Quantcast

Kinematic Analysis, Prototypation and Control of a Novel Gripper for Dexterous Applications

Research paper by Nahian Rahman, Luca Carbonari; Darwin Caldwell; Ferdinando Cannella

Indexed on: 04 Aug '18Published on: 01 Aug '18Published in: Journal of Intelligent & Robotic Systems



Abstract

Speed and flexibility are the primary concerns to whom a well designed industrial gripper should target. The first one leads to unquestionable pros in terms of production, while the second one to the ability of grasping and manipulating several payloads. However, these qualities are opposed to each other in terms of design requirements: speed requires a structure built of rigid bodies, flexibility would have to be favoured by the use of soft materials. As a common target, the human hand represents the most interesting inspiration source in this field, due to its natural dexterity and ability to perform in-hand manipulations. Thus, many bio-inspired or bio-mimicked grippers have been developed in the last decades with the final aim of replicating the terrific capabilities offered by the human hand. In such panorama, this paper presents the kinematic synthesis of a novel, modular, reconfigurable gripper, which is capable to manipulate a plurality of objects, being dexterous at the same time. Instead of using soft materials to achieve in-hand manipulation, the authors focused to use mechanisms to address the problem. The concept of manipulation is firstly evaluated in a multibody software environment, then a physical prototype was developed, and the necessary control laws were derived. Several experiments were conducted to test the effectiveness of the proposed structure. Results in terms of accuracy and repeatability are shown, and also the ability to address the three major tasks of grasp, in-hand manipulation and release with appropriate posture have been demonstrated.