Robot Grasp Target Selection Using Visibility Estimation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional robot systems face challenges in accurately selecting and grasping work targets due to visibility issues, such as partial occlusion and positioning within the camera's field of view, leading to increased failure rates in holding tasks.

Innovation Solution

A robot control system that includes an estimator to calculate the visibility of work targets based on image analysis, using a trained model to generate masks and estimate visibility, and a work target specifier to select the most visible and easily graspable targets for the robot.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the robot selects work targets based on simple image recognition, then the selection process is fast, but the accuracy of target selection deteriorates due to occlusion and positioning issues

Engineering Contradiction:

Improvetarget selection accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary visibility estimation and scoring of work targets before the robot attempts grasping. The estimator calculates visibility metrics for each detected target, and the work target specifier selects targets with optimal visibility scores, preventing failed grasping attempts before they occur.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The visibility estimator acts as an intermediary component between the image recognizer and the robot controller. It processes the recognized target information, evaluates visibility based on multiple criteria (occlusion, position, size), and provides optimized target selection feedback to the robot controller.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If the robot attempts to grasp all detected work targets, then the productivity is high, but the failure rate increases due to poor visibility of some targets

Engineering Contradiction:

Improvegrasping throughputVSAvoidgrasping success rate

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements a feedback mechanism where the visibility estimation results are fed back to the work target specifier, which then adjusts target selection based on visibility scores. This feedback loop ensures that only targets with sufficient visibility are selected, maintaining high success rates while preserving productivity.

Inventive Principle:

Principle #23Feedback

3Reliability

If the system uses detailed visibility analysis to select work targets, then the grasping success rate improves, but the processing time increases

Engineering Contradiction:

Improvegrasping success rateVSAvoidtarget selection time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs visibility analysis on all detected work targets simultaneously, evaluating multiple criteria (occlusion degree, position in field of view, size) for each target. This comprehensive partial analysis ensures high grasping success rates while maintaining efficient processing through parallel evaluation.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentEP4703096A1Robot control device, robot control system, trained model, and method for generating trained model

Publication Date: 2026.03.04 KYOCERA CORP

Figure 1
Figure 2
Figure 3

AI summary

A robot control apparatus includes an acquisition interface that acquires an image of a plurality of work targets of a robot, an estimator that estimates the visibility of each of the work targets based on the image, and a work target specifier that selects a specific work target to be held by the robot from among the plurality of work targets based on the visibility.