A hybrid system for knowledge-based synthesis of robot grasps

Addresses the grasp synthesis problem arising in task planning for robotic dextrous hands. For this purpose, a hybrid architecture is proposed, which relies on symbolic and subsymbolic computations to exploit heterogeneous sources of knowledge, such as practical experience learned through experiments with a real device, heuristic rules gained from human observation, geometric reasoning and, when applicable, analytical results. After a preliminary discussion of representation levels and techniques, this paper describes the design of a tool for selecting the feasible grasps of a robotic hand under various situations and for ranking them according to task-oriented criteria. The interaction of a rule-based expert system with a neural network-based classifier provides support to both explicit reasoning and direct learning from experience. The features taken into account by the tool are object geometry, hand kinematic capabilities, task requirements, (in terms of both accessibility and robustness) and workspace constraints.

[1]  Katsushi Ikeuchi,et al.  Grasp Recognition Using The Contact Web , 1992, Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems.

[2]  Stephen C. Jacobsen,et al.  The UTAH/M.I.T. Dextrous Hand: Work in Progress , 1984 .

[3]  J. Kenneth Kinematic and force analysis of articulated hands , 1985 .

[4]  Mark R. Cutkosky,et al.  On grasp choice, grasp models, and the design of hands for manufacturing tasks , 1989, IEEE Trans. Robotics Autom..

[5]  Stephen I. Gallant,et al.  Connectionist expert systems , 1988, CACM.

[6]  R. Howe,et al.  Human grasp choice and robotic grasp analysis , 1990 .

[7]  Thea Iberall,et al.  The nature of human prehension: Three dextrous hands in one , 1987, Proceedings. 1987 IEEE International Conference on Robotics and Automation.

[8]  C. Bonivento,et al.  The University of Bologna Robotic Hand Project: current state and future developments , 1991, Fifth International Conference on Advanced Robotics 'Robots in Unstructured Environments.

[9]  Huan Liu,et al.  Building a generic architecture for robot hand control , 1988, IEEE 1988 International Conference on Neural Networks.

[10]  Thea Iberall A Neural Network for Planning Hand Shapes in Human Prehension , 1988, 1988 American Control Conference.

[11]  Sharon A. Stansfield,et al.  Representing generic objects for exploration and recognition , 1988, Proceedings. 1988 IEEE International Conference on Robotics and Automation.

[12]  Andrew A. Goldenberg,et al.  Qualitative theory of robot grasping-A formalisation of 'intuitive' robot grasping , 1990, Proceedings. 5th IEEE International Symposium on Intelligent Control 1990.

[13]  Bernard Roth,et al.  Analysis of Multifingered Hands , 1986 .

[14]  Huan Liu,et al.  The multi-dimensional quality of task requirements for dextrous robot hand control , 1989, Proceedings, 1989 International Conference on Robotics and Automation.

[15]  Stephen H. Lane,et al.  Integrating neural networks and knowledge-based systems for robotic control , 1989, Proceedings, 1989 International Conference on Robotics and Automation.

[16]  George A. Bekey,et al.  A strategy for grasp synthesis with multifingered robot hands , 1987, Proceedings. 1987 IEEE International Conference on Robotics and Automation.

[17]  Antonio Bicchi,et al.  Whole-Hand Manipulation: Design of an Articulated Hand Exploiting All Its Parts to Increase Dexterity , 1993 .

[18]  Damian M. Lyons,et al.  A simple set of grasps for a dextrous hand , 1985, Proceedings. 1985 IEEE International Conference on Robotics and Automation.

[19]  E. Faldella,et al.  Grasp synthesis for a dextrous robotic hand using a multi-layer perceptron , 1991, [1991 Proceedings] 6th Mediterranean Electrotechnical Conference.