Joint coupling design of underactuated hands for unstructured environments

Aaron M Dollar,Robert D Howe

doi:10.1177/0278364911401441

Aaron M Dollar, Robert D Howe

Open Access

https://doi.org/10.1177/0278364911401441

Copy DOI

Abstract

This paper examines joint coupling in underactuated robotic grippers for unstructured environments where object properties and location may not be well known. A simplified grasper consisting of a pair of two-link planar fingers with compliant revolute joints was simulated as it grasped a target object. The joint coupling configuration of the gripper was varied in order to maximize successful grasp range and minimize contact forces for a wide range of target object sizes and positions. The number of actuators was also varied in order to test performance for varying degrees of underactuation. A normal distribution of object position was used to model sensing uncertainty and weight the results accordingly. There are three main results: distal/proximal joint torque ratios of around 1.0 produced the best results, both for cases in which sensory information available for the task was poor and when sensing was good; an actuator for each gripper finger performs no better than a single actuator for both fingers; and that for good sensing, the gripper should be positioned off-center from the object, resulting in an increased lever arm and lower unbalanced contact forces on the object.

Highlights

The ability to grasp diverse objects in human settings will provide robots with the functionality to perform a wide range of tasks in homes and workplaces
This paper examines joint coupling in underactuated robotic grippers for unstructured environments where object properties and location may not be well known
There are three main results: distal/proximal joint torque ratios of around 1.0 produced the best results, both for cases in which sensory information available for the task was poor and when sensing was good; an actuator for each gripper finger performs no better than a single actuator for both fingers; and that for good sensing, the gripper should be positioned off-center from the object, resulting in an increased lever arm and lower unbalanced contact forces on the object

Summary

Introduction

The ability to grasp diverse objects in human settings will provide robots with the functionality to perform a wide range of tasks in homes and workplaces. The majority of research in robotic grasping and manipulation has attempted to address this problem through elaborate multifingered hands, combined with tactile sensing and sophisticated planning and control algorithms, and often following an anthropomorphic approach While this methodology may provide good performance in the long term, it involves considerable systems-level complexity and significant implementation costs, and there has been little demonstrated success to date in grasping objects under the uncertainties typical of human environments. Much of the functionality of a hand can be retained by careful selection of joint compliance and coupling schemes, reducing the number of actuators and the overall complexity of the grasping mechanism Many of these grippers are ‘underactuated’, with fewer actuators than degrees of freedom See Dollar and Howe (2006) and Birglen et al (2008) for extensive surveys of underactuated hands

Objectives

Methods

Results

Conclusion