Paper:
Views over last 60 days: 324
Sensor-Based Artificial Skills with Force and Vision Information
Kosei Kitagaki*, Takashi Suehiro* and Motoyoshi Fujiwara**
*National Institute of Advanced Industrial Science and Technology (AIST), 2-1 Namikicho 1-chome Tsukuba Shi Ibaraki Prefecturure 305-8564, Japan
**Mie Prefectual Industrial Research Institute, 5-5-45 Takachaya, Tsu-shi, Mie Japan
Received:January 1, 2001Accepted:January 20, 2001Published:June 20, 2001
Keywords:manipulation skill, pseudo contact point monitoring, contact state recognition, feature-oriented coordination, calibration-free vision
Abstract
To add flexibility against changes in assembly processes to a manufacturing system, we introduce artificial skills with advanced sensor-based information. Tasks are divided into a phase in local (contact) and global (noncontact) states and we developed pseudo contact point monitoring and feature-oriented coordination to achieve artificial skills for tasks in each phase. Pseudo contact point monitoring is a new technique that detects contact states by monitoring a pseudo contact point calculated from a set of observed force and moment. Effectiveness of artificial skills is shown by experimental results of stamping, bearing roller positioning, and a peg-in-hole task. Feature-oriented coordination system is a new control method that does not require prior calibration between the vision system and manipulator. We detail system possibilities and limitations with simulation.
Cite this article as:K. Kitagaki, T. Suehiro, and M. Fujiwara, “Sensor-Based Artificial Skills with Force and Vision Information,” J. Robot. Mechatron., Vol.13 No.3, pp. 254-266, 2001.Data files:
Copyright© 2001 by Fuji Technology Press Ltd. and Japan Society of Mechanical Engineers. All right reserved.
