One of the major goals of robotics research is the construction of humanoid robots that enter human-centred environments. To promote man-machine interaction in daily life robots should have a roughly humanoid shape and use human-like movements. It is therefore crucial to understand how humans coordinate their upper-body movements during daily motor tasks. One of the key assumptions in motor control research is that information about the process of movement planning and control can be deduced from behavioural regularities. The purpose of the following study was to analyse multi-joint goal-directed movements in 3D space in a natural environment with respect to the identification of such regularities. This paper presents a study in which 20 subjects had to point to different targets in a kitchen. All pointing movements were tracked using an IRtracking system and joint angles were calculated using a kinematic model. The limb movements were analyzed in extrinsic and intrinsic kinematic Coordinates and the results were discussed against the background of theoretical assumptions of biological motor control and the generation of human-like movements on humanoid robots.
© Copyright 2010 International Journal of Computer Science in Sport. Sciendo. Alle Rechte vorbehalten.