Experimental Evaluation of the Schunk 5-Finger Gripping Hand for Grasping Tasks Steffen W. Ruehl, Christoper Parlitz, Georg Heppner, Andreas Hermann, Arne Roennau and Ruediger Dillmann Abstract— In order to perform useful tasks, a service robot needs to manipulate objects in its environment. In this paper, we propose a method for experimental evaluation of the suitability of a robotic hand for grasping tasks in service robotics. The method is applied to the Schunk 5-Finger Gripping Hand, which is a mechatronic gripper designed for service robots. During evaluation, it is shown, that it is able to grasp various common household objects and execute the grasps from the well known Cutkosky grasp taxonomy [1]. The result is, that it is a suitable hand for service robot tasks. I. I NTRODUCTION In contrast to industrial manipulators, service robots op- erate in unstructured environments with large varieties of different objects. In order to perform useful tasks, a service robot needs to manipulate objects in its environment. For example, in a ,,clean up the table” task, objects from the size of a toothpick up to a bottle may occur. The robot has to move those objects to a somehow defined place. Handling the complexity and variety of everyday environments is a major challenge for service robotics. The ability to relocate various objects reliably is an important skill required to tackle that challenge. Grasping depends on the gripper of the robot. In the indus- trial environment solutions with custom made, object specific gripper fingers may provide perfect form closure with one degree of freedom (DOF) two finger parallel grippers. But this is only feasible for the single object of the specific task. As mentioned, in service robot tasks, objects differ in size, shape and handling constraints. Thus, such solutions are not feasible. Instead, robot grippers with multiple fingers and DOF provide multiple potential contact points, which can be positioned according to the object using grasp control and planning algorithms. In an environment shared with humans, most objects are designed so they can be manipulated by humans. Thus, in order to create a gripper, which can handle those objects, many approaches mimic the human hand [2]. Anyhow, study- ing the work of those authors, the design of anthropomorphic hands has to deal with many challenges, such as the actuation within size and mass limits of the gripper, the transmission of forces and moments in the hand, sensor placement and Steffen W. Ruehl, Georg Heppner, Andreas Hermann, Arne Roennau, Ruediger Dillmann are with FZI Forschungszentrum Informatik, Karlsruhe, Haid-und-Neu-Str. 10-14, 76131 Karlsruhe, Germany, {ruehl|heppner|hermann|roennau|dillmann}@fzi.de Christoper Parlitz is with SCHUNK GmbH & Co. KG Spann- und Greiftechnik, Bahnhofstrasse 106-134, 74348 Lauffen Neckar, Germany christopher.parlitz@de.schunk.com Fig. 1. The Schunk 5-Finger Gripping Hand. control methods. Thus, optimized kinematics for grasping are not the main focus of many hand designs, and most hand designers have no experience in autonomous grasping. This leads to a situation, where different anthropomorphic hands perform very differently in grasping everyday objects due to kinematic properties. In this paper, we present an experimental evaluation of the grasping capabilities of the Schunk 5-Finger Gripping Hand (SVH) as shown in Fig. 1. Different everyday objects with a wide range of sizes are selected. Grasps from the Cutkosky [1] taxonomy are applied to demonstrate the capabilities of the hand. The paper is structured as follows: Sec. II presents a brief overview over robotic hands and grasping. Sec. III motivates and describes the mechanics and kinematics of the SVH. In Sec. IV we describe our experimental approach, which results are presented in Sec. V. In Sec. VI, we discuss the resulting insights and their relevance for the SVH and the design of hands for service robots in general. II. ROBOTIC HANDS AND GRASPING Today, a wide range of different robotic grippers is avail- able. [3] gives an overview over manipulation problems and gripping approaches. In this work, we structure the state of the art of robot hands by the number of fingers and DOFs. Industrial solutions for gripping are usually one DOF two finger grippers. Robust gripping is archived by either object specific finger design or by applying large forces. Two mechanical principles are common: parallel and angular movement of the fingers. The PR2 robot uses a parallel gripper [4]. [5] presents an angular gripper for manipulation with a six legged walking robot. © 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or is work in other works. DOI: 10.1109/ROBIO.2014.7090710 Robotics and Biomimetics (ROBIO), IEEE International Conference on 2465 - 2470, 2014