Dexterous object manipulation by combining sensory feedback and proprioceptive information(Fostering Joint International Research)
| Project/Area Number |
15KK0013
|
|---|
| Research Category |
Fund for the Promotion of Joint International Research (Fostering Joint International Research)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Intelligent robotics
|
|---|
| Research Institution | Kyushu University |
|---|
Principal Investigator |
Tahara Kenji 九州大学, 工学研究院, 准教授 (80392033)
|
|---|
| Research Collaborator |
Kao Imin The State University of New York at Stony Brook, Department of Mechanical Engineering, Professor, Executive Director of SUNY Korea
|
|---|
| Project Period (FY) |
2016 – 2018
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥10,790,000 (Direct Cost: ¥8,300,000、Indirect Cost: ¥2,490,000)
|
|---|
| Keywords | 柔軟センサ / 触覚 / 多指ハンド / 柔軟指 / 転がり拘束 / カメラ / 楕円体 / Friction Limit Surface / 触覚センサ / 多指ロボットハンド / 透明シリコン |
|---|
| Outline of Final Research Achievements |
Until now, modeling of a soft finger has been limited to a sphere, and there has been no discussion of shape. Therefore, in order to extend it to general shapes, we extend it to an ellipsoidal model. We mapped the ellipsoid onto the xy-plane and derived the contact center position by finding the geometric center of the ellipse for the mapping. In addition, the displacement direction was taken from the center position of the contact surface in the perpendicular direction, and the model was constructed so as to generate reaction force according to the volume of the deformed ellipsoid. We proposed an algorithm to estimate contact force and position using a prototype. The inside was hollow and beads were placed inside. In addition, the fingertip force component is decomposed and derived as a rigid ellipsoid in the object coordinate system in a form that includes constraint conditions, so that the target rigid ellipsoid can be set at the object center.
|
| Academic Significance and Societal Importance of the Research Achievements |
本研究では，柔軟な指先センサの構築を目的として，柔軟モデルの導出を行い，そのモデルをベースとして，非接触センサ（カメラ）を用いた接触センサの試作を行い，それを用いた物体把持アルゴリズムの提案を行った．
これまで，指先センサは繊細で高価な物が多く，また，検出された位置や力はノイズや誤差を多分に含むため，どのように物体把持に利用するかについて明示されていなかった．
本研究では，安価で簡易な接触センサを提案し，また，それによって検出されるノイズや誤差を含む位置や力情報を利用した物体把持アルゴリズムを提案することにより，多指ハンドによる物体把持をより産業において簡便に利用できるようにした．
|
Report
(4 results)
Research Products
(1 results)
