2018 Fiscal Year Annual Research Report
Interact with me: optimal human-humanoid physical interaction
| Project/Area Number |
18F18754
|
|---|
| Research Institution | National Institute of Advanced Industrial Science and Technology |
|---|
Principal Investigator |
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)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
HU YUE 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, 外国人特別研究員
|
|---|
| Project Period (FY) |
2018-11-09 – 2021-03-31
|
| Keywords | ヒューマン・ロボット・インタラクション / 最適化 / 人間動作解析 |
|---|
| Outline of Annual Research Achievements |
In FY2018 we designed and tested an experiment for measuring and identifying interaction factors, during an active physical human-robot interaction.
In order to have a higher control over the experiment and easier validation for later stages, we decided to use the Sawyer robot from Rethink. In the experiment the human is attached to the robot physically and uses it as joystick for playing a visual game, the robot can take an action on the human without prior notice, in order to evaluate the state and reaction of the person during such active interactions. During this experiment, the human is equipped with motion capture system, eye tracking, Galvanic Skin response sensor and heart beat monitor. From these measuring devices we can infer the interaction factors, those we are considering consist of (but not limited to): heart rate variations, emotional stress, frequency and timing of gazing at the robot, exchanged forces between the human and the robot, distance between human and robot, velocity and acceleration of the human motions. These factors will be used for a new optimal control architecture that will be designed in FY2019.
In order to guarantee safety of the user, we use a Quadratic Programming (QP) framework for the control of the robot, which allows to impose constraints, the robot is torque controlled during the interaction and the active force is bounded. The experiment has been approved by the ethical committee and from initial tests we consider it to be suitable to achieve the desired goal of measuring interaction factors.
|
| 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 current status is slightly behind schedule due to some sensors that were not available in FY2018, which we need to measure all interaction factors. Given that we need to perform extensive experiments in order to obtain statistically significant data to infer useful interaction factors, we would need to have all the sensor available before starting experiments with external participants. However, the setup is ready and the sensors will be available soon so the project will proceed as expected.
|
| Strategy for Future Research Activity |
In FY2019 we will perform extensive experiments with external participants, in order to collect data for the measurement of the interaction factors. We will evaluate the actual contribution of these factors and we also expect new factors to arise from the measurements. A new optimal control architecture that takes into account these factors will be implemented, then new experiments will be carried out in order to validate the proposed control architecture. We also expect to implement the new control architecture on other robots available in the laboratory for validation purposes.
|
