SystemX Blog – Trust AI

Published on 10/05/2022

Since its emergence in the 1950s, artificial intelligence (AI) has been growing rapidly, accentuated by the development of available computing power and the Internet in the 2000s. However, despite its many successes, AI can still be a source of concern and has many challenges to face, both on the industry side and on the user side, in order to be accepted by all. Among these challenges, there are the locks related to energy, security and confidentiality but also to the trust of its users.

In this article, Fabien Tschirhart (research engineer, IRT SystemX) and Bertrand Braunschweig (scientific coordinator of the Confiance.ai program) look at the issue of AI adoption in critical systems.

> Read more here <

See also

News

[On the blog] How to strengthen the generative power of a scientific and industrial ecosystem? The case of IRT SystemX, a double impact RTO

Why can IRT SystemX be described as a "dual impact" RTO? What do we mean by "double impact"? The Scientific Direction of IRT SystemX has initiated a project in 2021 in collaboration with the Scientific Management Center (Mines Paris-PSL) to answer this question. The objective of this project is to characterize the nature of the…

Published on 06/14/2022

News

[Video] Autonomous Train: the adventure of obstacle detection

The DOS project (Detection of Obstacles and Reading of Lateral Signals) led by SNCF, SystemX, Alstom and Systra, is coming to an end after three years of research and testing. This work provides an essential technological brick for the autonomous train. Discover in video the main results of this project: Find out more.

Published on 07/04/2022

News

[SystemX blog] AI-enhanced physical simulation: challenges and validation approach

The use of numerical simulation in industry faces several limitations, such as the difficulty in modeling certain physical phenomena or the computational cost which can be very high. In this context, Artificial Intelligence (AI) has emerged as a promising avenue to address this issue and allow a more efficient use of physical simulators in decision…

Published on 12/08/2022