Toward Damage-Less Robotic Fragile Fruit Grasping: A Closed-Loop Force Control Method for Pneumatic-Driven Soft...

Area of Science:

• Robotics
• Agricultural Engineering
• Control Systems

Background:

• Fragile fruit handling in postharvest processes is labor-intensive and prone to damage.
• Aging global populations necessitate automation, highlighting the need for robotic solutions.
• Controlling grasping force with pneumatic soft grippers is complex due to system nonlinearities and time delays.

Purpose of the Study:

• To develop a robust robotic grasping force control system for pneumatic grippers handling fragile fruits.
• To address challenges in damage-less grasping and ensure stable manipulation.
• To improve automation efficiency in the postharvest industry.

Main Methods:

• Development of a robotic grasping system integrating a pneumatic soft gripper and a matrix force sensor.
• Implementation of a double closed-loop control strategy using a Kalman filter (KF) for denoising and a proportion-integration-differentiation (PID) controller with a dead band.
• Exploration of force distribution for varying contact points and utilization of air pressure for force regulation.

Main Results:

• Achieved robust closed-loop grasping force control with a maximum steady-state error of 0.07 N.
• Validated the effectiveness of the Kalman filter for sensor data denoising and matrix force visualization for grasping mechanism analysis.
• Demonstrated the efficacy of multiple sensing units and the dead band in controller performance through ablation studies.
• Confirmed generalization performance and anti-disturbance capabilities of the developed system.

Conclusions:

• The developed double closed-loop control system effectively solves the problem of closed-loop grasping force control for pneumatic grippers.
• The proposed method offers a potential solution for automating fragile fruit handling in the postharvest industry, reducing damage and labor costs.