Robot Gripping Space Determination for Stable Collision-Free Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining gripping poses for robots are complex, require precise positioning, and do not account for multiple stable grip options, making them inefficient in dynamic environments and prone to collisions.
Innovation Solution
A method that defines gripping spaces and poses using an object's coordinate system, allowing for intuitive and easy determination of gripping areas by manually positioning the gripper and calculating additional poses based on structural data, enabling collision-free and stable gripping independent of the object's absolute position.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pre-programmed gripping poses are used, then the robot can execute stable grips, but the object must always be in the same position and orientation, reducing adaptability
Solution Approach 1:
The patent applies dynamics by transforming static pre-programmed gripping poses into dynamic gripping areas. Instead of fixed positions, the system defines spatial regions where multiple gripping poses are valid. The robot can dynamically select any pose within these areas based on real-time object position, maintaining grip stability while adapting to position variations.
Solution Approach 2:
The patent changes the parameter representation from discrete fixed coordinates to continuous spatial regions with multiple valid poses. By defining gripping areas as volumetric spaces containing numerous valid gripping configurations, the system allows continuous adaptation to object position changes while ensuring each selected pose maintains stable grip characteristics.
2Reliability
If individual gripping points are taught manually, then stable grips can be achieved, but the method is complex and time-consuming
Solution Approach 1:
The patent segments the complex task of determining individual gripping points by defining gripping areas as composite regions containing multiple valid poses. Instead of teaching each point separately, the system creates spatial zones that automatically encompass all valid gripping configurations, simplifying the programming process while maintaining comprehensive coverage.
Solution Approach 2:
The patent uses object structural data (CAD models or point clouds) to automatically generate and copy valid gripping poses throughout the defined gripping areas. Rather than manually teaching each pose, the system replicates valid gripping configurations based on the object's geometric features, significantly reducing programming complexity while ensuring physical feasibility.
3Object-affected harmful factors
If fixed gripping poses are used, then collision-free operation can be ensured, but the robot cannot accommodate multiple stable grip options
Solution Approach 1:
The patent transitions from zero-dimensional fixed points to three-dimensional volumetric gripping areas. This dimensional expansion allows the system to incorporate multiple stable grip options within each spatial region while maintaining collision avoidance through predefined boundary constraints. The added spatial dimensions enable both versatility and safety simultaneously.
4Reliability
If precise positioning is required for gripping, then stable object handling is achieved, but mobile robots with insufficient positioning accuracy cannot execute fixed grips
Solution Approach 1:
The patent applies dynamics by replacing fixed positioning requirements with dynamic gripping areas that accommodate positioning variations. Mobile robots can operate within these flexible spatial regions, selecting valid poses that maintain stable object handling despite lower positioning accuracy, while still ensuring collision-free operation through the predefined area boundaries.
Data Source
Figure 1a~1b
Figure 2a~2b
Figure 3
AI summary
The invention relates to a method and a system for determining gripping spaces on an object. The object is to be gripped by a robot using the specified gripping spaces. At least a first gripping pose of the robot on the object is taught and further gripping poses on the object are determined. A first gripping space is determined based on these gripping poses.