Abstract This paper presents a cognitive architecture for controlling robotic agents via a brain–computer interface (BCI). Aimed at patients with motor and speech impairments, the system translates high-level user intentions, decoded from EEG (Electroencephalography) signals, into conflict-free navigation for a group of wheelchairs. Experiments confirm the feasibility of multi-agent control based on a single global command, without requiring direct human intervention.
Akopyan et al. (Mon,) studied this question.