Abstraction of Assembly Tasks to Automatically Generate Robot Motion from Observation

ion of Assembly Tasks to Automatically Generate Robot Motion from Observation Jun Takamatsu,† Koichi Ogawara,† Hiroshi Kimura†† and Katsushi Ikeuchi† The ability of robots to learn human tasks from observation is one of the long-awaited demands in the field of robotics. Here, we limit the scope of the target tasks to assembly tasks because the domain is one of the central research areas in robotics and has a wide application area. We propose a method to recognize assembly tasks based on transitions of contact relations between two polyhedral objects. Concretely speaking, we propose a method for: (1)representing task models based on such transitions, (2)determining correct contact relations and transitions of them from noise-contaminated visual information, and (3) generating a corresponding sequence of movement-primitives (referred to as sub-skills) from those task models. We have implemented the system on our robot, with real time stereo system and a pair of arms with dextrous hands. In actuality, we have demonstrated the system’s effectiveness.

[1]  H. Kimura,et al.  Acquiring hand-action models in task and behavior levels by a learning robot through observing human demonstrations , 2000 .

[2]  Tamio Arai,et al.  Teaching of grasp/graspless manipulation for industrial robots by human demonstration , 2002, IEEE/RSJ International Conference on Intelligent Robots and Systems.

[3]  Kunikatsu Takase,et al.  Skill Based Manipulation System , 1990 .

[4]  Katsushi Ikeuchi,et al.  Sensor Modeling, Probabilistic Hypothesis Generation, and Robust Localization for Object Recognition , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  H. Hirukawa,et al.  Automatic determination of possible velocity and applicable force of frictionless objects in contact from a geometric model , 1994, IEEE Trans. Robotics Autom..

[6]  Katsushi Ikeuchi,et al.  Correcting observation errors for assembly task recognition , 2002, IEEE/RSJ International Conference on Intelligent Robots and Systems.

[7]  Stefan Schaal,et al.  Is imitation learning the route to humanoid robots? , 1999, Trends in Cognitive Sciences.

[8]  Jing Xiao,et al.  Toward obtaining all possible contacts-growing a polyhedron by its location uncertainty , 1996, IEEE Trans. Robotics Autom..

[9]  Feng Pan,et al.  Efficient contact state graph generation for assembly applications , 2003, 2003 IEEE International Conference on Robotics and Automation (Cat. No.03CH37422).

[10]  Bernard Roth,et al.  An Extension of Screw Theory , 1981 .

[11]  Yasuhisa Hasegawa,et al.  Direct teaching and error recovery method for assembly task based on a transition process of a constraint condition , 2001, Proceedings 2001 ICRA. IEEE International Conference on Robotics and Automation (Cat. No.01CH37164).

[12]  Katsushi Ikeuchi,et al.  Towards An Assembly Plan From Observation: Part II: Correction Of Motion parameters Based On Fact Contact Constraints , 1992, Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems.

[13]  Russell H. Taylor,et al.  Automatic Synthesis of Fine-Motion Strategies for Robots , 1984 .

[14]  Jing Xiao,et al.  A divide-and-merge approach to automatic generation of contact states and planning of contact motion , 2000, Proceedings 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings (Cat. No.00CH37065).

[15]  Katsushi Ikeuchi,et al.  Task-Oriented Generation of Visual Sensing Strategies in Assembly Tasks , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[16]  Bruce Randall Donald,et al.  A Search Algorithm for Motion Planning with Six Degrees of Freedom , 1987, Artif. Intell..

[17]  A. Tucker,et al.  Linear Inequalities And Related Systems , 1956 .

[18]  Michael M. Marefat,et al.  Determining feasible contact states of pairs of spatial polyhedra , 2000, Proceedings 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings (Cat. No.00CH37065).

[19]  Masayuki Inaba,et al.  Learning by watching: extracting reusable task knowledge from visual observation of human performance , 1994, IEEE Trans. Robotics Autom..

[20]  Katsushi Ikeuchi,et al.  Toward an assembly plan from observation. I. Task recognition with polyhedral objects , 1994, IEEE Trans. Robotics Autom..

[21]  Jing Xiao,et al.  Contact constraint analysis and determination of geometrically valid contact formations from possible contact primitives , 1997, IEEE Trans. Robotics Autom..