It can be found in the majority of objects and services that we use every day, including the Internet, computers, mobile phones, cameras, portable music players, televisions, game consoles and GPS receivers, but it is also hidden away, sometimes in a less visible manner, in vehicles, industrial tools, medical devices, decision support systems, etc.

In the space of barely a few years, all these technologies have already profoundly changed our lives, to the point of becoming indispensable. They will continue to do so in the future, thanks to applications that are currently being developed. This is the route taken by researchers at Inria, a public research body, which is inventing the digital world of the future.

Published in: Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Inria - 2011 annual report

  1. 1. ANNUAL REPORT 2011 Education MedicineInformation Transport Software Support Environment
  2. 2. PAGE 57ResearchPAGE at the heart1 of the debate The increasing importance of digital technology in everyday life, whether in a visible and tangibleDigital technology or invisible and intangibleat the heart manner, raises many PAGEof our lives 17 questions.As the printing press,the telephone and railways P. 58 We are alldid in the past, digital stakeholderstechnology is massively in regulationchanging our homes, our Isabelle Falque-Pierrotin,businesses, our industries 2011, President of the Frenchand our schools by at the heart National Committee forbroadening the options of Inria Data Protection (CNIL).for communication, The research landscape is P. 61 Fertile disorderby automating certain constantly evolving. The Daniel Kaplan, CEO of thetasks in order to increase institute’s management Next Generation Internetproductivity and by and researchers talk about Foundation (FING).inventing new ways of the past year.exchanging information. P. 64 Let researchers P.18 Key figures research! P.19 Strategy Henri Verdier, President From the point of view of the Cap Digital of Michel Cosnard, competitiveness cluster. Claude Kirchner and Hervé Mathieu. P.23 Assignments Research, partnerships, technology transfer and the spread of digital sciences. P.49 Coordination The challenges of modernisation, the eight research centres. • Pages bearing the mark opposite are interactive! Photograph the image using the Pixee application to access the content on line. Pixee is a free application, which is available on iPhone, Android and BlackBerry. • How does it work? Mobile phones are equipped with a built-in camera for taking photos... but not just photos. Mathematical formulae developed recently within Inriaallow mobile phones to decipher the content of images like the human brain does. Thus, the phonecompares the image captured with several thousand images to identify the closest one and to provide uswith the information.
  3. 3. DIGITAL TECHNOLOGY AT THE HEART OF OUR LIVES… … Now and in the future Digital technology now forms a central part of our everyday lives. It can be found in the majority of objects and services that we use every day, including the Internet, computers, mobile phones, cameras, portable music players, televisions, game consoles and GPS receivers, but it is also hidden away, sometimes in a less visible manner, in vehicles, industrial tools, medical devices, decision support systems, etc. In the space of barely a few years, all these technologies have already profoundly changed our lives, to the point of becoming indispensable. They will continue to do so in the future, thanks to applications that are currently being developed. This is the route taken by researchers at Inria, a public research body, which is inventing the digital world of the future. 1
  4. 4. Inria How should we handle this overdose? It is well known that too much information can be as useful as no information at all. The fact that we have access to knowledge does not necessarily mean we are fully benefiting from it. With the development of digital technologies and the democratisation of communication tools, we are now swamped with information of all kinds, which is becoming increasingly difficult to analyse, rank and store. Huge numbers of e-mails, tweets and other messages Information circulate everyday on networks and even the most minor request via a search engine returns dozens of web pages, all of which represent sources of information that need to be patiently explored. What are we supposed to do, given that our brains cannot absorb everything? Numerous Inria teams are carrying out research aiming to organise and clarify this ever-increasing mass of data, in order to make it usable and understandable for everyone.This is precisely the case of the Orpailleur project-team, which is developing smart tools for managing and extracting knowledge from databases, or of Aviz , which is working on new ways of visualising and summarising information while browsing through large masses of data, by focussing on contextualising it. 2
  5. 5. Digital technology at the heart of our lives3
  6. 6. Inria The target is zero defects On 4th June 1996, Ariane 5 exploded in mid-air during its inaugural flight. The cause was a simple bug in one of the pieces of software managing Software the rocket’s autopilot. In fact, in computer sciences, the smallest design error can have disastrous human or economic consequences, above all when this involves a critical embedded program, like those within control aerospace, nuclear, medical or automotive systems. To prevent this kind of accident occurring, several Inria teams, specialising in algorithms and programming, have devoted themselves to the issue of software reliability by developing tools designed to track down faults in programs, from the design phase onwards.Thus, Abstraction has developed Astrée , a code analyser that is able to detect errors by performing automatic calculations. Gallium is seeking to improve the reliability of software thanks to Objective Caml , a high-level programming language, which is more reliable and expressive, and CompCert , a compiler that faithfully translates source code, without introducing any errors. Numerous ongoing projects, making use of powerful computing and mathematical concepts, have the same goal: to make software perfectly reliable, whether it is used for navigation, control, automation or communication. For the sake of all our safety! 4
  7. 7. Digital technology at the heart of our lives5
  8. 8. Inria The Magrit team Virtual technology for surgeons Preparing for an operation using a computer simulation, or even performing an operation by superimposing additional information within the surgeon’s field of vision; these practices, which appear to come straight from a science fiction novel, could actually become commonplace in the near future.This is due in Medicine particular to augmented reality, a technique that allows the surgical field to be enhanced by adding information such as 3D models of organs or vessels, therefore enabling specialists to see beyond what is actually visible, in order to better guide their actions. The Shacra team is working on this topic within MIX-Surg, one of three university hospital institutes (IHU) with which Inria has formed a partnership, of the six selected as part of the Future Investments programme launched in 2010. As for the Magrit project-team, it is working on faithful digital representations of the vascular network and on simulations of organs. By combining imaging, simulation, modelling and even robotics (automating and improving the reliability of certain technical operations), these approaches should significantly improve minimally invasive surgery and help to advance medicine and patient comfort. 6
  9. 9. Digital technology at the heart of our lives7
  10. 10. Inria The Wam team Progression towards smart cities More than half the world’s population now lives in an urban environment, and this trend is set to intensify over the course of this century. In order to adapt to this profound change and its social and environmental effects, cities need to revise their infrastructures and services by taking maximum advantage of the opportunities offered by digital systems and new technologies. Numerous smartphone applications are already showing the way by using Internet connections, for example, to allow you to find out, in real time, the location and arrival time of the bus you are waiting for, Urban or the address of a restaurant in planning an area you are unfamiliar with. Nevertheless, there is still a great deal to do to really make cities smart, in particular by providing residents with the means to sustain contextual information networks, which are of use to everyone. Inria is devising and preparing for this change.The Wam project-team has developed an augmented reality application allowing users to visit Grenoble with the benefit of additional visual and audio information on a smartphone. By using web geolocation data and wireless communication technologies, including wi-fi, 3G telephony and RFID tags on devices, a multitude of interactive services designed to assist disabled people, guide motorists and provide users with information, will soon be available in digital cities. 8
  11. 11. Digital technology at the heart of our lives9
  12. 12. Inria 10
  13. 13. Digital technology at the heart of our lives The Coprin team Digital home help By 2020, Europe will be home to almost 84 million people above the age of 75. With steadily increasing life expectancy and eroding family and social ties, increasing numbers of elderly people are finding themselves completely isolated, and not everyone is able to make useSupport of home help services to assist them with their day-to-day activities. This is a problem also encountered by disabled people, who are forced to perform all kinds of “tricks” in order to carry out everyday tasks. What if solutions could be provided by digital technology and robotics? As part of the large-scale PAL initiative (Personally Assisted Living) , Inria has devoted nine teams, spread across its research centres, to developing solutions to improve the lives of dependent people.The Coprin project-team has therefore developed Marionet-Assist, an automated system allowing people with reduced mobility to get out of a chair or to get up after a fall (the most common accident). E-motion is exploring artificial systems with sufficiently advanced perception, decision- making and action capabilities to interact with humans, and Maia is working on stand-alone machines with artificial intelligence, allowing us to imagine, one day, genuine domestic robots that are able to help us in our daily lives. 11
  14. 14. Inria The art of sharing electricity Everyone is now aware that energy is precious and that managing it is a major challenge for the 21st century, in both economic and ecological terms. Faced with the intrinsic limits of fossil resources, there is a need for new production and consumption patterns, favouring and exploiting renewable energies, including solar, hydraulic and wind power. Nowadays, private individuals and companies are Energy able to produce electricity and use it for their own needs or sell it on to a national operator, in order for it to be shared. However, in order for the system to work and develop, powerful digital tools are needed to manage this real-time distribution via smart grid type networks. Inria is working towards this goal by means of the Metis I-Lab , a joint structure bringing together the company Artelys and the Tao and Maxplus project-teams.Within this joint laboratory, researchers and industrial operators are pooling their resources and expertise to create a smart grid-based energy management system. Specialising in optimisation, statistics and decision support systems, Artelys uses the teams’ research into algorithms and modelling, in order to speed up the development of operational solutions to be offered to clients in France and around the world. 12
  15. 15. Digital technology at the heart of our lives13
  16. 16. Indiscretions on the Internet On the net, everyone believes they are safely hidden behind a pseudonym. However, a study by the MIT (Massachusetts Institute of Technology), involving Inria researchers, reveals that the majority of Internet users have the same user names on the various web sites that they visit. With the development of social networks and e-commerce, it is relatively easy to connect these virtual identities in order to target, track and even identify users forCommuni- commercial, political or criminalcation purposes. At a time when the Internet and digital communication are at the heart of society, the question of the security and confidentiality of personal data is crucial, in terms of both privacy and sound functioning of the economy. Around fifty Inria teams are currently working on this topic, which is a major challenge in this era of communication. Some of them, including Secsi and Caramel , are concentrating on protocols and encryption keys allowing the confidentiality of information exchanges to be guaranteed. Others, such as Planete , are examining security gaps in popular communication software and the anonymity of user names.The Secret project-team has been involved in the development of Shabal , an algorithm allowing the encryption standards for confidential data to be tested. These are so many avenues to explore, methods to invent and tools to be developed in order to make use of digital technology in a completely safe environment. 14
  17. 17. Digital technology at the heart of our lives15
  18. 18. Digital technology, an engine of growthDigital technology lies at the heart A study by Coe-Rexecode, publishedof our society, in our homes, of course, in May 2011, estimates that the digitalbut also in our schools, companies economy was on average responsibleand industries. As the printing press, (directly and indirectly) for 26% ofthe telephone and railways did in the growth in France between 1980 andpast, digital technology is massively 2008, 32% in Germany and Japan andchanging our economy by broadening 37% in the United States. This provesthe options for communication, its importance and shows the potentialby automating certain tasks in order for growth that is yet to be exploitedto increase productivity and by inventing in ways of exchanging information.Although digital technology still relies Inria is playing an active part in thison physical devices (processors, change with its research and technologycomputers, servers, networks, etc.), transfer projects, in particular by meansit also gives a proper value to intangible of I-Labs, joint structures combiningaspects, whether these are cultural teams of researchers and SMEs-SMIsassets (music, cinema, etc.), services within joint laboratories to workor intellectual creation. The same on industrial projects.applies to software tools. What’s more,digital technology is changing theeconomic landscape by significantlyreducing the cost of innovation, asillustrated by the story of e-commerceplatforms and social networks. 16
  19. 19. 2011,AT THE HEART of Inria 17
  20. 20. Inria — Annual Report 2011STRATEGYKey figures • 83 %7start-up created. 179 project-teams (on 1st January 2012), 83% of which are joint teams with universities and other research institutions.4,432reference publications. 4,351 people who have contributed to Inria’s assignments, of which106 68% are scientific personnel. • 68 %software applications submitted , 1282to the APP (France’s softwareprotection agency). PhD students within research teams.23 66initial patents filed. teams linked to foreign universities and laboratories. 304 theses produced. 18
  21. 21. 2011, at the heart of Inria Interview “INCREASING THE IMPACT OF OUR RESEARCH IS VITAL”Interview“INRIA ESTPRÊT ÀCosnard, Michel ABORDERLA DÉCENNIE Chairman and CEOQUI S’OUVRE” Michel Cosnard, Président-Directeur général 19
  22. 22. Inria — Annual Report 2011STRATEGYInterview On a European level, our global approach encompasses education, research, innovation and uses. In 2011, the institute established Engineering Schools (CDEFI) and our new sites in Bordeaux, Lille and Saclay, as institute – and the year in which numer- well as opening the PCRI in Saclay. How are ous future investments were initiated. you managing to go ahead with these The creation of Allistene provides proof developments against the background of a that digital sciences are a priority for our freeze on state funding? country, in the same way as health, en- Michel Cosnard • On the ergy, the environment and human and one hand, by means of an social sciences.This alliance has allowed increase in our own re- us to strengthen our partnerships and sources, largely associated to define the national priorities for dig- with the success of our ital science and technology in a con- teams in terms of calls for certed manner.The Future Investments projects, and on the other programme has also provided the op- hand, by means of improved efficiency in portunity to launch new partnership terms of their use, with the aim of achiev- activities, in particular in the fields of ing excellence in the field of management. health, the environment and human and Our first edition of the Inria awards also social sciences. Many of our research perfectly reflects these two aspects. teams are involved in laboratories of excellence (Labex) and excellence-in The French research landscape has equipment (Équipex) projects with local changed a great deal. How have these academic partners. As a national insti- changes affected the institute? tute, Inria has partnerships with three • M. C. In 2011 we saw the first full year university hospital institutes (Paris, of Allistene – a dedicated consultation Strasbourg and Bordeaux), and, as part structure for digital science and technol- of the digital economy programme and ogy, which brings together the French even technology research institutes Atomic Energy Commission (CEA), the (IRT), our partners are industrial opera- French National Centre for Scientific tors, major groups and SMEs. Research (CNRS), the Institut Mines- Télécom, the French Conference of European research has also been University Chancellors (CPU), the booming since the creation, in 2010, of EIT Conference of the Heads of French ICT Labs, for which Inria coordinates the research arm. Where do we stand now? Our aim is to achieve • M. C. EIT ICT Labs, one of the European excellence in the field Institute of Innovation andTechnology’s of management. (EIT) three knowledge and innovation communities devoted to digital sci- ences, is taking shape and looks very promising. It should allow a bridge to be built between research and industry by funding innovation, as far as actual 20
  23. 23. 2011, at the heart of Inriatechnological development – includingpatent applications, software, start-up Claude Kirchner,companies and other technology trans- Executive Officer for Research andfer methods. Its two strengths are: joint Technology Transfer for Innovationprogramming (the 23 partners proposeresearch topics and fund them with Eu- 2011 Inria Awardsropean support) and a new global ap-proach incorporating education, “The first emblematicresearch, innovation and uses, one of the embodimentvariants of which could, for example, betraining for new student profiles, both of our approach”scientists and entrepreneurs. “The purpose of the Inria Stéphane Donikian, the Do these developments influence Awards is to recognise creator of the spin-off and reward scientists who Golaem and its digitalyour research subjects? have made an outstanding human behaviour •M. C. Our research retains a certain con- contribution in the field simulators in 3Dtinuity and is in keeping with the burgeon- of computer sciences and environments.The Awarding development of digital technologies. mathematics, or who have for Research andIn this “world under construction”we are , made major contributions Innovation Support wasparticularly attentive to increasing the to technology transfer or won by Julien Wintz, innovation in these fields. a young engineer whoimpact of our research. We are increas- They also recognise has established and leadsingly interested in uses, an approach that engineers or technicians a scientific computingencompasses studying social networks, who have made a notable platform allowing– whether this is for privacy protection or contribution in terms of researchers to test theirthe right to oblivion – the Internet of ob- support or backing for software in their fields research. In 2011, the of application. Finally,jects, human-machine interaction and the Research Support Grand Prize was awardedhome help, as well as issues relating to to Gérard Huet, a pioneer Department Awardsustainable development. from the French school of recognised theThis is why we are in close contact with computer science, whose outstanding teamworkmajor economic and social players in work has been crucial in that contributed tothe digital world. We have also asked terms of guaranteeing the certification of the quality of software.The institute’s accounts,three of them to contribute on the fol- Young Researcher Award the first stage in thelowing pages. highlighted the talent of process of simplifying Bruno Lévy, a geometry and harmonising enthusiast who is practices within Inria.” radically changing 3D representation for both computer game graphics and computer simulation images.The Inria-Dassault Systèmes Award for Innovation recognised 21
  24. 24. Inria — Annual Report 2011STRATEGYRetrospective“Permanence in termsof strategy and renewedambitions”“Initial observation: the of research teams. All theseappropriateness of the aspects are summed upfundamental principles, perfectly by the motto foraround which the institute the institute’s first strategicwas built, as set down by plan, in 1994: scientificthe first chairman of Inria, excellence and technologyJacques-Louis Lions, in transfer.1979. More than forty yearslater, they are still relevant, Second observation: thewhether this lies in the ambitious plan to doublecombination of computer Inria’s activities and resourcesscience and mathematics, (staff and budgets) in tenthe importance of years, driven by Bernardtechnology transfer to Larrouturou (Chairman fromindustry, the structure 1996 to 2003), with the aimof research in small and of taking the institute to the highest global level, Hervé Mathieu,responsive project-teams, Executive Officer for Resourceseducation/training by formalised by the agreementwelcoming PhD students concluded with the and Service Administrationand young engineers or government in 2000, has until March 2012international openness. been an undeniable success.Over the course of the years Proof of this is provided byand successive chairmen, the international Visitingthis has been demonstrated Committee, which, having with universities, schoolsby the creation of start-up, evaluated the institute at and businesses. Yet, thereinvolvement in various the end of 2008, rated it as is still a great deal to beEuropean programmes and the European leader and done for computer sciencethe establishment of lasting one of the world’s best to be recognised not onlypartnerships, such as the research organisations in as a tool but also asErcim organisation, its field. Another telling a science in its own right. ”involvement in W3C since indicator is the large numberthe inception of the web, of grants that our researchersnumerous subcontracting obtain from the Europeanactivities, strong international Research Council (ERC).relationships and even While maintaining thisthe stringent assessment strategy, our aim is to further open up our research to industry and society, on the one hand, by incorporating social and environmental issues into our research, as we have done in the past for telecoms and life sciences, and on the other hand, by playing the role of a catalyst for research and a facilitator of projects, in order to support strong partnerships 22
  26. 26. Inria — Annual Report 2011ASSIGNMENTS… from the pointof view ofour researchers have already been envisaged, in particular in the environmental field: for example, detecting forest On completion of her engineering fires using temperature studies at the Insa in Lyon, sensors, monitoring air where she specialised in or water pollution and even networks and telecommunications, studying the behaviour of Nathalie Mitton joined groups of animals by fitting the Pops project-team. them with sensors. What As an Inria Research Scientist, I like about Inria is its she now leads the brand new openness: dialogue with Fun team, which is examining other regional research heterogeneous networks of centres, relationships with communicating objects. numerous researchers around the world and close “Having been attracted to contacts with industry. We science since my childhood, have concluded an external I studied as an engineer research agreement (CRE) at the Insa in Lyon. with FranceTélécom, and As I had chosen to created the EtiPops I-Lab specialise in networks with the local SME Etineo. and telecommunications, This joint structure, within I continued with a thesis which we are working on on wireless networks within geolocation, has a number the framework of CITI, an of advantages: it allows us Insa laboratory which was to pool our resources, define working in partnership with our research with a 3-year Inria.This research really road map and set ourselves fascinated me and I wanted industrial goals, with the to continue within Inria. creation of products, which In 2006, I joined the Pops is a way of giving our project-team, where research a highly visible I gradually took on form. ” responsibilities, beforeNathalie Mitton, setting up the Fun team, which will extend its workFun Team Leader, on future ubiquitousLille – Nord Europe Research Centre networks from 2012 onwards. We are actually aiming to develop and“Working with unify networks of RFID tags, sensors and roboticcompanies on actuators, which are battery-operated andindustrial projects” communicate wirelessly using radio waves. Numerous applications 24
  27. 27. 2011, at the heart of InriaRémi Gribonval, François-Xavier Le Dimet,Inria Senior Research Scientist, Professor Emeritus at Joseph-Fourier University,Rennes - Bretagne Atlantique Research Centre Grenoble – Rhône-Alpes Research Centre“My research combines “Inria has trusted memy two favourite to continue my research”subjects: musicand mathematics”“It was my love of mathematics that drew meto ENS, and my love of music that led me to choosea master’s degree combining the two. I continuedin the same vein by writing a thesis on soundprocessing with the IRCAM and the ÉcolePolytechnique. I subsequently left to take upa post-doctoral position at the University of SouthCarolina in the United States, where I studied thetheoretical aspects of the algorithms I was usingin greater depth. Since I was invited to join theMetiss team in 2001, I have been exploring thenotion of parsimony, which specifically allows audiosignals to be represented and processed with very François-Xavier Le Dimet is conductingfew parameters.The resulting techniques allow his research within the Moise project-team,audio data to be compressed, voices to be isolated which has taken over the Idopt project,and instruments to be separated within a recording, for which he was responsible. He also teachesfor instance. However, my research goes far beyond at Florida State University and has just beenjust sound, with applications for medical imaging, appointed a fellow of the Americanbiology, astronomy and, even more broadly, Meteorological Society.machine learning. Some of these have alreadybeen examined within the framework of Small “The life of a researcher majority of weather centresand Échange, two projects that I coordinated, one is made up of encounters. around the world, therebya European project and the other for the ANR I owe a great deal to helping to improve their(the French National Research Agency).These are Jacques-Louis Lions, the forecasts. At present, I amjust some of the many avenues to be explored Chairman of Inria at the continuing to develop itwith the team, which I will soon be leading. ” beginning of the 1980s. within the Moise team, I attended his classes for by specifically attempting my post-graduate diploma to make use of the dynamicA graduate of the École normale supérieure, in numerical analysis and information provided byRémi Gribonval is currently working on the concept he has always actively and satellite images. However,of parsimony in signal processing within the wholeheartedly supported the question of a link betweenMetiss project-team. In 2011, he was awarded the me in my research.This data gathered by observationBlaise Pascal prize by the Académie des Sciences(French Academy of Sciences) for his work. human dimension has been and mathematical modelsHe received an ERC (European Research Council) vitally important, as have representing complexgrant for his Please project. the discussions I have had phenomena clearly has with other scientists on my implications for areas other travels to the United States, than climate, such as Russia, China, etc. Jacques- oceanography, hydrology, Louis Lions’ optimal control biology and even petroleum theory also formed the basis research. ” for my concept of a data assimilation method, which was of interest to meteorologists. Initially, this approach was faced with the limited computational capacities of computers at the time, but it has since been adopted by the 25
  28. 28. Inria — Annual Report 2011ASSIGNMENTS… from the point of viewof our researchersMichel Banâtre,Inria Senior Research Scientist,Rennes - Bretagne Atlantique Research Centre“Carrying out originalresearch in directcontact with the worldof business“Michel Banâtre has conducted a real environment, or even,numerous projects with where possible and relevant,the LSP , Solidor and Aces for industrial use. Obviously,teams, focussing his research we shouldn’t hesitateon distributed systems, fault to immerse ourselves in ourtolerant architectures and chosen areas of applicationambient computing. Guided by and not ignore issuesthe constant goal of technologytransfer, he is behind the filing surrounding industrialof twenty or so patents, the protection rights, includingestablishment of numerous patents and prototyping.large-scale collaborations with This is what has guided myindustrial partners and the entire career as a researcher,creation of two start-up. and what still drives me. My first experience in this area logistics, with all its“I took my first steps in dates back to 1979-1985, potential applications inresearch as an engineer at with the Enchère system, sectors as varied as luxurythe University of Rennes 1, a perfect example of my goods and medical items.where I studied for my approach. My aim was My approach is not opposedentire degree in the field to apply the principle of to more academic research.of computer science, which distributed systems, which In my opinion, it should gohad just been born, at the I had been studying as part without saying when thebeginning of the 1970s, of my doctoral thesis, to concepts or tools proposedbefore becoming a senior auctions held by fruit and by researchers could formresearch scientist for Inria, vegetable producers in my part of industrial solutions,in 1986. I like research that native Brittany. Undertaken or even be the productsis off the beaten track, in close cooperation with of the future.”research that allows the local cooperatives, thissystems and services project formed the subjectof the future to be shaped. of a technology transferHowever, I am still convinced with an SME based inof the need for technology Rennes, but did not resulttransfer as a means to in it being marketed. Thisvalidate the results of was where I really beganresearch. For many to understand the difficultyresearchers, the promotion of turning a great idea intoof a scientific idea often a commercial solution, butends with a list of the I persevered with Bull, Texaspublications it may have Instruments and Alcatel-generated. For my part, Lucent in the field ofI want to see it tested in architectures and systems. Similarly, my research on ambient computing has formed the subject of a technology transfer to the JCDecaux Group. More recently, my work on paired physical objects has allowed me to tackle the issue of reliable traceability within 26
  29. 29. 2011, at the heart of InriaCatherine Bonnet, Martin Hachet,Disco Team Leader, Potioc Team Leader,Saclay – île-de-France Research Centre Bordeaux – Sud-Ouest Research Centre“Exchanging ideas with “Making interactive 3Dothers opens our minds“ technologies accessibleAs an Inria researcher andleader of the Disco team, which in complex environments, to everyone“ we are working on acutecomprises around fifteen myeloid leukaemia, amembers, Catherine Bonnet specific form of this working on the analysis We are working inand control of interconnected partnership with a doctorsystems in complexenvironments. Being extremely from the Saint-Antoinecommitted to scientific hospital and his team ofmediation, she is also a member biologists, who are helpingof the board of the ’Femmes et us to test and refine themathématiques’ association. models we are developing. Within the institute, I also“After completing my like being in close contactsecondary education, with all kinds of scientists,I studied mathematics, on trips, at conferences orbecause I liked it, without on the premises of Supélec,any idea of where it would where we are based, as wetake me. I continued with a know that exchanging ideasthesis on the modelling of opens our minds andhelicopter rotor blades for sustains us. I also take partAérospatiale. When I arrived in outreach activities forat Inria in 1994, I worked on mathematics and research, From the University of Bordeaux,an engine control issue for by explaining to young girls Martin Hachet joined the institute to takeRenault. This experience that these fields are not part in the Iparla project. As an Inriahelped me to understand incompatible with life as research scientist, he is currentlythe challenges of theoretical a woman, and that they leading the brand new Potioc team,research and exact are not forced to cut which is developing interactive 3Dmethods, which allow themselves off from the technologies for use by the general public.generic models and tools outside world, as is stillfor resolving specific too often the perception. ” “Initially attracted to the audiovisual sector,problems to be developed. I began studying computer science rather bySince then, I always think of chance. I rapidly recognised the creative potentialpotential applications even of this science: you start from scratch and shapewhen my research appears everything! During the course of my thesisvery abstract. Therefore, as on virtual reality, I produced a device allowingpart of Disco , which relates manual actions to be performed using 3D objectsto the analysis and control displayed on a large screen. However, theof interconnected systems solutions available at that time, mainly developed for the automotive and aeronautical industries, required complex and expensive devices. What is driving my current research is precisely the desire to make these tools accessible to as many people as possible, in order for everyone to be able to use them intuitively. I have been working along these lines within the Iparla project, where we have made use of the touch-screen interfaces of tablets and smartphones to redefine human-computer relations within mobility situations. I am hoping to go further with Potioc , by exploring new ways of manipulating 3D images while placing greater emphasis on the pleasure of use, in particular in the fields of art, culture, education and even personal assistance. ” 27
  30. 30. Inria — Annual Report 2011ASSIGNMENTSResearch New scientific and social challengesCONTRIBUTORS The institute has identified and is respondingPascal Guitton,Senior Research Scientist to new scientific and social challenges. FurtheringStéphane Ubeda, knowledge and developing technological expertiseDirector of Technological and its effects is crucial for the research communityDevelopment and industrial operators.Jean Roman,Deputy Scientific Directorwith responsibility for Applied Digital systems, which have permeated all aspects of our pro-Mathematics, Computing fessional and private lives, are at the heart of our modernand Simulation society. “We are working on numerous aspects associatedRoberto Di Cosmo, with this revolution, which is frequently only perceived byDirector of Irill users, recalls Pascal Guitton, “but our role is also to produce ” breakthrough research, which tackles new challenges that areJean-Pierre Banâtre, often hard to identify on the face of it.  We specifically exam- ”Director of EuropeanPartnerships ine these new subjects with Heads of Science and Deputy Scientific Directors, by setting up meetings between research- ers on shared topics, for example. Generating new research topics “In 2011, we organised three national seminars on cross- disciplinary topics, he adds. Researchers swap ideas and work together for several days. New connections and even, occa- sionally, new multidisciplinary projects emerge.  Thus, fifteen ” or so people met to discuss their work on brain-computer interfaces; devices that allow computers or machines to be controlled on the basis of measurements of the brain’s electri- cal activity. Although it is still at the early exploratory stage, this research will, one day, allow severely disabled people to access part of the digital world (games, virtual museums, etc.). In addition, Inria has independently organised days designed to present the results of various joint scientific actions.These talks have generally only been attended by a small number of researchers.To increase their visibility, scientific workshops covering all incentive actions were organised for the first time 28
  31. 31. 2011, at the heart of InriaAlgorithmic geometry:quadrics intersection,Vegas team.Christine Guillemot,Inria Senior Research Scientist, Rennes –Bretagne Atlantique Research Centre“Standardising is a wayof promoting our experience“For the majority of audiovisual alongside them. This is extremelyand multimedia applications, time-consuming, but, nevertheless,as for telecommunications, strategic.This ensures direct contactstandardisation is essential to with the technological challengesguarantee interoperability between of the moment and, in the event ofequipment developed by industrial a successful outcome, it guaranteesoperators. As soon as a new that our results have a markedtechnology emerges, like 3D impact within applications. The ”television did recently, a group standardisation process generallyof industrial operators takes the lasts a number of years, during 42initiative (in this case MPEG, which the algorithm is improved.Motion Picture Experts Group) This entails comprehensive develop standards. “As part “For these aspects of softwareof our research into compression development, we benefit fromand image transmission three years funding for an engineer standardisationalgorithms, standardisation via a technology development assignments wereis a natural and important step action. This allows us to finalise undertaken in 2011,in promoting our results, our algorithms and to improve in particular with bodieseven though publication our chances, even though success including the IET F, ISO,remains an essential means may not always be forthcoming. ” W3C, ETSI and OGF.of making our work and progressknown to the scientificcommunity. Nevertheless, ”given what is at stake economically,it is mainly industrial operatorswho submit solutions tostandardisation bodies suchas the ISO. Researchers generallywork upstream, transferringtheir technology to an industrialoperator who files patentsand protects the solution. “Weoccasionally need to contributeChristine Guillemot is leader of the Temics project-team, which hasbecome Sirocco . With four researchers, two engineers and 11 PhD studentsand post-doctoral researchers, she is working on 2D and 3D imageand video sequence processing including, specifically, on aspects relatingto the compression and communication of content. 29
  32. 32. Inria — Annual Report 2011ASSIGNMENTSResearchOn the left: Social robotics, in November 2011. “These joint research presentations, whichprototype robots in the form are often packed with information, have therefore been ben-of lamps designed to interact eficial to around a hundred researchers from all the researchwith humans in their daily centres, explains Pascal Guitton. “Forty or so subjects, ”environment, Flowers team. grouped into seven topics, have been discussed. These meet- ”On the right: An immersive ings will continue in 2012.and interactive video wall(CadWall) from Inria Sophia Supporting large-scale projectsAntipolis – Méditerranée. To encourage synergy between teams and promote multi- disciplinary research, the institute is supporting large-scale initiatives; these are four-year projects, which bring together several project-teams that complement each other to tackle a complex challenge.The aim is for this research to have a sig- nificant impact – including outside Inria – whether the result is finalised software, a tried and tested approach or a major7 theoretical breakthrough. In this way, the large-scale PAL (Personally Assisted Living) initiative is enabling nine project- teams (Grenoble, Sophia Antipolis, Nancy and Rennes) to jointly design and test personal home assistance technologiesactive large-scaleinitiatives in 2011. (for elderly, disabled or injured individuals). “What makes the system original is the ability to combine the acquisition of data, indicating a fall, for example, with a re- sponse by autonomous robotic systems, explains Stéphane ” Ubeda, “all at an affordable cost. We have fitted two apartments19 (in Nancy and Grenoble) with non-intrusive sensors and con- ducted initial automated experiments, as well as protocols, with around thirty patients. Academics and industrial opera- ”active collaborative tors have also been able to share these experiences.To increaseresearch actions (ARC) the impact of this research, the institute has decided to broad-in 2011. ly strengthen its support for large-scale initiatives in 2012. Promoting technological development Software development forms part of the daily lives of every Inria team, within the framework of European or French National Research Agency (ANR) projects or, quite simply, 30
  33. 33. 2011, at the heart of Inria Anne Canteaut, Inria Senior Research Scientist, Paris – Rocquencourt Research Centre “We are evaluating the security of cryptographic algorithms“ It’s a strange game that cryptography researchers play: on the one hand, they invent algorithms for authenticating confidential files, and on the other, they launch attacks on competing algorithms to expose their flaws. This is precisely how, in 2004, the dozen “hash” standards – key functions that create a sort of file fingerprint – were found to havewhen illustrating or explaining a research task. Inria has flaws. The American NIST (Nationalcreated a dedicated organisation to ensure that all the teams Institute of Standards andTechnology)are able to develop high quality code or put hardware launched a competition to define a new and more robust standardplatforms in place (such as the wheelchairs or hinged arms by the end of 2012. “Having reacheddeveloped within PAL). the semi-finals with our algorithmEngineers from the Experimentation and Development known as Shabal, we have beenDepartments (SED) within each Inria research centre – all evaluating the security of the fivescientists and the majority of them with a PhD – and the fifty finalists that are still in the running,or so young contracted engineers who assist them every since 2009. Seeking and examining new attacks also gives us new ideasyear provide the researchers with support. for functions. For example, knowingWithin the framework of technology development actions that the high level of polynomials(ADT), the engineers join research teams for at least two for hashing is one of the relevantyears. These ADTs, launched via internal calls for projects, security criteria, we developedare expanding with between 30 and 50 new ones every year. a brand new mathematical tool for evaluating it correctly. This resulted in the Japanese Luffa Increasing fruitful collaborations algorithm being eliminated fromThe Mobsim ADT, for example, is bringing together three the finalists. It has become the ”engineers as part of the development of the NS3 radio net- basis for more general theoreticalwork simulator which, promoted by Inria, is in the process work. “In order to successfullyof becoming the worldwide standard. quantify the security of hashing functions, we have developedWithin the framework of the excellence-in-equipment (Équipex) new techniques, which preciselyFIT (Future Internet ofThings) project, which was launched in evaluate the quality of the algorithm’s2011, web-based communication by networks of several hun- construction and the impactdred physical measurement sensors installed on five sites of imperfections. ”(Grenoble, Lyon, Rennes, Lille and Paris) is tested, thanks totechnologies developed by around a dozen engineers. Thisequipment is made available to all French researchers. Anne Canteaut has been the SecretFinally, as part of the Sofa research platform, which brought project- team leader since 2008. Within her team, with three permanenttogether five project-teams to work on a realistic simulation researchers and around ten PhD studentsof organs, in 2011, four engineers completed the development and post-doctoral researchers, she takesof a real-time medical simulation software suite (Mix-Surg), an interest in the protection of information,which will be used in the future university hospital institute in particular in the security of cryptographic algorithms. In 2012, she will chair the(IHU) in Strasbourg. programme committee for the domain’sThis therefore means that development is able to break international conference: Fast Softwarethrough barriers: “Engineers from Inria and the Fraunhofer Encryption (FSE) in Washington. 31
  34. 34. Inria — Annual Report 2011ASSIGNMENTSResearchOn the left: ANG, Institute in Darmstadt (Germany) will take three to four yearsan experimental smart to create a fully interoperable generic environment forwalking frame, Coprin team. controlling virtual or augmented reality systems, which are applicable to all domains, with the focus on remote learningOn the right: wifibot all-terrainguided robots, Pops team. for surgical operations and remote industrial maintenance, ” explains Stéphane Ubeda. Creating experimental platforms Dedicated to scientific experimentation for the evaluation of algorithms, for example, experimental platforms, generally developed with territorial partners, are managed by research- ers and, mainly, by development engineers. They are open to scientists and companies in a wide variety of research and400 application fields. “Num3sis in Sophia Antipolis is a modular architecture devoted to scientific computation and digital simulation“ , explains Jean Roman. “It allows software to be rapidly inte-development engineers. grated, complex simulations to be run and results to be viewed. Thanks to the work undertaken by the SED, it can potentially be adapted for all applications, including medical imaging and fluid dynamics. ” Jean Roman also refers to the PlaFRIM (Bordeaux), a fed- erative high-performance computing platform for research in computer science and mathematics. Operational since May 2010, it allows algorithms and scientific computation codes to be tested before being deployed by large national computing centres. 32
  35. 35. 2011, at the heart of Inria ©GeoEye ©ASI On the left: Data visualisation using Graphdice software, Aviz team. On the right: Automatic satellite image classification (©GeoEye and ASI), Ayin/Ariana team.The European Research Councilrecognises the visionary aspect of Inria x2For the fourth year in succession, new call for projects. “Seventhis European funding programme researchers from our teams won(the European Research Council, ERC awards in 2011: Xavier Rival,more widely known by its English Rémi Gribonval, Erwan Faou andacronym the ERC) is rewarding Andreas Enge in the young Inria’s workforce hasresearchers with particularly researchers’ category, and doubled in ten years.ambitious and visionary ideas. The Marie-Paule Cani (INP Grenoble),first ever prize winners from 2007 , Nicholas Ayache and Dale Millerwill submit their final results in in the experienced researchers2012. “Inria has had winners every category. What is needed for ” 10time, which is rare, both for young this funding is to be individual.researchers (for which funding can Selection takes account of thebe up to 1.5 million euros) and for candidate’s potential and theexperienced researchers (up to excellence of his/her career as well2.5 million euros). The 23 winners as the quality of his/her project, to 15 Inria candidateswithin Inria project-teams – 18 for the originality of his/her approach apply for each ERC awardscomputer science research and and the anticipated impact. The programme.5 for applied mathematics – benefit ERC has already awarded morefrom exceptional conditions, says ” than 2,200 grants across allJean-Pierre Banâtre, Director of disciplines. In the field of computerEuropean Partnerships. In concrete science, Inria is the Europeanterms, thanks to these unusual establishment which is home tofinancial resources, they create or the largest number of winners.strengthen their team and devote The ERC’s ambitious and globally75 to 80 % of their time to their unique funding programmeproject, with a great deal of (7 billion euro over five years) willfreedom. This is every researcher’s be repeated and even expandeddream, which is why there is an for the period between 2013 andincreasing interest in the ERC, 2020. This funding has noweven though the selection process become coveted and candidatesis rigorous (the success rate is from all around the world apply10 to 15 %). There are now between to conduct research in Europe.10 and 15 Inria candidates for each 33
  36. 36. Inria — Annual Report 2011ASSIGNMENTSResearch On the left: Scan of white human brain matter (coronal image), Asclepios team. On the right: Molecular manipulation via a multi- touch screen, Mint team. Irill or free software as a research topic6%market share for freesoftware and related The Initiative for Research and Innovation in Free Software (Irill) was launched in October 2010 by Inria, Pierre-et-Marie-Curie University and Paris-Diderot University. The research centre welcomes researchersservices in France(2.5 billion euros on a who are interested in new problems posed by thenational market forsoftware and services development of free software, who are prepared toworth 40 billion euro). create tools to resolve them and to promote them with developers. This is a particularly innovative scheme in terms of both shape and substance.x2The quantity of freesoftware has almost “Nowadays, almost all software users make use of free soft- ware, claims Roberto Di Cosmo, Director of Irill. This means ” that this movement, which began in the 1980s, has expanded. “We cannot support the idea that all software should be free, but this will become the rule for technical reasons: there needsdoubled every two yearswithin Debian distribution to be access to all the source code in order to create and de-(used by Ubuntu). velop increasingly complex computer systems, which are subject to increasingly stringent quality requirements. ” Therefore, the monitoring, management and communication tools used by developers of free software need to evolve: the size of these “technical objects” is growing as the amount of software rises and the communities developing them expand. Irill’s primary mission is to become a globally unique centre of expertise, attracting researchers who are able to identify the issues of these growing code bases and to finalise solu- tions and tools permitting this collaborative development. “These kinds of candidates are still fairly rare. In the world of computer science, researchers who publish articles and source code developers rarely speak to each other. Our aim is to con- nect these communities with each other and with industrial operators, in particular the SMEs who use free software. ” 34
  37. 37. 2011, at the heart of Inria Julia Lawall, Inria Senior Research Scientist, Paris – Rocquencourt Research Centre “It provides the opportunity to interact closely with this community“ Julia Lawall is a quiet campaigner for free software. Nevertheless, her research has already helped radically change practices within the Linux community, with its Free software disrupts the lines… of code free operating system developedThe efforts of Irill, which is home to around twenty people, in 1991 now being one of the mostare starting to bear fruit. Fifteen or so meetings were organised reliable and widely used. Shein 2011 to bring dozens of developers from different commu- became interested in free softwarenities together (GNU, Debian, LibreOffice suite) for several somewhat by chance, in 2005,days and to foster contacts. when studying the problems of switching from one version of Linux“We don’t offer any permanent posts, but welcome research- to another, with Gilles Muller, aters working for our partners. Our goal is to support projects the time at the École des Minesthat bring together researchers who share our view, to make in Nantes (now at the Inria Paris –them more visible and allow them to benefit from our dy- Rocquencourt Research Centre).namic approach for a few years.They will subsequently make They identified a key issue for which they developed a languageroom for other researchers and projects and remain in contact, that was able to identifybuilding a large community. At present, Irill is hosting three ” conversions from millions of linesflagship research projects, all designed to improve the qual- of Linux code, to specify them andity of code, programmed in OCaml language developed by even to search for errors, a uniqueInria: the Coccinelle project, involving the development of generic tool. This was a fantastic giftLinux kernel code, headed by Julia Lawall, the Mancoosi for Linux, incorporated into the kernel of the system in 2010project examining the quality of free software distribution, and now used by the majoritywith, specifically, Stefano Zacchiroli, the leader of the Debian of developers. “Working in the fieldproject and the Ocsigen project, investigating a new way of of free software, which is outstandingprogramming rich web applications. in terms of quality and functions,Irill’s second mission is to help modernise education in order provides a fantastic opportunity toto prepare computer science students for these new forms of interact directly with the developers and to discover new issues forcollaborative working, from both a technical and human point research. It also guarantees thatof view.The working environment can actually be confusing, our work will have a direct impact. ”with discussions between contributors occasionally being Although Julia Lawall is a specialanimated. “We began by getting computer science research- case within the Linux community,lecturers together to swap experiences and best practice. having joined Irill in order to continue with her project, she isUltimately, this could give rise to a shared model for courses, now surrounded by scientistsor even masters degrees specialising in free software. ” engaged in similar research.Its third mission is technology transfer and development.“This is what will take the longest, acknowledges Roberto Di ”Cosmo. For forty years, the software business has been based Following a degree in mathematics andon the sale of licences, a model that is now reaching its limits, computer sciences in Ohio (United States),including for music and video.There is now a need to promote a PhD thesis at the University of Indiana on programming languages, followed bythe adaptation of software, its qualification and certification, post-doctoral research into the optimisationtraining and other related services – all emerging fields that of programs at Inria’s Rennes – Bretagnerequire new skills. Atlantique Research Centre, Julia Lawall has been a lecturer at the University of Copenhagen (Denmark) for eleven years working on the robustness of operating systems. She is now head of the Coccinelle project at Irill. 35
  38. 38. Inria — Annual Report 2011ASSIGNMENTSPartnerships Increasingly structured partnershipsCONTRIBUTORS Research structuring plans within computationalAntoine Petit,Deputy Managing Director sciences are continuing to bear fruit. At the same time,Hélène Kirchner, the numerous calls for proposals for the FutureDirector of International Investments Programme, which punctuated the yearRelations 2011, have allowed for the consolidation ofOlivier Festor,Senior Research Scientist partnership and technology transfer policies, whichat EIT ICT Labs are a priority for the institute.Thierry Priol,former Senior Research “Partnerships, whether in France or around the world, are es-Scientist at EIT ICT Labs sential if we are to be at the forefront in our fields of research and known and recognised on the international stage, says Antoine ” Petit, Deputy Managing Director of Inria.This view guides the institute’s partnership policy on all levels.That is why, this year, Inria signed a framework cooperation agreement with the French National Centre for Scientific Research (CNRS). “This agreement underlines the complementarity of our actions and our desire to work together” explains Antoine Petit. “It also documents our joint contribution to the establishment of site policies in the field of digital sciences, in close cooperation with universities and schools. Thus, holding an annual site committee meeting will ” allow the two organisations, universities and schools to discuss their priorities and establish a genuine coordinated policy. Investments that consolidate Inria’s policy Within the framework of calls for proposals for the Future Investments Programme (PIA), Inria has supported universities and schools by becoming involved in 17 selected laboratories of excellence (Labex) and three excellence initiatives (Idex). “This programme has also provided the opportunity for the institute to assert its presence within life and environmental sciences, ” explains Antoine Petit. For example, the institute is involved in three of the six selected university hospital institutes, in a number of bioinformatics projects and two institute of excellence projects 36
  39. 39. 2011, at the heart of InriaOptimisation of semantic webdata storage, Leo team.Marc Thiriet,French National Centre for Scientific Research (CNRS)researcher, Paris – Rocquencourt Research Centre“Our action has two aspects:research and dissemination““Tony Wen-Hann Sheu, from earned us a Franco-TaiwaneseNational Taiwan University, spent award, each receiving 19,000nine months at the Jacques-Louis- euros. This money allowed me toLions laboratory in Paris in 2001. fund the summer school (Cermacs)We had a mutual interest in the at Luminy, and Tony Wen-Hannbiomechanics of respiratory and Sheu organised a similar seminarcirculatory systems and began the following year in Taiwan. Wea collaboration, which has since are currently looking to establishextended to the Taida Institute a twinning programme betweenfor Mathematical Sciences and our two laboratories in order,the Taiwanese high-performance in particular, to explore acupuncture 83%computing centre. We also devoted remedies. We will be submittinga great deal of energy to the an application as part of thedissemination of knowledge, cooperation agreement concludedby establishing student exchanges between the National Sciencefor summer schools and by Council of Taiwan (NSC) and Inria, of project-teams are jointregularly holding conferences. which will enable associate teams teams with universities,Our activities were supported both to be formed between the two grandes écoles and otherby France (Inria and Égide, the countries. ” research organisations.French Institute in Taiwan) andTaiwan (the National ScienceCouncil and representativesin France). For 2007-2008, wespecifically benefited from a“Hubert-Curien partnership” (PHC)for the development of scientificexchanges, funded by the FrenchMinistry of Foreign and EuropeanAffairs with the support of theFrench Ministry of Higher Educationand Research. This dual aspectof research and disseminationMarcThiriet is a member of the Reo project-team. In 2011, together with his colleagueTonyWen-Hann Sheu, from the University of Taipei, he received the Franco-Taiwanese Scientific Foundation Award, created by the National Science Council ofTaiwan (NSC) and the French Academy of Sciences.This prize was awarded for theirwork on physiological fluids, which they began together around ten years previously. 37