Interactive non-prehensile manipulation for grasping via POMDPs
Authors:
- 1 May 2013
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 3257-3264
Abstract
This paper develops a technique for an autonomous robot endowed with a manipulator arm, a multi-fingered gripper, and a variety of sensors to manipulate a known, but poorly observable object. We present a novel grasp planning method which not only incorporates the potential collision between object and manipulator, but takes advantage of this interaction. The natural uncertainty and difficulties in observation in such tasks is modeled as a Partially Observable Markov Decision Process (POMDP). Recent advances in point-based methods as well as a novel state space representation specific to the grasping problem are leveraged to overcome state space growth issues. Simulation results are presented for the combined localization, manipulation, and grasping of a small nut on a table.Keywords
This publication has 17 references indexed in Scilit:
- The next best touch for model-based localization2013
- Synthesizing grasp configurations with specified contact regionsThe International Journal of Robotics Research, 2010
- Grasping POMDPsProceedings 2001 ICRA. IEEE International Conference on Robotics and Automation (cat. No.01ch37164), 2007
- GraspIt!IEEE Robotics & Automation Magazine, 2004
- The mechanics of fine manipulation by pushing2003
- A complete algorithm for designing passive fences to orient parts2002
- Parts orienting with shape uncertainty2002
- An exploration of sensorless manipulationIEEE Journal on Robotics and Automation, 1988
- The motion of a pushed, sliding workpieceIEEE Journal on Robotics and Automation, 1988
- The Complexity of Markov Decision ProcessesMathematics of Operations Research, 1987