Guillaume is in charge of the “Evaluation of Artificial Intelligence” department at LNE. In this context, he conducts scientific and technical work on the definition of evaluation protocols, metrics, and test environments (databases, simulators, physical test benches) for intelligent systems.
He coordinates various consortia and working groups at the national, European and international levels with the aim of defining and eventually converging on a fundamental and operational metrology for AI (Guillaume coordinates in particular the European consortium METRICS, made up of 17 evaluation centers for intelligent robots, as well as a working group of about fifteen industrial companies aiming to define a first certification standard for AI).
It also participates intuitu personae in standardization committees on AI and robotics (Afnor IA, ISO/IEC JTC1/SC42, CEN-CENELEC IA Group Focus, UNM81, etc.).