Robot Hand Grasp Generation with Physics-Based Stability Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing grasp planning technologies for robotic hands struggle to efficiently generate stable grasps for diverse objects, particularly due to the high-dimensional joint space of robotic hands, which complicates the search for stable grasp postures.
Innovation Solution
A method and system for generating feasible grasps by searching within a configuration space of a robot hand model, simulating grasping in a physics engine, applying wrench disturbances, and assigning grasp stability scores to identify stable grasps, utilizing eigengrasp spaces to reduce dimensionality and enhance grasp planning efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the robot hand model searches within the full configuration space to generate candidate grasps, then the grasp coverage and versatility are improved, but the computational complexity and time consumption increase significantly
Solution Approach 1:
The configuration space is segmented into multiple local configuration spaces, each corresponding to a specific grasp type (e.g., power grasp, precision grasp, lateral grasp). By dividing the search space into manageable segments, the system can efficiently explore different grasp categories without being overwhelmed by the full configuration space complexity, thus maintaining high grasp coverage while reducing computational burden.
Solution Approach 2:
The patent introduces a grasp type dimension to organize and constrain the configuration space search. Instead of searching the entire high-dimensional configuration space uniformly, the system adds a categorical dimension that groups configurations by grasp type, enabling more efficient navigation through the configuration space while ensuring diverse grasp coverage.
2Reliability
If multiple simulations with wrench disturbances are applied to evaluate grasp stability, then the reliability of grasp selection is improved, but the computational time and processing resources increase
Solution Approach 1:
The system performs preliminary filtering of candidate grasps using geometric and kinematic constraints before applying computationally intensive wrench disturbance simulations. By pre-screening grasps that satisfy basic stability criteria, the system reduces the number of simulations needed, thereby maintaining high reliability in grasp selection while minimizing computational time consumption.
Solution Approach 2:
The patent applies wrench disturbances selectively to a subset of the most promising candidate grasps rather than all candidates. This partial action approach focuses computational resources on evaluating only those grasps that have high potential stability, achieving reliable grasp selection without the excessive time cost of simulating every possible grasp.
3Adaptability or versatility
If the robot hand model engages diverse objects with multiple grasp types, then the adaptability to different objects is improved, but the difficulty of detecting and measuring stable grasp postures increases
Solution Approach 1:
The patent introduces grasp templates as intermediary structures that encode stable grasp postures for different object types and grasp categories. These templates serve as reference models that simplify the detection and measurement process by providing pre-defined geometric and kinematic constraints, making it easier to identify stable grasps across diverse objects without directly analyzing the full complexity of each configuration.
Data Source
AI summary
A method of grasp generation for a robot includes searching within a configuration space of a robot hand model for robot hand configurations to engage an object model with a grasp type. The method includes generating a set of candidate grasps based on the robot hand configurations. Grasping of the object model with the robot hand model is simulated in a physics engine using simulated grasps generated based on a given candidate grasp. A simulated grasp is assigned a score based on a response of the object model to an applied wrench disturbance when the object is engaged with the simulated grasp. The method includes generating a set of feasible grasps for the given candidate grasp based on the respective simulated grasps having a score above a score threshold at a target wrench disturbance.