Stiffness control of 2-DOF exoskeleton for brain-machine interfaces

Previous demonstrations of brain-machine interfaces have shown the potential for controlling a neuroprosthesis under pure motion control, i.e. predicting end effector kinematics from neural ensemble activity. For real world tasks, however, pure motion control lacks the information required for versatile manipulation in which the dynamic interactions of forces and torques between the musculoskeletal system and the environment play a crucial role. Thus, our current efforts aim at enabling a subject using a brain-machine interface to volitionally control the mechanical impedance of the prosthetic device. Here we propose the use of a two-link arm exoskeleton to investigate upper limb stiffness in non-human primates. We show that the device can be used to experimentally measure end-point limb stiffness, as well as to control the stiffness when the exoskeleton is used in slave-robot mode. Experimental results show that this platform allows for both stiffness measurement and control of the robotic device.