The central nervous system stabilizes unstable dynamics by learning optimal impedance

To manipulate objects or to use tools we must compensate for any forces ari sing from interaction with the physical environment. Recent studies indicat e that this compensation is achieved by learning an internal model of the d ynamics(1-6), that is, a neural representation of the relation between moto r command and movement(5,7). In these studies interaction with the physical environment was stable, but many common tasks are intrinsically unstable(8 ,9). For example, keeping a screwdriver in the slot of a screw is unstable because excessive force parallel to the slot can cause the screwdriver to s lip and because misdirected force can cause loss of contact between the scr ewdriver and the screw. Stability may be dependent on the control of mechan ical impedance in the human arm because mechanical impedance can generate f orces which resist destabilizing motion. Here we examined arm movements in an unstable dynamic environment created by a robotic interface. Our results show that humans learn to stabilize unstable dynamics using the skilful an d energy-efficient strategy of selective control of impedance geometry.