Résumé
This work presents the use of a brain-computer interface (BCI) system in order to steer a humanoid robot. We aim at designing a ready-to-use system allowing the user to manipulate the robot in an unknown environment. This design induces some constraints on the BCI system design that we present in this paper. Given these constraints, two steering paradigms based upon the steady-state visually evoked potentials (SSVEP) phenomenon are proposed, implemented on the HRP-2 humanoid robot, and compared on the basis of both robotics and interface criteria through an usage case study.