Exploiting affordance symmetries for task reproduction planning

Many tool use tasks exhibit symmetries, such as the fact that a carrot can be cut anywhere along the blade of a knife, or that a nail can be struck from many directions by a hammer. This paper uses a previously proposed concept of affordance symmetries and extends it to enable such freedoms to be captured for robot programming by demonstration (RPbD). A reproduction planner is proposed which leverages these concepts by reducing the planning space to one where robot redundancy and symmetries are the only redundancies. Naturally, this improves planning time, but also allows the robots to properly exploit task redundancies to remain within the kinematic limitations of the robot. To illustrate the performance of the algorithm a simple example of teaching the robot to pour from one cup to another is used. The robot was able to perform the task successfully because it exploited task symmetries.

[1]  Aude Billard,et al.  Dynamical System Modulation for Robot Learning via Kinesthetic Demonstrations , 2008, IEEE Transactions on Robotics.

[2]  Jonathan Claassens,et al.  An RRT-based path planner for use in trajectory imitation , 2010, 2010 IEEE International Conference on Robotics and Automation.

[3]  Markus Schneider,et al.  Robot Learning by Demonstration with local Gaussian process regression , 2010, 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[4]  Yoshihiko Nakamura,et al.  Mimesis Scheme using a Monocular Vision System on a Humanoid Robot , 2007, Proceedings 2007 IEEE International Conference on Robotics and Automation.

[5]  Sethu Vijayakumar,et al.  Synthesising Novel Movements through Latent Space Modulation of Scalable Control Policies , 2008, SAB.

[6]  Jun Morimoto,et al.  Task-Specific Generalization of Discrete and Periodic Dynamic Movement Primitives , 2010, IEEE Transactions on Robotics.

[7]  Aude Billard,et al.  On Learning, Representing, and Generalizing a Task in a Humanoid Robot , 2007, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[8]  Changchang Wu,et al.  SiftGPU : A GPU Implementation of Scale Invariant Feature Transform (SIFT) , 2007 .

[9]  Siddhartha S. Srinivasa,et al.  The MOPED framework: Object recognition and pose estimation for manipulation , 2011, Int. J. Robotics Res..

[10]  Christopher G. Atkeson,et al.  Constructive Incremental Learning from Only Local Information , 1998, Neural Computation.

[11]  Rajesh P. N. Rao,et al.  Learning Nonparametric Models for Probabilistic Imitation , 2006, NIPS.

[12]  Siddhartha S. Srinivasa,et al.  Efficient multi-view object recognition and full pose estimation , 2010, 2010 IEEE International Conference on Robotics and Automation.

[13]  Oliver Kroemer,et al.  A flexible hybrid framework for modeling complex manipulation tasks , 2011, 2011 IEEE International Conference on Robotics and Automation.

[14]  S. LaValle.  Rapidly-exploring random trees : a new tool for path planning , 1998 .

[15]  Rüdiger Dillmann,et al.  Representation and constrained planning of manipulation strategies in the context of Programming by Demonstration , 2010, 2010 IEEE International Conference on Robotics and Automation.

[16]  Yiannis Demiris,et al.  Generalising human demonstration data by identifying affordance symmetries in object interaction trajectories , 2011, 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[17]  Jun Nakanishi,et al.  Movement imitation with nonlinear dynamical systems in humanoid robots , 2002, Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292).

[18]  Jonathan Claassens,et al.  An analytical solution for the inverse kinematics of a redundant 7DoF Manipulator with link offsets , 2010, 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems.