2015 Fiscal Year Research-status Report
人工手指を自分の手指のように動かす：ヒト脳活動を用いた人工手指の自然な学習
| Project/Area Number |
15K12623
|
|---|
| Research Institution | National Institute of Advanced Industrial Science and Technology |
|---|
Principal Investigator |
GOWRISHANK AR.G 国立研究開発法人産業技術総合研究所, 知能システム研究部門, 国際客員研究員 (10570244)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
宮脇 陽一 電気通信大学, 先端領域教育研究センター, 准教授 (80373372)
|
|---|
| Project Period (FY) |
2015-04-01 – 2017-03-31
|
| Keywords | Brain machine interface / embodiment / machine learning / natural learning |
|---|
| Outline of Annual Research Achievements |
In this year we designed experiments to investigate the free neural activity (FNA) in the brain and collected data from 3 subjects with a visual task for now. We worked on the algorithm to isolate the FNA. This work is still progressing.
A critical issue for our project is the ability of the subjects to embody the robot finger. Extensive experiments were conducted with subjects to understand sensory feedback constraints on embodiment and ownership when a robot limb is ‘attached’ to one’s body. These experiments were published in 2 conferences and two journals, while one more journal article is now in review.
I have also started working on the design of the finger and feedback device which should be made in the next year.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
The project is progressing very smoothly. The robot finger and feedback device have not yet been developed as initially planned, but we were able instead do extensive experiments on embodiment, that were initially planned for the second year. The FNA algorithm is a very challenging issue for the project, but this algorithm development is going smoothly
|
| Strategy for Future Research Activity |
Last year we had investigated embodiment of robot limbs, but with whole hands. In this year, In collaboration with the LIRMM lab in France in June, I plan to extend these experiment to a robot finger, which is the main goal of the project. The design of the artificial finger and feedback device will be completed and they will be developed using 3d printing by September. The FNA algorithm needs to be finished and then integrated with the artificial finger. With these we plan to start user experiments by December next year to evaluate our hypothesis.
|
| Causes of Carryover |
Money was left in the first year because
1) The robot finger and feedback device are not yet developed. Instead we I spent more time on the experiments
2) Due to lack of time the money for conference travels were not used.
|
| Expenditure Plan for Carryover Budget |
Next year, most of the money that is forwarded will be used for hardware development (of the robot finger and feedback device) and partly for the visit to LIRMM in France, where we plan a new experiment using a robot finger/hand system
|
