Robot Grasp Teaching from Human Demonstration and Pose Optimization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing robot grasp teaching techniques are inefficient and costly due to manual teaching methods, and automated techniques often result in low-quality grasps that fail to account for environmental factors and specific grasping requirements.

Innovation Solution

A method for robotic grasp teaching by human demonstration, where a human demonstrates a grasp on a workpiece, and images are analyzed to determine a hand pose and grasp region. Grasp optimization is employed to generate multiple stable, high-quality grasps, which are then selected and added to a grasp database.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual grasp teaching is used, then grasp quality can be high, but teaching time and cost are excessive

Engineering Contradiction:

Improvegrasp qualityVSAvoidteaching time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system captures images of human hand poses during demonstration and uses these visual copies to train the vision system and generate grasp poses, eliminating the need for time-consuming manual teaching while preserving high grasp quality through human expertise

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent replaces manual mechanical teaching operations with an automated vision-based system that captures images, processes hand pose data, and generates grasp poses algorithmically, significantly reducing teaching time while maintaining reliability

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If automated grasp generation is used, then teaching speed is fast, but grasp quality is low

Engineering Contradiction:

Improveteaching speedVSAvoidgrasp quality

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The vision system acts as an intermediary that captures human demonstration images and translates them into robot grasp poses, combining the speed of automated processing with the quality of human expertise to achieve both fast teaching and high grasp quality

Inventive Principle:

Principle #24Intermediary (Mediator)

3Productivity

If existing automated techniques are used, then teaching is efficient, but environmental factors are not accounted for

Engineering Contradiction:

Improveteaching efficiencyVSAvoidenvironmental factor accounting

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The system determines specific grasp regions on the workpiece by analyzing hand pose images and identifying contact points, allowing adaptation to local environmental factors and specific workpiece geometries while maintaining teaching efficiency through automated processing

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS12304070B2Grasp teach by human demonstration

Publication Date: 2025.05.20 FANUC LTD

AI summary

A technique for robotic grasp teaching by human demonstration. A human demonstrates a grasp on a workpiece, while a camera provides images of the demonstration which are analyzed to identify a hand pose relative to the workpiece. The hand pose is converted to a plane representing two fingers of a gripper. The hand plane is used to determine a grasp region on the workpiece which corresponds to the human demonstration. The grasp region and the hand pose are used in an optimization computation which is run repeatedly with randomization to generate multiple grasps approximating the demonstration, where each of the optimized grasps is a stable, high quality grasp with gripper-workpiece surface contact. A best one of the generated grasps is then selected and added to a grasp database. The human demonstration may be repeated on different locations of the workpiece to provide multiple different grasps in the database.