Impedance Control Architecture for Robotic-Assisted Mold Polishing Based on Human Demonstration

Abstract

This article presents a computed-torque impedance control architecture for robotic-assisted mold polishing based on human skills.The mold polishing process for plastic injection is critical since tiny surface imperfections are easily visible in the plastic product, having strong implications in the overall quality. A Cartesian impedance control referred to the end-effector frame with posture optimization is proposed to tackle this problem. A collaborative robot manipulator from Franka Emika and a steel mold are used in the experiments to validate our approach. Human skills are captured both in comanipulation and free-hand modes, where position and force patterns are analyzed. Those patterns are then autonomously replicated by the robotic system, validating the proposed methodology.