Redundancy problem in writing: from human to anthropomorphic robot arm

Area of Science:

• Robotics
• Biomechanics
• Control Systems

Background:

• Redundant anthropomorphic arms, like the human arm, exhibit complex motion patterns during tasks such as writing.
• Existing robotic control methods often struggle with the inverse kinematics of redundant mechanisms.
• Biological systems demonstrate inherent synergies, where proximal joint movements dictate distal joint actions.

Purpose of the Study:

• To analyze and model the motion of a redundant anthropomorphic arm during writing.
• To introduce and validate the concept of distributed positioning (DP) as a control strategy.
• To demonstrate the potential for improving robot control by mimicking biological systems.

Main Methods:

• Decomposing prescribed arm movement into smooth global and fast local motions (distributed positioning).
• Distributing these motions to individual arm joints based on inertial properties and actuation capabilities.
• Applying the developed model to an anthropomorphic robot arm performing a writing task.

Main Results:

• The proposed distributed positioning approach provides a unique solution for the inverse kinematics of redundant mechanisms.
• The model successfully captures the synergistic control observed in human writing.
• Demonstrated that a biological analog control schema can enhance robot arm capabilities.

Conclusions:

• Distributed positioning is an effective method for modeling and controlling redundant anthropomorphic arms.
• Mimicking biological synergies offers a promising pathway for advancing robotic control.
• This research provides a foundation for developing more capable and human-like robots.