Robust neuromechanical control in uncertain load environments

Humans can control their actions in unpredictable environments, such as adjusting a tennis serve on a windy day. Here we examined how our nervous system uses both predictive and feedback control in response to uncertain mechanical loads. Participants reached to a target in the horizontal plane. An exoskeleton applied loads perpendicular to the movement of the hand. Across blocks of trials, load magnitude was either constant or randomly drawn from a uniform probability distribution. A robust optimal feedback control (rOFC) model predicted hand trajectories and enhanced feedback gains with uncertain loads. We observed enhanced muscle feedback responses with uncertain loads. Our data suggests that the nervous system uses robust control by enhancing feedback gains when moving in an uncertain environment.