The statistical study of biomechanical arms

A statistical approach to the analysis of biomechanical manipulators is presented. The new method stems from previous studies on mechanical joint-actuated manipulators (see ibid. p.424-9). Motivated by the kinesiological aspects of the human arm, the authors demonstrate that biomechanical structures have better performance than standard manipulators. It is observed that joint-actuated robotic structures are nonoptimal because they have to support the direct impact of the operational requirements, whereas, muscle-actuated arms are intrinsically superior because the anatomic levers adapt the operational exigencies to the actuators. This observation provides a clear basis for the design of new mechanical manipulator structures, with performances close to those of the biological systems.<
>