2021 Fiscal Year Research-status Report
Design and Geometric Motion Planning Studies of a Novel Agile 3-DoF Fingertip Grasping Mechanism
| Project/Area Number |
21K20391
|
|---|
| Research Institution | Tohoku University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2021-08-30 – 2023-03-31
|
| Keywords | Geometric mechanics / Differential geometry / Motion planning / Control / Agile fingertip / Dexterous manipulation / Design / Robots |
|---|
| Outline of Annual Research Achievements |
We studied the motion planning problem of a spin-rolling sphere as our fingertip model. We found a new differential geometry-based transformation for the ball-plate kinematics. The strategy transformed underactuated ball-plate kinematics into fully-actuated one. Based on the newly transformed kinematics, we designed a motion planning strategy for a spin-rolling sphere (fingertip) on a plane surface. Also, the new framework was used in obtaining a new assistive geometric controller that was tested in mobile robots. Although the Coronavirus pandemic caused certain delays, we are in the progress of finishing the first prototype of a 3-DoF spin-rolling fingertip. Also, we are researching on expanding the motion planning strategy for a multi-finger grasping with arbitrary surfaces.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
We published our results in well-known journals and international conferences based on our successful development of differential geometry-based techniques. We are now working on the first prototype of a spin-rolling fingertip. Because of the Coronavirus pandemic, we had some issues in getting the required items at the proper time. However, we are aiming to verify our findings in an experiment by considering effort division and correlation of the research with the ongoing Moonshot Project. Moreover, we have achieved relevant research outputs in the mobile robots platform. Recently, We are testing and improving the performance of the fingertip mechanism. Also, we are working on developing a general solution for motion planning by using our proposed differential geometry-based technique.
|
| Strategy for Future Research Activity |
The following goals are planned in a recursive manner: Extending the motion planning strategy based on Darboux-frame to arbitrary surfaces, Developing the first prototype of 3-Dof spin-rolling fingertip, Verifying the effectiveness of the grasping and agility of the prototype, Extending the setup to a multi-finger model, Applying object manipulation algorithms to have a feed-forward control on different objects, Develop a visual servo technique to pose estimate and control the grasped objects using the depth camera, Experiment the motion planning strategy based on differential geometry with the new prototype.
|
| Causes of Carryover |
We plan to use the amount in several things. First, we plan to purchase motors for the finger body to increase the degrees of freedom. Also, we are planning to purchase the required sensory systems e.g., a motion capture camera to get accurate tracking of the manipulated object. Moreover, we might purchase a manipulator to mount the new agile fingertip grasping mechanism to the robot tool-tip for pick-and-place tasks. We will research manipulation of different objects to check the performance of the developed mechanism.
|
