Winching up heavy loads with a compliant arm: a new local joint controller

Area of Science:

• Robotics
• Biomimetic Control
• Mechanical Engineering

Background:

• Constrained movement tasks, common in robotics, often rely on complex central controllers and force-torque sensors.
• Existing methods for contour following in compliant manipulators can be limited, especially when facing counteracting forces.

Purpose of the Study:

• To introduce and evaluate a new decentralized control system, power-controlled relaxation LPVF, for compliant manipulators performing constrained movement tasks.
• To demonstrate the system's ability to overcome significant opposing forces during contour following.

Main Methods:

• Developed a control system inspired by stick insect locomotion, utilizing decentralized joint controllers with local positive velocity feedback (LPVF).
• Extended LPVF by incorporating local mechanical power measurement to create power-controlled relaxation LPVF.
• Tested the system on a two-joint compliant manipulator performing a crank-turning task to winch a weight.

Main Results:

• The power-controlled relaxation LPVF successfully enabled contour following in a constrained crank-turning task.
• The system demonstrated robustness against considerable counter forces dragging the manipulator along the contour.
• The control approach eliminated the need for kinematic knowledge or end-point force/torque sensing.

Conclusions:

• Power-controlled relaxation LPVF offers an effective, decentralized solution for compliant manipulators in constrained tasks.
• This bio-inspired approach simplifies robotic control by removing the need for complex sensing and central coordination.
• The method shows promise for applications requiring precise and robust manipulation in challenging environments.