Development of Robotic Hand with Passive Variable Transmission

Abstract

This paper presents a way to improve performance of a tendon driven robotic hand using Passive Variable Transmission. The concept of research was inspired by the human pulley mechanism which changes the moment arm of tendon by pulleys, which makes it possible for a human hand to rapidly and powerfully grasp an object. To mimic the pulley mechanism of the human hand, Passive Variable Transmission is applied, which changes the path of the tendon wire passively as the human hand pulleys do. This PVT Mechanism is a transmission for a tendon driven mechanism, which varies a moment arm of a tendon wire by changing the lengths of compliant material when the tension of the tendon wire is changing. If the tension of the tendon wire is small, the moment arm remains short. As the tension gets bigger, the moment arm increases. Thus, when the joint rotates without load, the moment arm is short and the joint rotates rapidly. On the other hand, if the joint is blocked by an obstacle, the tension increases, which increases the moment arm, and the joint generates a bigger moment. The goal of this research was to develop a suitable PVT design for a small finger structure. After trying a number of concept designs, the spring type PVT was chosen. This spring type PVT has been tested to certificate that this PVT can change the tendon wire moment arm passively. Before fabricating the robotic hand, the parametric study was conducted. Based on the results of the parametric study, parameters of the spring type PVT was decided, and a robotic hand was fabricated. By comparing with an existing tendon driven robotic hand, H2 hand from Meka, it was confirmed that the fabricated robotic hand used 20% rated output and showed 40% grasping performance. There has been a structural limitation on robotic hands in mimicking the grasping performance of human hands. This paper will show prospective view of using the Passive Variable Transmission Mechanism to improve the grasping performance of robotic hands.