A Dexterous Reorientation Strategy for Precision Picking of Large Thin Objects.

Area of Science:

• Robotics
• Artificial Intelligence
• Mechanical Engineering

Background:

• Handling large, thin objects presents challenges for robotic systems due to their low profile and tendency to adhere to surfaces.
• Existing robotic grasping methods often struggle with objects resting on flat surfaces, requiring complex maneuvers or specialized tooling.

Purpose of the Study:

• To present and analyze the tilt-and-pivot manipulation technique for robotic object acquisition.
• To demonstrate the effectiveness of this method for picking thin, low-profile objects from supporting surfaces.

Main Methods:

• The tilt-and-pivot technique involves reorienting the robotic gripper around the object's contact point.
• This allows a robotic finger to access the gap between the object and the surface without sliding.
• The method facilitates reliable pinch grasps on the object's faces, supported by kinematic principles and planning strategies.

Main Results:

• Experimental results demonstrate the effectiveness of tilt-and-pivot manipulation in bin-picking scenarios.
• The technique successfully enabled reliable pinch grasps on various low-profile objects.
• Integration of vision and force-torque sensing enhanced the manipulation's success rate.

Conclusions:

• Tilt-and-pivot manipulation offers a practical and effective solution for robotic grasping of thin objects.
• The technique shows promise for broader applications in complex object-handling tasks.
• This robotic manipulation strategy enhances dexterity and reliability in object acquisition.