On March 29, 2018, during the “AI for Humanity” day, Emmanuel Macron announced the “National Strategy for Artificial Intelligence”, inspired by Cedric Villani’s report which called for “the awakening of France and Europe” in terms of AI. For France to have a role as a world leader in AI, this report recommended the creation of a network of Interdisciplinary Institutes of Artificial Intelligence. Four three 3IAs were finally selected and financed viaa 1.5 billion euro plan.

The creation in France of a network of Interdisciplinary Institutes of Artificial Intelligence had three interdependent objectives:

• “The (re)constitution of attractive and prestigious research environments, capable of significant advances on a global scale, grouped under a single, visible and recognized label;
• The dissemination of high-level scientific training in AI, for the researchers, engineers and entrepreneurs of tomorrow;
• The fluidity of interfaces, between disciplines and between academic research and the industrial world, promoting the rapid transformation of ideas into proofs of concept (POC), scientific applications, technological advances and intellectual property, able to create the fabric of startups and SMEs on which tomorrow’s industry depends.

Following an AMI launched by the French National Research Agency (ANR) in July 2018, the jury selected four 3IA institute projects from the sites of Grenoble, Nice, Paris and Toulouse and requested their labeling.

These four AI research clusters each have their specificities:

• MIAI Grenoble Alpes: AI for humans and the environment
• 3IA Côte d’Azur: AI for health and regional development
• PrAIrie: AI for health, transport and the environment
• ANITI: AI for mobility, transport and the industry of the future

Specificities that do not prevent them from operating as a network and creating excellent conditions for collaboration between public and private, academic research and innovation players of all sizes.

Assessment of the 3IAs’ actions

Since their creation, the four 3IAs have had as a project a collaboration between academic research and companies which makes them play an economic role in their ecosystems
On the other hand, the 3IAs have recruited foreign researchers with a brilliant track record, developed international collaborations and have each welcomed young researchers.

On the training side, the objective is to double the number of people trained in AI within 4 years through training at all levels and for all, including continuing education:

• Introduction of AI modules in existing training courses,
• Increase in the number of students,
• Opening of new training courses on the levels bac+3, bac+5 and bac+8,
• Implementation of transversal training modules in AI at the interface with other disciplines,
• Development of continuing education in order to meet the short-term challenges expressed by professional
  Development of continuing education in order to meet the short-term challenges expressed by the professional sectors, etc.

The mapping of the offer has thus notably evolved in two years, significantly increasing the readability and attractiveness of AI training on the 3IA sites. The concern for the impact on the value chain is omnipresent in the activities of the 3IA. Ethics, the impact of AI on society, the diversity of the teams, and the dissemination of the scientific culture of AI are all essential topics within 3IA.

Making France a leading country in AI

The 3IAs organize themselves into a network and carry out joint actions to make France a globally recognized leader in AI. A few examples illustrate this ambition:

• Among the very first joint events, ROB@3IA, a workshop on robotics
  organized by ANITI (LAAS-CNRS) and PRAIRIE took place in Toulouse on October 24 and 25, 2019.
  October 2019.
• AI4HEALTH Winter School, organized from January 4 to 8, 2021 in collaboration between the Health Data Hub’, the 3IA MIAI Grenoble Alpes, 3IA Côte d’Azur and PrAIrie, as well as
  AIM, the University of Paris and the Institut du Cerveau de Paris – ICM
• In collaboration with PrAIrie, MIAI Grenoble Alpes organizes every year since
  2019 the PAISS (PrAIrie/MIAI Artificial Intelligence Summer School). This
  school welcomes each year more than two hundred students from all over the world (300
  (300 participants of 52 nationalities in 2021), selected from several hundred applicants.

Some figures to remember about the 3IA

• A total budget of 251.5 million euros,
• 149 chairs,
• 563 academic researchers,
• 454 doctoral and post-doctoral students funded,
• 168 partner companies,
• 186 researchers or engineers from companies,
• 13678 people trained in AI each year,
• 1471 publications.

The 3IAs – actions adapted to the territories

3IA Côte d’Azur

Supported by the Université Côte d’Azur, Inria, CNRS, Inserm, Eurecom and SKEMA Business School, the 3IA Côte d’Azur scientific program is developing applications for health (medicine and biology) and intelligent territories, as well as specific tools to boost collaboration with companies and technology transfer, notably

• the “3IA Techpool”, a pool of shared AI development engineers set up in early 2021 to meet the demands of
  The “3IA Techpool”, a pool of shared AI development engineers to be set up in early 2021 to meet the demands of collaborative engineering for the evaluation of tools and the development of proofs of concept and prototypes;
• the “Start-it up” scheme to support the economic and technological maturation of AI start-ups;
• Training: + 33% of students trained in AI in two years: master AI in co-diploma with the University of Laval, creation of an inter-EUR minor in AI, development of a continuing education offer adapted to the needs of companies, creation of the DU AI & Health, the 3IA Côte d’azur mobilizes all the tools to adapt its training offer to the needs of today and tomorrow

The 3IA Côte d’Azur collaborates and dialogues with all the new AI and digital objects created on its territory concomitantly with its labeling: the Maison de
The 3IA Côte d’Azur collaborates and dialogues with all the new AI and digital objects created in the region at the same time as it was awarded the label: the Maison de l’intelligence artificielle (MIA), which disseminates scientific culture in AI to schoolchildren and the general public; the Observatoire sur les impacts économiques et sociétaux de l’intelligence artificielle (OTESIA), which studies the impact of AI in different areas of society; the EuropIA Institute, which is building events for the general public around AI on the Côte d’Azur.

Olena Kushakovska, SAP, and Jean-Michel Sauvage, Icair, said:

“The 3IA Côte d’Azur label is of significant importance to the partners of the Industrial Council for Artificial Intelligence Research (ICAIR), which represents 14 major international industrial companies present on the Côte d’Azur. The label serves as a guarantee of the quality of the AI research ecosystem present in the region, which is particularly important for our majority of non-French members of ICAIR. This recognition, interest and investment from the French government.”

MIAI Grenoble Alpes

Led by Grenoble Alpes University alongside Inria, CNRS, Inserm, CEA, Grenoble INP-UGA
Grenoble INP-UGA, Grenoble Ecole de Management, Grenoble-Alpes University Hospital and
and all of its private and industrial partners, MIAI Grenoble Alpes brings together all of the
Grenoble Alpes’ strengths in the field of AI.

The automatic decision-making systems currently deployed on a large scale are already affecting the lives of citizens. These systems raise many scientific challenges in terms of, for example, security, robustness, confidentiality, fairness and efficiency of data.

The 3IA MIAI Grenoble Alpes addresses all these challenges through interdisciplinary
collaborations and makes fundamental and applied contributions to artificial intelligence.
intelligence. In particular, MIAI Grenoble Alpes develops the next generation of AI models and systems, from hardware and embedded architectures to software, with a particular focus on health, environment and energy, and industry 4.0. It also examines the legal and ethical issues raised by the increasing use of AI tools.

In addition, MIAI Grenoble Alpes has launched, in partnership with the Minalogic competitiveness cluster, a series of webinars “Les éclairages de l’AI”, which disseminate AI culture to companies, and set up the “MIAInnovation Quaterly Journal”, a journal open to all people wishing to share their experience on AI innovation.

On the training side, the actions implemented within the short and long training courses
have made it possible to exceed the initial objective of increasing the number of AI graduates from 700 in 2018 to 1400 in
2022. This was made possible, on the one hand, by the redefinition of existing training courses and two certificates: “AI and application” and “Core AI”.

Claude Le Pape-Gardeux, Schneider Electric, states:

“Beyond the AI applications on which we were already active, the exchanges organized within the MIAI Institute convinced us to launch ourselves locally, with the teams of the University of Grenoble-Alpes, on long-term topics, such as the partially automatic analysis of normative or regulatory documents, ‘structured’ reinforcement learning or the systematic evaluation of the unreliability of a prediction.

The monthly meetings are particularly useful in generating thinking about potential applications and techniques. Similarly, we have enjoyed the multi-company brainstorming sessions organised by the ‘AI for data-driven and self-configurable supplychains’ chair. In addition, the launch in 2021 of multi-company short courses is an excellent initiative that we are currently looking into.”

PrAIrie

Supported by the CNRS, Inria, the Pasteur Institute, PSL University, the University of Paris and a club of industrial partners, the PrAIrie institute is distinguished by a concentration, exceptional on a global scale, of AI talent. The core of the institute, which is highly interdisciplinary, is currently made up of chairs whose specialties span the sub-disciplines of AI, from machine learning to optimization, including automatic language processing, vision, robotics, collective decision-making, biology, medicine and statistical physics. It has two major application areas: health and transport.

This 3IA is also part of the PariSanté Campus, whose ambition, as stated by the French President on December 4, 2020, is to make France a world leader in digital health. As part of the global landscape of artificial intelligence institutes, PRAIRIE is involved in structuring the European landscape by coordinating the Ile-de-France node of the ELLIS network.

The institute has a proactive policy for young people and for the promotion of diversity. This policy is reflected in the awarding of excellence grants to Master’s students, mobility grants as part of the PAISS summer school and the recruitment of six new junior chairs.

In terms of training, the dynamic set in motion by PrAIrie has enabled the creation ex-nihilo of new training courses on the foundations of AI (Master PSL IASD for example), but also at the interfaces, with a strong desire to develop dual skills (creation of a new double Bachelor’s degree in AI and Management). The implementation of a Transverse Program backed by new training modalities in medicine (Masterclasses, DU AI and health) illustrates this dynamic. In addition to initial training, work-study programs, continuing education and a high-level professional training program for the Société des Ingénieurs de l’Automobile have been created.

Jean-Luc Beylat, President Nokia Bell Labs France, assures:

“The PRAIRIE institute has, in a very short time, both brought together very good researchers in
The PRAIRIE institute has, in a very short space of time, both brought together very good researchers in Artificial Intelligence and federated around this project a wide range of leading industrialists who are strongly involved in the transformations brought about by AI.

This unique concentration of top international talent combined with the open innovation approach of industrialists makes PRAIRIE a unique place of excellence and attractiveness for AI in France. The quality of scientific publications and the growing number of companies involved are the best illustrations of this.

Grégoire Ferre, Senior Vice President, Faurecia, adds:

“PRAIRIE has been an innovative knowledge space for us, allowing us to support our
teams in building their own. PRAIRIE’s teams are focused on fundamental research but understand the operational reality of a manufacturer like us. I am delighted that we are part of this consortium.”

ANITI

Supported by the Toulouse Midi-Pyrénées Federal University, ANITI is developing a new generation of
generation of artificial intelligence, known as hybrid intelligence, which integrates machine learning techniques
ANITI is developing a new generation of hybrid artificial intelligence, combining data-driven machine learning techniques and models for expressing constraints and performing logical reasoning.

Based on an inclusive collaborative model, involving nearly 14 laboratories and some thirty private
and some thirty private partners, large groups, SMEs and emerging start-ups, ANITI
intervenes on a broad spectrum of interdisciplinary research, from fundamental to
research to application research, in order to collectively develop the building blocks
the use of trusted AI for critical applications in the fields of
mobility, transport, robotics/cobotics for the industry of the future
for example.

ANITI is also working to strengthen AI training on the Toulouse site, in a co-construction approach, involving all the academic actors of the site (schools and universities) but also the industrial actors needing resources trained in AI.

The evolution of the training offer and the attractiveness of the label proposed by ANITI have already allowed to increase the number of students training in AI to nearly 3400, that is to say an increase of 70% in two years, including a doubling on the core and integration perimeter.

Particular emphasis has also been placed on lifelong learning: highlighting the existing
of existing academic training courses and the creation of new courses,
actions to bring together students and companies.

With regard to the transfer to the economic sphere, ANITI relies on three levers:

• an original model of academic/company partnership that takes into account the specificities of different types of companies
• Support for the creation and development of start-ups through programmes developed with partners such as SATT Toulouse Tech Transfer
• the PEPITE entrepreneurship support scheme, as well as a strong registration in the local ecosystem to promote connections with all the players.Florent Teichteil Koenigsbuch, Airbus, Data Science Senior Scientistet Key Account Manager Aniti states:

“ANITI is a real opportunity for our engineers at Airbus to be trained in the most modern Artificial Intelligence methods. Our engineers will eventually be working in a plateau mode at B612 with academic researchers holding chairs in a virtuous mutual enrichment process: we bring industrial problems to the researchers, who in return enrich their portfolio of methods to solve them.

The exchange of issues and know-how with the other industrial partners of the consortium is also a strength for building tomorrow’s standards on certifiable, acceptable and collaborative AI, which are a sine qua non condition for embedding AI technologies on board our intelligent and communicating aircraft and satellite fleets.”

Translated from Retour sur le premier bilan des 3IA sur la recherche, la formation et le développement économique