The purpose of this research is to develop robot hand system which interacts with human hand motion. Develop tele-operation system which robot hand replicates human hand motion, and feedback algorithm between the robot and the human are the two major goals of this research.
Build an interface that robot hand resembles human hand motion, and develop 3D modeling algorithm to find unknown object's characteristics with sensors and active sensing mechanism in order to develop safe tele-operation technique. Moreover, in order to accomplish robust task for robot hand, developing grasping force optimization algorithm, and transition algorithm for avoiding singularity while switching position/force control are required. Therefore, applying these technologies we build a system for human hand and remote object interaction for real-time robot hand tele-operation control framework.
Develop the technology to improve realistic haptic feedback between human and robot in tele-operation control for remote object
1. Develop mapping algorithm between robot hand and human hand
Human replicates robot hand motion and collects human hand motion data in order to improve the corelation. By stacking these data we build a database for more precise similarity between robot hand and human hand motion.
2. Develop haptic glove module for recognizing human hand motion and grasping force
Through haptic glove we develop a module that recognize the operator's grasping force and direction using EMG signal pattern generation algorithm. After analysing the EMG pattern we develop position/force simulator.
3. Unknown object modeling using robot hand's force reflection sensing
Develop 3D modeling of unknown object by using vision and touch sensor. Moreover, adding F/T sensor we suggest force reflection sensing and object modeling technique using sensor fusion algorithm.
4. Develop force/position transition algorithm
Develop stability guarantee guide algorithm to prevent rapid torque fluctuation during control mode switch, and prevent unstability while robot is in contact with an object.
5. Develop grasping force distribution algorithm for stable grasping
Estimate unknown object's stiffness , mass change, and develop grasping force distribution algorithm in contact state. Also develop safety guarantee guiding technology during operation.