The program
Confiance.ai
Led by a collective of 13 founding partners, major French industrial and academic players, Confiance.ai was a French technological research program designed to help manufacturers integrate trusted AI into their mission-critical systems.
Funded by France 2030 and an integral part of the national strategy for artificial intelligence, this program has enabled France to become a leader in the field of industrial and responsible AI. Its methodologies and tools are now state-of-the-art and serve as the technological foundation for the international Confiance.ai community.
The legacy of this program lives on through the community’s initiatives, which continue and accelerate the development of industrial and responsible AI.
More than 50 partners
already committed to building the future of trustworthy AI
Confiance.ai has brought together an ecosystem of over 50 partners, including manufacturers, researchers, start-ups and research laboratories, to develop methodologies and tools for integrating trusted AI into mission-critical systems.
This ecosystem has fostered the exchange of expertise and the creation of solutions to technological and regulatory challenges, while contributing to the definition of European regulations, such as the AI Act.
Today, thanks to the Confiance.ai community, this ecosystem continues to grow and evolve, supporting the expansion of trusted AI on an industrial scale.
The initial vision
to meet the challenge of industrializing AI
The Confiance.ai program was born of the conviction that artificial intelligence could transform industry and society, provided certain challenges are overcome. Manufacturers need simple, secure solutions to integrate AI into their critical systems with complete confidence.
Thanks to the program’s advances, many initial hurdles have been cleared, establishing a new state of the art in trusted AI engineering. However, with the emergence of new technologies such as generative AI, hybrid AI and human-machine collaboration, new challenges are emerging.
The Confiance.ai community is committed to meeting these new challenges, sustaining the results achieved, expanding its ecosystem and accelerating the industrialization of trusted AI to meet current and future industry needs.
Confiance.ai
use cases
| Partner | Use case | Data used |
 |
Air Liquide: Demand ForecastingPrediction of monthly industrial gas demand to optimize production and distribution. |
Time series |
 |
Renault group: Welding inspectionDetection of weld defects on production lines using a dataset of 15,000 HD photos to guarantee the reliability of AI methods and industrialize the process. |
Images |
 |
Valeo: Scene understanding with 2D perceptionDetection and identification of objects from images captured by a fisheye camera on cars, with the ultimate aim of processing videos for temporal consistency. |
Images |
 |
Airbus: Anti-collision system for UAVsReplacement of the vast ACAS Xu tables used to avoid mid-air collisions, with lightweight models that can be integrated into on-board systems, indicating the maneuvers to be carried out (change of course or altitude) to avoid a collision between a drone and another aircraft. |
Time series |
|
Air Liquide: Cylinder CountingIdentification and tracking of gas cylinders entering and leaving a warehouse using video captured by RGB cameras. |
Videos |
|
Renault group : Opinion analysis of customer commentsAnalysis of the sentiments expressed in comments left by Renault dealership customers, with specific challenges such as understanding irony. |
Texts |
 |
Thales: Interpretation of aerial photographsReal-time detection and classification of objects such as buildings, vehicles and other features in aerial photographs, using computer vision and supervised learning models. |
Images |
 |
Naval Group: Industrial visual inspectionDetection of anomalies on industrial parts from photos, identifying out-of-domain cases such as cracks to improve reliability and ensure that an alert is raised if the algorithm malfunctions. |
Images |
|
Airbus : Landing strip detectionDetection of landing strips from aircraft images. |
Images |
|
Airbus: Aircraft detection on satellite imagesDetection of aircraft on satellite images, with a major challenge linked to the performance gap between synthetic and real data. |
Images |
 |
Safran : Hybrid planning with temporal constraintsSolve planning problems with temporal, uncertain or disjunctive constraints, integrating reactive strategies to manage these uncertainties. |
Time series |
|
Thales: Enriching ontologies with NLPExtraction of key information from NOTAMs (aviation text messages) to build a knowledge graph, simplifying access to critical data for pilots. |
Tabular data |
 |
Atos: Visual similarity learning for re-identificationIdentification of people or objects detected by one camera in the field of another, based on visual similarity learning. |
Videos |
Ecosystem
of Confiance.ai
Large groups
Research centers / Universities
Standards / norms
Ecosystem
Start-up / PME
Supports
Last news
of Trustworthy AI
Look back at Confiance.ai Day 2024, the must-attend event for trustworthy AI
Confiance.ai Day, the must-attend event on trustworthy AI, was held on March 7, 2024 at the Direction Générale des Entreprises (DGE). For the 360 participants in attendance, it was a unique opportunity to meet the AI engineering and scientific communities and discover the methods and tools deployed by the Confiance.ai collective. Lire plus
Digital Factory Day: the place to discover Confiance.ai tools and methodologies
Home > News > The Confiance.ai collective is organizing the Digital Factory Day, an online event dedicated to disseminating ... Lire plus
The future of Confiance.ai: sustainability and new horizons for trustworthy AI
Home > News > The Confiance.ai program, the cornerstone of France's national strategy for artificial intelligence (AI), will ... Lire plus