Temporal and geometrical analysis of human motor behavior to redundant solutions

Abstract

All goal-directed tasks are redundant at some level of analysis, to the extent that an infinite number of equivalent solutions exist for achieving the task goal. A solution manifold is composed by the combinations of execution variables that lead to a perfect task result. In previous studies of skilled task performance, variability in the execution variables was aligned with the solution manifold. This suggests the central nervous system determines the motor commands in the execution space where variability has the least detrimental effect on the result (task). These previous studies focused on the data set distributions in execution space, whereas recent studies have developed new analyses for the temporal structures. After reviewing these two analyses in detail, I examined their characteristics by applying these analyses to the experimental data set of a bimanual pinch-force generation task. The results demonstrated the two analyses could produce different conclusions for performance according to the task duration.