ideaflow

Patent overview

Unique aspects of the patent

Traditional issues

Solution

Patent features

Applicant / Rights holder

Inventors

Ideas using this intellectual property

Adaptive Gripping Planning Technology for Robots

US20210308866A1

Fanuc Corp

Patent overview

This patent provides adaptive gripping and motion planning technology for industrial robots to efficiently grip parts arranged randomly and move them to designated positions. By analyzing the shape of the parts and identifying multiple robust gripping options and stable intermediate postures, it selects the optimal gripping method. The robot uses cameras and sensors to identify the position of the parts and develop a motion plan. This optimizes the robot's actions and achieves high processing capability while avoiding collisions.

Unique aspects of the patent

Identification of gripping methods based on part shapes
Motion planning using intermediate postures
Position identification using cameras and sensors

Traditional issues

Conventional industrial robots found it difficult to efficiently grip parts that were randomly arranged and move them to designated positions. Especially when parts are arranged in random positions and postures, it is necessary to calculate a unique motion path for each part, requiring efficient operations. Additionally, there are often obstacles in the work environment that affect the robot's motion path, necessitating motion planning to avoid these obstacles. Moreover, it was challenging to find the optimal gripping method depending on the shape and arrangement of the parts, necessitating the optimization of the robot's actions.

Solution

This patent solves conventional challenges by analyzing the shape of parts to identify multiple robust gripping options and stable intermediate postures. The robot uses cameras and sensors to identify the position of the parts and develop a motion plan. If no feasible grip is available, it selects an intermediate posture and plans to move the parts from the initial posture to the intermediate posture and then to the target posture. This optimizes the robot's actions and achieves high processing capability while avoiding collisions.