Functional electrical stimulation is a promising technique for restoring arm function to those with paralysis from a high spinal cord injury. While simple controllers are easy to implement, model-based controllers are likely better equipped to leverage the arm’s kinematic and dynamic complexity, particularly for the high variations associated with functional arm movement. One modelling technique for a model-based controller is Gaussian Process Regression. Previous simulation work has shown promise leveraging whole-arm error data to identify the arm’s various subsystems, but used perfect simulated data. We asked caregivers to correct a robotic arm’s movement as simulated muscles generated torque. The simulated muscles were controlled as if they were electrically stimulated human arm muscles. This study demonstrates non-expert caregivers’ ability to collect this error data via whole-arm corrections, and provides insight into their ability to improve arm subsystem models made with Gaussian Process Regression. Despite significant error in caregivers’ ability to provide force corrections to hold the robot in a static configuration, these corrections were leveraged to significantly improve muscle models; the muscles that improved the most were the ones primarily used to move the physiologically actuated robot.

Functional electrical stimulation (FES) is a promising technique for restoring reaching ability to individuals with tetraplegia. To this point, the complexities of goal-directed reaching motions and the shoulder-arm complex have prevented the realization of this potential in full-arm 3D reaching tasks. We trained a Gaussian process regression model to form the basis of a feedforward-feedback control structure capable of achieving reaching motions with a paralyzed upper limb. Over a series of 95 reaches of at least 10 cm in length, the controller achieved an average accuracy (measured by the Euclidean distance of the wrist to the final target position) of 3.8 cm and an average error along the path of 3.5 cm. This controller is the first demonstration of an accurate, complete-arm, FES-driven 3D reaching controller to be implemented with an individual with tetraplegia.

In this paper we present a method to stimulate multiple muscles in a human arm to perform interaction tasks, using an implanted Functional Electrical Stimulation (FES) neuroprosthesis. The unstable effect arising from interaction tasks is considered, and the arm stability is directly treated as one of the control objectives in the controller design. By exploiting the kinematic and muscular redundancy of the system, we can control the interaction force and the arm's stiffness property simultaneously, thus ensuring the stable execution of interaction tasks. A representative example of such interaction tasks, namely the “pushing with a stick” task, is simulated. It is found that using our proposed controller, as compared to a previously developed feedforward FES controller that does not consider arm stiffness or stability, the stability of the arm is guaranteed while the task of force control is correctly achieved.

We present a method to experimentally identify the inverse dynamics of a human arm. We drive a person's hand with a robot along smooth reaching trajectories while measuring the motion of the shoulder and elbow joints and the force required to move the hand. We fit a model that predicts the shoulder and elbow joint torques required to achieve a desired arm motion. This torque can be supplied by functional electrical stimulation of muscles to control the arm of a person paralyzed by spinal cord injury. Errors in predictions of the joint torques for a subject without spinal cord injury were less than 20% of the maximum torques observed in the identification experiments. In most cases a semiparametric Gaussian process model predicted joint torques with equal or less error than a nonparametric Gaussian process model or a parametric model.

We present a method for controlling a neuroprosthesis for a paralyzed human arm using functional electrical stimulation (FES). The subject has surgically implanted electrodes for stimulating muscles in her shoulder and arm. Using input/output data, a model is identified that describes the mapping from muscle stimulations to the endpoint force measured at the subject's hand. To compute the muscle stimulations given a target endpoint force the model is inverted. Because the system is redundant, we compute the inverse by minimizing muscle activations and use this inverse for feedforward control. This is the first published demonstration with a human subject with a high spinal cord injury of an FES controller that treats the arm with shoulder and elbow as a multiple-input multiple-output system and can achieve arbitrary goals.