2019 Fiscal Year Final Research Report
A Study for Learning of Robotic Tool Use by Logging of Human Motions
| Project/Area Number |
17K12761
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Intelligent robotics
|
|---|
| Research Institution | Chukyo University (2019)
Keio University (2017-2018) |
|---|
Principal Investigator |
Shuichi Akizuki 中京大学, 工学部, 助教(テニュア) (40796182)
|
|---|
| Project Period (FY) |
2017-04-01 – 2020-03-31
|
| Keywords | 機能認識 / 日用品理解 / ピッキング / 把持 / 動作ロギング |
|---|
| Outline of Final Research Achievements |
In this research, for the purpose of realizing the manipulation of daily tools (such as kitchen or DIY utensils) by robotic arms, we have purposed an instruction manual of them with 3D shape, which describes the usage method. By analyzing the human demonstrations, tactile histories that occurred on the tools are accumulated to the tool’s surface as the "Tactile Log", which can help to achieve the plausible tool use by robot arms. Since the tactile history is logged directly to the shape of the object,
it is possible to recognize the correct usage of the daily tools only by recognizing them from the environment. We have also developed a method for 3D pose estimation algorithm required for the tactile logging, and a robotic picking system.
|
| Free Research Field |
知能ロボティクス
|
| Academic Significance and Societal Importance of the Research Achievements |
生活支援ロボットには，さまざまな日用品を扱うための汎用性が求められる．日用品の操作を実現するためには，長い時間をかけて一品一様に手作業での動作教示が必要であった．本研究では，人間が日用品を使用する様子を記録し，その使用履歴を物体形状自体に書き込む，いわば「形を持った取扱説明書」を提案した．このデータをロボットが参照することによって，人間のような日用品の取り扱い動作を実行することができる．
|
