Real-time estimation of human impedance for haptic interfaces

Humans significantly vary the impedance of their limbs during many manipulation tasks and during interaction with their environment. Humans vary their impedance to reduce contact forces, to increase positional control, and to stabilize unstable dynamics. The neuromuscular system modulates overall limb impedance by strategically selecting appropriate limb configurations and by selective activation of musculature. To date, haptic interfaces neither respond to nor measure these effects. We seek a real-time measurement technique to estimate user impedance, with the hope of enabling variable impedance based control of telerobots and haptic environments. We propose that such a measure might be generated through the introduction of controlled vibrations to the user through a haptic device. A model for such a system and its implications on device design is described. The results of preliminary study using a prototype device are compared to reference measurements of user stiffness and grip force.