In search of risk

© Frédérique Plas / CNRS Images
© Frédérique Plas / CNRS Images
In connection with the France 2030 investment plan, the CNRS recently unveiled its new programme to identify and support bold scientific projects that could spark major technological advances.



Along with the CEA, Inrae, Inria, and Inserm, the CNRS is one of five research organisations to launch a Risky Research programme as part of France 2030 1 , as sought by the government since January 2024. The CNRS has inaugurated the Risky and High-Impact Research programme or (RI)², with the goal of identifying-very far in advance-the basic or innovative research that can lead in the coming decades to conceptual, technological, or strategic breakthroughs for France amid international competition. Twelve projects were selected at the CNRS as part of this programme (see text box). " The CNRS has always emphasized the importance of basic research, which is the foundation of risky research, as well as the source for the knowledge that has built our societies. This new programme confirms and reaffirms that there is no opposition between basic research and impactful research. Quite the contrary ," explains Antoine Petit, the Chairman and CEO of the CNRS.

"The CNRS identified our team during a discussion regarding a programme on fluids, and asked us to submit something more risky," relates Bérengère Dubrulle, a CNRS Senior Researcher at the Condensed Matter Physics Laboratory 3 who was recently selected for the (RI)² programme. For her project, this specialist on turbulence, who was Female Scientist of the Year in 2022, set out from a 20-year-old question involving spontaneous bifurcations 4 . Her goal is to predict unpredictable phenomena, especially those that are climate-related, extreme, rare, and on the generally very long timescales characteristic of fluid dynamics: "these types of issues are so risky and bold that I was never able to fund such research via existing means." The objective of the CNRS’s Risky and High-Impact Research Programme is to position itself in the vanguard by funding projects that would not have otherwise found takers on the national or European level.

Another example is Jérôme Casas and his project for a bioinspired artificial nose (see text box), which was refused in 2019 by the European instrument of the European Research Council’s (ERC). "There is a genuine scientific and technological risk. The idea of sexual pheromones seems far-fetched," explains the ecologist, whose project studies how animals, and insects in particular, detect very small concentrations of molecules in the air, a process that remains poorly understood.

An ideal format for scientists

The CNRS decided to allocate these means to projects that can break down barriers and revolutionize one or more fields, but whose funding represents a scientific wager, so to speak. "These are projects that we are unable to fund with the existing instruments. We help open new pathways with these projects," explains Frédéric Villiéras, the Director of the CNRS’s National Programmes Mission (MipN).

The particularity of the programme resides also in its simplification, as there is no call for proposals in the identification and selection of (RI)² projects. It is instead based on the experience of the organisation’s national scientific management, which travels across France throughout the year to discuss science, to accompany and support laboratory teams, and to identify future gems. "It is an ideal format for scientists in a number of respects, where science is the only consideration," says Dubrulle. The instrument is also simplified in terms of monitoring, for risky research entails agility. The mindset is very different from investments for the future programmes 5 , where the use of funds must be justified in detail. (RI)² functions experimentally based on a fixed-amount format that focuses exclusively on scientific monitoring. "This was a request made by various establishments, and it will be a relief to research teams," indicates Villiéras. It is a model of simplification that could potentially be applied to other research programmes...

The CNRS funds major projects, with a budget of approximately 2-3 million euros each. The programme is designed in two stages. During the first year, the demonstration stage, the idea or intuition is tested and assessed, with a "go/no go" halfway through at two years. If the demonstration is positive, this leads to the second stage, which allows for the project to be completed (in 5 years). "This means about ten selected projects for over 100,000 researchers, since the programme involves all CNRS laboratories, and is not just reserved to CNRS scientists," explains Alain Schuhl, the Chief Scientific Officer of the CNRS. The pre-selection decisions were deliberately made through the ten CNRS institutes in order to capitalise on the organisation’s foresight efforts and the work of specialized inter-organizational commissions, in addition to that of our university partners in our joint research units.

Twelve projects announcing revolutions

To improve the chances of breaking scientific or industrial barriers in emerging subjects, the twelve selected projects are led by "scientists who have already been singled out, and who have left a mark on the history of French research," Schuhl points out. These projects are led by researchers who have built internationally recognized experience and expertise in their field.

While the CNRS sought risk in these selections, it also wanted to promote projects with scientific, technological, or societal impact. And the projects selected indeed herald revolutions! For instance, the one coordinated by Bérengère Dubrulle could pave the way for new neuromorphological artificial intelligence (AI), which would function based on the differences or links between objects, thereby creating a new kind of AI that is more similar to neuronal function.

On the societal level, the transfer of knowledge in the climate and environmental sciences could lead to a system that predicts extreme events. "It tells itself that some marine birds detect storms before traditional forecasting systems such as Météo France, thanks to their ability to detect minute changes in their environment," Dubrulle cites by way of example. Exploiting this idea could revolutionize our approach to climate and environmental forecasting.

For Jean-Luc Moullet 6 , the Chief Innovation Officer of the CNRS, "the Risky and High-Impact Research Programme establishes a spectrum between research and innovation." While the work conducted in the (RI)² programme is not "applied research projects," specifies Schuhl, the CNRS implements technology transfer pathways toward industry for the research results if they are successful.

The ERC, an engine for risky research

This instrument is part of a new dynamic pursued by the government, inspired by the United States and the Defense Advanced Research Projects Agency (DARPA) 7 , probably the best-known American Advanced Research Projects Agency (ARPA). The project led by Thomas Ebbesen, who was also selected for the (RI)² programme, represents an emerging field that has already attracted the interest of the American strategic agency. "DARPA contacted me in 2012 when we published our first demonstration that the principles of quantum electrodynamics could be used to influence chemistry. I was impressed by their extraordinary monitoring capacity. Then in 2021, the agency launched a programme to encourage American researchers to explore this new field," underscores Ebbesen, who was awarded the CNRS Gold Medal in 2019, thereby stressing the importance of keeping his idea for modifying the properties of matter firmly within France.


    1 Investment plan launched in 2021 by French President Emmanuel Macron in response to the major challenges of our time.

    2 PEPRs involve priority fields selected and funded by the government, with a budget of 3 billion euros. The CNRS co-manages three quarters of the fifty programmes launched to date, notably all "exploratory" PEPRs, which support emerging transformations.

    4 A bifurcation occurs when a small change in a physical parameter produces a major change in the organisation of the system. During spontaneous bifurcation, the system’s organisation changes without an external parameter of change.


    5 The PIA is an initiative launched by the French government in 2010 to fund innovative and strategic projects for the country’s future.

    6 Jean-Luc Moullet was appointed the chief of staff for Patrick Hetzel, the Minister for Higher Education and Research, on 23 September 2024.

    7 An agency of the U.S. Department of Defense, tasked with research and the development of new technologies for military use.



DARPA was created by President Eisenhower in February 1958 in the middle of the arms race, in response to the Soviet launch of Sputnik 1. Famous for its bold projects and ability to transform innovative ideas into concrete reality, such as the Internet or the Global Positioning System (GPS), it is the perfect example of successful risky research. The very indicator that DARPA prefers to use is the percentage of its projects that do not succeed: 80%.

"That is risky research. Failure is part of the game. The idea is a willingness to invest millions of dollars, because what we are looking for is innovation that will transform society," explains Mireille Guyader, the Counsellor for Science and Technology at the Embassy of France in the United States. Jérôme Casas, the project leader for Innovating in Agroecology and the Fight against Terrorism, adds: "In addition to the finished product, DARPA is interested in mobilising and keeping afloat brilliant minds."

What if Europe could rival U.S. budgets? For Ebbesen, whose project received two ERC Advanced grants in 2008 and 2018 1 , the European Research Council (ERC) is the "engine behind European risky research." A pillar of the EU’s basic research programme for research and innovation, with a very low rate of acceptance, the ERC seeks to fund exploratory projects on the frontiers of knowledge. "I partially ’derisked’ the ideas behind my project for my ERCs. At the time, we did not know if our research would yield results." For the ERC indeed has just one selection criteria: scientific excellence. One essential point is that these grants are not oriented toward particular issues, and come with no short-term expectations for results.

Towards a European ARPA

"The CNRS is the primary beneficiary of European framework programmes, and more specifically ERC grants," details Jean-Stéphane Dhersin, the Director of the CNRS’s Brussels office. Since 2007, 14,000 "white" ERC projects have been funded, with 1,900 (13%) becoming ERC Proof of Concept Grants (which transfer the results growing out of ERC research). "This is clear proof that Europe has a system that can transform risky projects," Dhersin adds. However, it is perhaps necessary to go further, with the French President affirming, during his speech on Europe at the Sorbonne last April, the need "to continue up until this European DARPA, that we do not yet entirely possess." Will a response come in the form of the European Innovation Council (EIC), a key initiative of Horizon Europe with its budget of 10 billion euros for 2021-2027 seeking to develop groundbreaking technologies and innovations? Not in its current state. "Unlike ARPAs, which concentrate on proof of concept (TRL 3-4) 2 , the EIC devotes half of its budget to its Accelerator programme for projects with TRLs of 8 and above," Dhersin continues. According to a recent report, only 5% of Horizon Europe’s total budget supports groundbreaking innovation far removed from commercial applications. 3

The discussions surrounding FP10 (10th Framework Programme), the future framework programme for research and innovation-the successor of today’s Horizon Europe starting in 2028-will be decisive for the future of innovation in Europe. "The CNRS has affirmed the importance of projects with lower TRLs, and hence with greater risk, within the three pillars of the future programme," 4 emphasizes Dhersin.


    1 These grants allow nationally and internationally recognized scientists to conduct innovative, high-risk projects that open new pathways in their discipline or other fields.

    2 The TRL evaluates, on a scale of 1 to 9 (highest), a project’s technological readiness level. The Accelerator call supports start-ups and enterprises in the development of their key innovations.

    4 Scientific Excellence is devoted to basic research, with the ERC, Marie Sklodowska-Curie actions, and research infrastructure; Global Issues and European Industrial Competitiveness supports major collaborative projects and is organised into "clusters," with the largest share of funding at 53.5 billion euros; and the third pillar is the European Innovation Council (EIC).

The 12 projects selected for the CNRS’s (RI)² programme

Using a neuromorphological approach to detect and model weak signals announcing extreme events, with a view to improving climate forecasting and our understanding of complex systems.

Bérengère Dubrulle, Guillaume Balarac, Mickaël Bourgoin.

Exploring the impact of vibrational strong coupling on the chemical reactivity and properties of materials, providing a new way to transform the fundamental properties of materials.

Thomas Ebbesen, Cristiano Ciuti, Cyriaque Genet.

An innovative detector at the Modane Underground Laboratory tracks dark matter with increased sensitivity, extending the search over 12 orders of magnitude in terms of mass, and capable of detecting the antineutrinos of nuclear reactors.

Julien Billard, Stefanos Marnieros, Silvia Scorza.


Accelerating the transfer of mathematical ideas towards chemistry and physics, with a notable impact on quantum technology and molecular simulation.

Mathieu Lewin, Eric Cancès, Julien Toulouse.

Exploring transportation mechanisms and the olfactory cuticle surface to improve explosive detection and to optimise the use of pheromones in managing agricultural pests.

Jérôme Casas.

Developing an analytical method for cell proteins that preserves information regarding their micro-environment, with potential applications in medicine, as well as for research in virology, neurobiology, and beyond.

Raphaël Gaudin, Myriam Ferro, Jean-Christophe Olivo-Marin.

Developing an energy-recovering linear electron accelerator (ERL) in Orsay, with considerably reduced electricity consumption. This demonstrator provides France with European leadership.

Maud Baylac, Walid Kaabi.

Hierarchical and functional self-assembly for heterogeneous and dynamic catalysis in confined liquid interfaces Developing complex fluids (foams, emulsions) functionalised by enzymes to take advantage of the synergy between the physical properties of these fluids and the (bio)chemical activity of enzymes. This involves varying the characteristics of the complex fluid/enzyme duo in order to eliminate bacterial biofilms, or to degrade/upcycle microplastics.

Wiebke Drenckhan, Christophe Chassenieux.

Developing an advanced in situ system for the circulation of fluids in underground and ocean environments in order to characterize their dynamics and the effects linked to change.

Olivier Bour, Anthony Sladen.

Describing the diversity of languages in the Pacific, including languages of oral tradition that are vulnerable today, and studying their transmission in a multilingual context, all while situating them within the universal tendencies of language.

Jacques Vernaudon, Alejandrina Cristia, Alexandre François, Marie Salaun.

The project takes an interdisciplinary, integrated, and sustainable approach to the synthesis of the primary active ingredients in the pharmaceutical industry, with potential applications in oncology.

Claude Grison, Peter Hesemann.

Exploring DNA-based molecular computing, enabling requests for large masses of structured data. The goal is to provide proof of concept for a molecular computer (data coding, implementing data instructions, reading results), and measuring its environmental impact. The effort especially involves the development of algorithms that manipulate graphs of metadata 100 times greater than the state of the art.

Anthony Genot, Nicolas Schabanel.