Postural synergies of the UB Hand IV for human-like grasping

https://doi.org/10.1016/j.robot.2013.12.008Get rights and content

Highlights

  • A method to derive the three predominant synergies of the UB Hand IV is proposed.

  • The control strategy exploiting synergies for reach to grasp action is described.

  • Synthesis of new grasps not in the set used for synergies evaluation is achieved.

  • The method for synergies derivation is applied to other two robot hands.

  • The obtained synergies for different hands kinematics have been compared.

Abstract

In this paper, the postural synergy configuration subspace given by the fundamental eigengrasps of the UB Hand IV is derived from experiments, and a simplified synergy-based strategy for planning grasps is proposed. The objectives of this work are, on one side, the simplification of grasp synthesis in a configuration space of reduced dimensions and, on the other side, the attainment of a human-like behavior for anthropomorphic hands. A reference set of 36 hand postures, chosen with the goal of covering the entire grasp variety referring to a recently proposed taxonomy, has been considered for the evaluation of the hand synergies. With the aim of defining general properties of the three predominant synergies, the reference set of hand postures has been applied to other two anthropomorphic robot hands, and the obtained synergies have been compared with the ones computed considering the UB Hand IV kinematics. Moreover, the synthesis of new grasps, not contained in the reference set of hand postures, has also been achieved by means of the synergy subspace. The experiments carried out demonstrate that the adopted synergy-based planning method works efficiently for all the considered grasps even if not contained in the reference set used for the evaluation of the postural synergies.

Introduction

In order to interact with humans directly, the robots of the future will require enhanced manipulation capabilities similar to those of human beings. For this purpose, complex dexterous hands with advanced sensorimotor skills and human-like kinematics are needed. The human hand is an excellent example of dexterous bio-mechanical architecture with versatile capabilities to perform different kinds of tasks. The undergoing research in the field aims at the reproduction of human’s abilities not only by means of anthropomorphic design but also by adopting human-inspired control strategies. Recent advances in neuroscience have shown that control of the human hand during grasp is dominated by movements in a continuous configuration space of highly reduced dimensionality with respect to the number of DoFs [1], [2]. The goals of this paper are the simplification of grasp synthesis in a configuration space of reduced dimensions and the attainment of a human-like behavior for anthropomorphic hands. In order to allow a direct qualitative comparison with the results obtained on the human hand, the approach reported in  [1] has been adopted in this work both for the selection of the posture reference set and for the data analysis. The postural synergies, also called simply synergies or eigengrasps, of the UB Hand IV (see Fig. 1) have been derived using the Principal Component Analysis (PCA) by considering a reference set of 36 hand configurations, shown in Fig. 2, divided in four main groups. According to the grasp classification reported in [3], the first three groups of the reference set can be divided in precision, intermediate and power grasps. Each of these grasps involves objects used in the human everyday life. Since we are interested in planning the whole hand motion during the reach-to-grasp phase and not in describing the grasp only, a fourth group of open hand configurations has been added to the reference set for a complete coverage of the possible hand movements, allowing control of the execution of any grasp. It is worth noticing that considering different relative positions of the object with respect to the hand will result in different grasp configurations and, as a consequence, on different synergies.

The changes in synergies with respect to the change in position and orientation of the objects relative to the palm have not been investigated in this work for three main reasons.

  • (i)

    Usually humans adapt the hand position with respect to the object before grasping to execute it in a proper way (accordingly also to the task to be performed with the object); hence also for robot hands the grasps can be in some way standardized considering a single object/hand relative position for each grasp, as shown in Fig. 2, and adjusting the hand position before grasp execution.

  • (ii)

    If the grasp is robust enough (from the point of view of the grasp stability) and the object is not constrained, the hand may change the object position and orientation before the complete grasp is achieved in case one or more fingers contact the object before the others, resulting in an adaptation of the object position to the grasp.

  • (iii)

    By adopting the synergy-based grasp synthesis, we are not ensured that the location of the contact points, and then the object position too, will be the same considered for synergy derivation; hence constraining the object/hand relative position may result in a degradation of the grasp stability with respect to the case in which an adaptation of the object position by the hand itself during the grasp execution is allowed.

Aiming at the definition of common characteristics of the three predominant synergies, the postural synergies of other two anthropomorphic robot hands, a five-fingered hand with significant differences in the thumb kinematics and a four-fingered hand, have also been evaluated adopting the same posture reference set, and the results obtained from the different kinematic models have been compared. The PCA has been used in this work for synergy computation because, due to its linearity, it allows planning the reach-to-grasp movements of the robot hand by means of a simple linear interpolation of the synergies. Moreover, PCA is faster than other methods, it allows finding global optima and it shows good performance in representing new grasps, as the experiments on new object/grasps pair synthesis reported in this paper demonstrate. Indeed, experimental results show that the grasp planning performed on the UB Hand IV by considering the three predominant postural synergies allows synthesizing and performing the whole reference set of grasps, which includes objects with different shapes and dimensions. The new contribution of this work with respect to the state of the art, reported in the next section, consists in the enrichment of the results obtained in previous works  [4], [5], [6], by carrying out additional experiments in order to prove the efficiency of the synergy-based planning method in synthesizing new grasp/object pairs not contained in the reference set. Moreover, the general properties of the three predominant eigengrasps have been analyzed and compared with the ones of the other two anthropomorphic robot hands.

The paper is organized as follows: Section  2 describes research work related to postural synergies. In Section  3 the UB Hand IV design characteristics are illustrated together with its kinematics and the one of the other two robotic hands used for comparing the results on postural synergies. Section  4 provides the description of the method adopted for deriving the hand postural synergies and some evaluations on the kinematic patterns of the three predominant eigenpostures. Moreover, the synergy-based planning method adopted for the experimental execution of the grasps is described. In Section  5 the use of the three predominant synergies for the grasp synthesis, the comparison with the case of two synergies only and the synthesis of new grasp/object pairs beyond the set of 36 hand configurations are shown through experimental results. Finally, Section  6 provides the conclusions and a sketch of future work.

Section snippets

Related works

Recently, the studies of postural synergies have collected interest of many researchers not only belonging to the field of neuroscience but also working on control theory and mechanical design of artificial hands.

In  [1] the PCA has been used to calculate the postural synergies from real-world data collected on a variety of human hand postures by means of a data glove. Moreover, the authors show that a wide set of hand postures during grasping operation evolves continuously within a linear

Description of the robotic hands used for synergies evaluation

In this section, the devices considered in this paper for the evaluation of the postural synergies are briefly described. Particular emphasis is given to the UB Hand IV since it has been adopted also in the experiments hereby reported. The evaluation of the postural synergies has also been applied to other two robot hands with the aim of deriving common features that do not depend on the particular kinematics nor on the number of fingers.

It is worth noticing that, since the aim of this work is

Postural synergies of the UB Hand IV

Drawing inspiration from the studies on the human hand motion [1], and since the UB Hand IV presents human-like kinematics, a set of postural synergies of the UB Hand IV configuration space has been found. The details of this analysis are reported in the following.

Experimental evaluation of the synergy-based grasp synthesis

The hand controller developed in the MATLAB/Simulink environment is based on the RTAI-Linux realtime operating system. The MATLAB Realtime Workshop toolbox has been used for the automatic generation of the real-time application of the UB Hand IV controller. The user interface to the real-time application has been implemented by means of the Simulink External Mode capabilities, for which the RTAI-Linux support has been purposely developed. In the experiments, starting from the zero-offset

Conclusions and future work

In this paper the experimental evaluation of the three predominant postural synergies of the UB Hand IV by means of PCA has been presented. For this purpose, a suitable reference set of hand postures has been taken into account. The kinematic patterns of the three predominant postural synergies have been described and the benefit given by the introduction of the third synergy, added to improve over a previous work, has been enlightened. Moreover, the introduction of the third synergy has been

Acknowledgments

The authors want to thank all those who have contributed in the recent years to the development of the UB Hand IV prototype. Further, the support of Umberto Scarcia and Ugo Fabrizi during the experiments described in this paper is gratefully acknowledged.

Fanny Ficuciello has obtained the Laurea degree magna cum laude in Mechanical Engineering from the University of Naples Federico II in 2007. She received the Ph.D. degree in Computer and Automation Engineering at the University of Naples Federico II, in November 2010. From September to March 2010 she was a visiting scholar in the Control Engineering Group at the University of Twente (Netherlands) under the supervision of Prof. Stefano Stramigioli. She is member of PRISMA (Projects of industrial

References (42)

  • F. Ficuciello, G. Palli, C. Melchiorri, B. Siciliano, Experimental evaluation of postural synergies during reach to...
  • L. Villani et al.

    Grasping and control of multifingered hands

  • M. Ciocarlie, C. Goldfeder, P. Allen, Dimensionality reduction for hand-independent dexterous robotic grasping, in:...
  • M. Ciocarlie et al.

    Hand posture subspaces for dexterous robotic grasping

    Int. J. Robot. Res.

    (2009)
  • S. Sun, C. Rosales, R. Suarez, Study of coordinated motions of the human hand for robotic applications, in: Proc. IEEE...
  • T. Wimboeck, B. Jan, G. Hirzinger, Synergy-level impedance control for a multifingered hand, in: Proc. IEEE Int. Conf....
  • G. Gioioso, G. Salvietti, M. Malvezzi, P. Prattichizzo, Mapping synergies from human to robotic hands with dissimilar...
  • C. Brown, H. Asada, Inter-finger coordination and postural synergies in robot hands via mechanical implementation of...
  • D. Prattichizzo, M. Malvezzi, A. Bicchi, On motion and force controllability of grasping hands with postural synergies,...
  • M. Gabiccini, A. Bicchi, On the role of hand synergies in the optimal choice of grasping forces, in: Proc. of Robotics:...
  • A. Bicchi et al.

    Modelling natural and artificial hands with synergies

    Philos. Trans. R. Soc. B

    (2011)
  • Cited by (40)

    • The PRISMA Hand I: A novel underactuated design and EMG/voice-based multimodal control

      2020, Engineering Applications of Artificial Intelligence
      Citation Excerpt :

      Both the transmission system and the compliance introduced by elastic tendons allow the hand adapting to different object shape and size. About the second motor, studies on fully actuated robotic hands demonstrate that thumb adduction/abduction motion depends especially from the third synergy that is responsible of successful precision grasps (Gabiccini and Bicchi, 2010; Ficuciello et al., 2014, 2012b, 2016a; Ficuciello, 2019b). Considering the important role of the thumb on grasp stability, in particular for the opposition role, we decided to use the second motor for adduction/abduction thumb motion.

    • cHand: Open Source Hand Posture Visualization in CHAI3D

      2023, 2023 International Symposium on Medical Robotics, ISMR 2023
    View all citing articles on Scopus

    Fanny Ficuciello has obtained the Laurea degree magna cum laude in Mechanical Engineering from the University of Naples Federico II in 2007. She received the Ph.D. degree in Computer and Automation Engineering at the University of Naples Federico II, in November 2010. From September to March 2010 she was a visiting scholar in the Control Engineering Group at the University of Twente (Netherlands) under the supervision of Prof. Stefano Stramigioli. She is member of PRISMA (Projects of industrial and service robotics, mechatronics and automation) research group on robotics headed by Prof. Bruno Siciliano. Currently she is holding a Post Doctoral position at the University of Naples Federico II. Her research activity is focused on robotics, in particular on human–robot interaction, grasping and manipulation control of anthropomorphic hand–arm systems.

    Gianluca Palli received the Laurea and the Ph.D. degrees in automation engineering from the University of Bologna, Italy, in 2003 and 2007, respectively. He was a visiting student at the Robotic Institute of the Deutsches Zentrum fur Luft-und Raumfahrt (DLR), Munich, Germany, in 2006. Currently he is Assistant Professor at the University of Bologna. His research interests include design and control of robotic hands, modeling and control of robots with variable stiffness joints, design of compliant structures and actuation systems for robotics applications and development of real time systems for automatic control applications.

    Claudio Melchiorri (M’92-SM’03) received the Laurea degree in electrical engineering and the Ph.D. degree from the University of Bologna, Bologna, Italy, in 1985 and 1990, respectively. In 1998, he was an Adjunct Associate in engineering with the Department of Electrical Engineering, University of Florida, Gainesville. In 1990–1991 he was a Visiting Scientist with the Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge. Since 1985 he has been with the Department of Electrical, Electronic and Information Engineering, University of Bologna, where he is currently a Full Professor of robotics. His research interests include dexterous robotic manipulation, haptic interfaces, telemanipulation systems, advanced sensors, and nonlinear control. He is the author or coauthor of about 270 scientific papers presented at conferences or published in journals and of 13 books on digital control and robotics.

    Bruno Siciliano is Professor of Control and Robotics, and Director of the PRISMA Lab in the Department of Computer and Systems Engineering at University of Naples Federico II. His research interests include force and visual control, human–robot interaction and service robotics. He has co-authored 7 books, 70 journal papers, 170 conference papers and book chapters. He has delivered 100 invited lectures and seminars at institutions worldwide, and he has been the recipient of several awards. He is a Fellow of IEEE, ASME and IFAC. He has served on the editorial boards of several peer-reviewed journals and has been chair of program and organizing committees of several international conferences. He is Co-Editor of the Springer Tracts in Advanced Robotics, and of the Springer Handbook of Robotics, which received the PROSE Award for Excellence in Physical Sciences & Mathematics and was also the winner in the category Engineering & Technology. His group has been granted twelve European projects. Professor Siciliano is the Past-President of the IEEE Robotics and Automation Society.

    View full text