Applications et Recherche pour une Intelligence Artificielle de Confiance
Objective of the project
The TRAIL initiative (Trusted AI Labs) was launched on 10 September 2020 in the presence of the Minister-President of the Walloon Government, with the support of the Digital Wallonia strategy.
It brings together the five French-speaking universities (UCLouvain, UMONS, ULB, ULiège and UNamur), the four accredited research centres active in AI (Cenaero, CETIC, Multitel and Sirris), as well as the Digital Agency and AI4Belgium.
Its ambition is to pool artificial intelligence research in the Wallonia-Brussels Federation and is concretely translated through the research project “ARIAC by DigitalWallonia4.ai”, based on a convention between the Walloon Region (SPW Recherche) and the actors forming the TRAIL consortium.
The ARIAC (Applications and Research for Trusted Artificial Intelligence) project spans 6 years and is structured around 5 WP (Work Package):
- human-AI interaction,
- Confidence Mechanisms for AI,
- model-AI integration,
- optimized AI implementations,
- TRAIL Factory.
TRAIL and the ARIAC research project are part of the DigitalWallonia4.ai regional programme, which aims to accelerate the development of artificial intelligence technologies in Wallonia.
Contribution of Multitel
In addition to participating in the various WP research projects, Multitel is in charge of the Factory axis whose objectives are:
- Responding to the challenges of today and tomorrow in the socio-economic fabric by directing applied research and providing key and differentiating technologies through major collective challenges
- Increase productivity by creating an environment conducive to testing through hardware, software and development of AI operationalization tools
- To position research actors in the innovation process by facilitating access to research results, improving transfer and ownership of technology components
- To contribute to the spread of research in Wallonia and Brussels and to improve interaction with the economic fabric by means of a one-stop-shop and demonstrators.
Multitel is also the leader of two major collective challenges:
- Machine learning with low supervision
- Trusted AI for critical systems.