מושגי ליבה
iRoCo optimizes robot control through smartwatch and smartphone for intuitive human-robot collaboration.
תקציר
I. Introduction
iRoCo framework introduced for human-robot collaboration.
Challenges with traditional motion capture systems.
Potential of smart devices for motion capture explored.
II. Related Work
Utilization of smart devices in enhancing human-robot interactions.
Wearable devices to reduce user fatigue in robot control tasks.
Proposal to utilize differentiable filters for accurate arm pose estimation.
III. Methodology
A. Data Collection, Observation, and State
Data collection setup using smartwatch, smartphone, and Optitrack system.
Definition of observations and states based on sensor data.
B. Differentiable Ensemble Kalman Filter
Description of DEnKF model structure for state estimation.
C. Control Modality
Introduction of a control modality leveraging human pose estimations for robot control.
IV. Evaluation
A. Training and Test Datasets
Collection methodology approved by IRB at ASU.
B. Model Performance
Evaluation metrics used to assess the performance of the DEnKF model.
C. Results Comparison with Related Works
V. Real-Robot Applications
A. Teleoperation Task Setup and Procedure
B. Drone Piloting Task Setup and Procedure
C. Results Comparison between Remote Control and iRoCo System
VI. Conclusion
סטטיסטיקה
iRoCoは、ドローン操縦タスクで平均して19秒早く完了しました。
Optitrackを使用したピックアンドプレースタスクでは、平均して13秒遅く完了しました。
ציטוטים
"Studies have specifically examined the use of a single smartwatch for motion capture."
"Our findings strongly suggest that iRoCo is a promising new approach for intuitive robot control."