次世代協働ロボット：行動神経学に基づく「安心できる」ロボットの動きの解明

2018 Fiscal Year Annual Research Report

• Project/Area Number: 17H01804
• Research Institution: National Institute of Advanced Industrial Science and Technology
• Principal Investigator: 山野辺 夏樹 国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 主任研究員 (90455436)
• Co-Investigator(Kenkyū-buntansha): GOWRISHANK AR.G 国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 客員研究員 (10570244) ; 吉田 英一 国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 研究部門付 (30358329) ; Venture Gentiane 国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 特定フェロー (30538278)
• Project Period (FY): 2017-04-01 – 2020-03-31
• Keywords: 協働ロボット / 行動神経学 / 動作計画

Outline of Annual Research Achievements

For object handover experiments, which is one of human-human collaboration tasks, we increased the variation of the test subjects and verified our hypothesis, which was observed last year, that humans can predict the partner's handover behavior and there are some important variables including physical and social features of the pairs, which effect human motion modification. A paper was submitted to the journal, PLOS ONE (https://journals.plos.org/plosone/), based on the extended experiments and is the final stage of reviewing. In addition to human-human tasks, we conducted human-robot experiments for object handover to investigate how humans predict partner’s behavior when the partner is a robot. The tendency of prediction was slightly different with the human-human scene.
For understanding human-robot collaboration in co-worker scenario, we also conducted human-human experiments in which subjects’ arms cross each other. Based on the preliminary experiments and the results obtained by the handover investigation, which is that human perception becomes slightly different with a robot partner, we redesigned experimental setup of simple human-robot collaboration and conducted experiments with several test subjects. We are now analyzing the data for defining evaluation functions related to human emotion in co-working situation.

Current Status of Research Progress

2: Research has progressed on the whole more than it was originally planned.

Reason

The data analysis for defining peri-limb space, which was the target for FY2018, is still ongoing, but we were able to submit a journal paper related to handover experiments. Moreover, we adopted a new idea for the evaluation functions of human emotion, which would be an additional result to the peri-limb space. The project seems to be going well.

Strategy for Future Research Activity

As we submitted for FY2019 research plan, we will continue the experiments of human-robot collaboration for defining evaluation functions related to human emotion in co-working situation. Based on such definitions and additional human-robot collaboration experiments, we will then model the peri-limb space, which is our project goal.