The Use of BSN for Whole Body Motion Training for a Humanoid Robot

Sensor based motion capture system enables motion analysis with applications ranging from entertainment, healthcare and robotics. It can provide an intuitive interface for human to provide motion training and control a humanoid robot. In this paper, we propose a novel framework for the imitation of human motion for a humanoid robot. In the proposed framework, human motion data is directly captured from a wireless, wearable motion capture platform (Biomotion+). The reconstructed posture is then converted into joint angle trajectories. Due to the structural differences between the joints of the robot and those of the human, the trajectories are then optimized to satisfy the mechanical constraints of the robot and to maintain appropriate balance. To validate the proposed framework, different motion trajectories were verified. The results demonstrate the stability and effectiveness of the proposed framework to reproduce realistic human motion for a humanoid robot and the potential for a tele-rehabilitation application. The proposed framework offers a new way of imitating human motion for a humanoid robot.