An asynchronous BMI for autonomous robotic grasping based on SSVEF detection

Severely impaired persons could greatly benefit from assistive devices controlled by brain activity. However, the low information transfer rate of noninvasive neuroimaging techniques complicates complex and asynchronous control of robotic devices enormously. In this paper we present an asynchronous brain-machine interface (BMI) relying on autonomous grasp planning. The system enables a user to grasp and manipulate objects with a minimal set of commands. We successfully tested the system in a virtual environment with eight subjects. Our results suggest that the system represents a promising approach for real-world application of brain-controlled intelligent robotic devices.