Exciting new work from the R-Pad Lab’s Wenxuan Zhou and David Held features a simple gripper that can solve more complex manipulation tasks if it can utilize the external environment such as pushing the object against the table or a vertical wall, known as “Extrinsic Dexterity.”
From the researchers:
A simple gripper can solve more complex manipulation tasks if it can utilize the external environment such as pushing the object against the table or a vertical wall, known as “Extrinsic Dexterity.” We present a system for the “Occluded Grasping” task with model-free RL which shows emergent extrinsic dexterity behaviors to solve the task. The learned policy demonstrates dynamic and contact-rich motions with a simple gripper that generalizes across objects with various size, density, surface friction, and shape with a 78% success rate.
Read the paper on the project website.