Low level control in a semi-autonomous rehabilitation robotic system via a Brain-Computer Interface

In this work, a connection between a semi-autonomous rehabilitation robotic system and Brain-Computer Interfaces (BCI) is described This paper focuses on a system for user intervention in low-level movement control of an assistive robotic arm. The rehabilitation robotic system allows tetra-plegics to control the system with high-level commands (e.g., "grab the bottle"), and then to intervene in the execution of the task, if they see that something is going wrong. In such a case, the user gets the opportunity to continue the task with a low-level control of the robot arm. A system for such a control on a low abstraction level by a Brain-Computer Interface based on P300 and steady-state visual evoked potentials (SSVEP) will be described in this work.