Robot Object Designation Using 2D Shape Input for Complex Grasping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing robot systems struggle to accurately recognize objects, especially when objects are complex, intermingled, or have reflective surfaces, leading to difficulties in grasping and moving them correctly.

Innovation Solution

A designation device that allows a worker to input and display a two-dimensional image of an object's external form, enabling the robot system to grasp and move objects based on this designated form, even if the robot system cannot correctly recognize the object's pre-movement state.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If expensive industrial cameras are used for automatic recognition, then the robot system can recognize objects in normal conditions, but it fails to appropriately recognize individual objects when objects are complex, intermingled, reflective, or wrapped in cushioning material

Engineering Contradiction:

Improveobject recognition accuracyVSAvoidability to handle complex object conditions

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent uses a two-dimensional image as a copy or representation of the object's external form, allowing the robot to grasp and move objects based on this simplified representation rather than relying on complex automatic recognition of the actual object. This copying approach enables accurate object handling even when the original object has complex features that are difficult to recognize automatically.

Inventive Principle:

Principle #26Copying

2Extent of automation

If automatic recognition systems are used, then the robot system can operate without human intervention in simple cases, but it cannot correctly recognize objects in complex scenarios requiring manual designation

Engineering Contradiction:

Improveautomatic object recognitionVSAvoidobject recognition reliability

Core Design Contradiction:

Extent of automationVSReliability

Solution Approach 1:

The patent introduces a two-dimensional image as an intermediary between the object and the robot's grasping action. Instead of directly recognizing complex objects, the system uses the simplified 2D image representation as a mediator that the robot can reliably process, combining automatic operation with reliable object identification.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Productivity

If the robot system uses traditional automatic recognition, then it can handle simple objects efficiently, but it struggles with objects that have reflective surfaces, transparent materials, or are wrapped in cushioning material

Engineering Contradiction:

Improveobject handling efficiencyVSAvoidobject detection difficulty

Core Design Contradiction:

ProductivityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent extracts only the essential external form information needed for grasping and moving objects, represented as a two-dimensional image. By taking out only the necessary geometric information rather than attempting to fully recognize complex object properties, the system achieves efficient object handling while avoiding the difficulties of detecting reflective, transparent, or wrapped surfaces.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20250144807A1Designation device, robot system, designation method, and recording medium

Publication Date: 2025.05.08 NEC CORP

AI summary

A designation device receives an input designating at least a portion of an external form of an object to be moved in a robot system that moves the object to be moved by following a predetermined algorithm in accordance with a work goal. The designation device makes a display device display a two-dimensional image including the object to be moved, and the external form received by the reception means.