Estimating Physical Assistance Need Using a Musculoskeletal Model

Technologies that provide physical assistance during tasks are often required to provide assistance specific to the task and person performing it. An example is robotic rehabilitation in which the assistance-as-needed (AAN) paradigm aims to provide operators with the minimum assistance required to perform the task. Current approaches use empirical performance-based methods which require repeated observation of the specific task before an estimate of the needed assistance can be determined. In this paper, we present a new approach utilizing a musculoskeletal model (MM) of the upper limb to estimate the operator´s assistance needs with respect to physical tasks. With capabilities of the operator defined at the muscular level of the MM, an optimization model is used to estimate the operator´s strength capability. Strength required to perform a task is calculated using a task model. The difference or gap between the operator´s strength capability and the strength required to execute a task forms the basis for the new AAN paradigm. We show how this approach estimates the effects of limb pose, load direction, and muscle impairments on a person´s ability to perform tasks.