A new approach to the analysis and design of robot manipulators is presented. The novel feature resides on a non-standard formulation to the modelling problem. Usually, system descriptions are based on a set of differential equations which, in general, require laborious computations and may be difficult to analyse. These facts motivate the need of alternative models based on different mathematical concepts. The proposed statistical approach to the Fourier modelling gives clear guidelines towards the optimisation of the robot kinematics and points out structural characteristics of the trajectory planning algorithms.

A statistical analysis of biomechanical manipulators is presented. The approach presented stems from studies of the properties of joint-actuated manipulators. Motivated by the kinesiological aspects of the human arm, it is demonstrated that biomechanical structures have better performances than standard manipulators. In fact, joint-actuated robotic structures are non-optimal because they have to support the direct impact of the operational requirements, while muscle-actuated arms are superior because they have anatomic levers and actuators more adapted to the transients required by the operational tasks. Therefore, these results are a step towards the design of a new generation of better mechanical manipulators.>