Development of Human-Like Motion Planning Method Using Task Constrained for Humanoid Robotic Arm With Non-Offset Wrist

Abstract

Abstract In industry 4.0, to increase flexibility of a robotic arm system, high-DOFs (degrees of freedom) robotic arm system is used for different applications. A traditional way of dealing with high-DOFs configuration control is to utilize kinesthetic teaching in Cartesian space by guiding the robot’s end-effector to targets. Usually, the teaching process takes a long time, and the configuration of the robotic arm makes them only workable for specific application needs. This research presents a novel mechanical design of 7-DOFs humanoid robotic arm and an algorithm to generate human-like motions for the anthropometric robotic arm to solve the problem. Because the humanoid robotic arm exists one redundant DOF, in this paper, this extra DOF is utilized to optimize the human-like motion in performing tasks with minimum angle variance, adjusting force transmission with manipulability ellipsoids, and avoiding obstacles in the environment. Finally, a 7-DOFs humanoid robotic arm is designed and used in a task which is typically performed by bartenders, namely, pouring wine into the glass.