Deep machine learning methods and apparatus for robotic grasping

This invention describes a robot that attempts to pick up an object. After the robot's gripper closes, it moves the gripper slightly away while still holding the object. The robot then compares images taken before and after the move, using the difference in pixels to determine if the grasp was successful.

Why it matters: Since 2021, deep learning models for robotic manipulation have become significantly more capable and efficient, making the training and real-time servoing described in the abstract more practical. While the claims detail a basic pixel-difference success check, modern computer vision could enhance this verification.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedRobot hardware, deep learning software, image processing.

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10