Bi-Directional Human-Robot Handover Using a Novel Supernumerary Robotic System

Abstract

Cooperative human-robot systems are the key technology to improve efficiency in agile production. In this work, a novel on-shoulder supernumerary robotic limb (SRL) is developed for human-robot collaboration. The SRL system is mainly composed of a 6-degree-of-freedom (DoF) lightweight robotic arm and a wearable and adjustable base, which allows different users to regulate the arm installation position to fit their physical dimensions and expand their workspace. As the SRL is close to the human head, nullspace-based collision avoidance is designed to achieve safe motion planning. Meanwhile, a Kalman filter is designed to actively predict human hand position and yield anticipatory motion control of our SRL in handover tasks. Furthermore, we conduct a bi-directional handover case study to verify the capability of the SRL system and the effectiveness of the proposed collaborative handover. The result shows that our control method can guarantee intuitive and natural handover behavior.