Task-dependent impedance improves user performance with a virtual prosthetic arm

Recent studies of neuromotor control have shown that humans modulate the impedance of their arms for different tasks, motivating the development of a prosthetic arm with user-selectable impedance characteristics. While impedance control has been extensively studied in robotics, prosthetic arms present a unique combination of human and robotic control, in which human capabilities and preferences play an important role. To understand the desirable impedance characteristics of prosthetic arms, we studied human control of a one-degree of-freedom virtual prosthetic limb with variable stiffness and damping in tasks involving (1) force minimization and (2) trajectory tracking. Subjects performed best with different patterns of impedance modulation depending on task goals and available feedback, suggesting that the ability to modulate the impedance of a prosthetic limb may be beneficial to the wearer. The results of this study inform the future design of prosthetic limbs in which the wearer can vary limb impedance to improve performance in a variety of manipulation tasks.