Robotic Picking Control Using AI Weight Estimation From Shape

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing picking systems face difficulties in securely holding objects with varying shapes and weights, such as food products, due to the need for precise adjustment of holding force, which can lead to inefficiencies and increased production costs.

Innovation Solution

A picking system that includes a shape obtainer, a weight estimator using artificial neural networks for quick weight estimation based on shape information, and a controller to optimize the holding force of the gripper, eliminating the need for a weight sensor during the picking operation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a robot uses a fixed gripping force for picking operations, then the device complexity is reduced, but the robot cannot securely hold objects with varying shapes and weights

Engineering Contradiction:

Improveability to handle objects with varying shapes and weightsVSAvoidcomplexity of weight estimation system

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent replaces the mechanical weight sensor system with an artificial intelligence-based weight estimation system that uses image processing and neural networks. Instead of physically measuring weight through sensors, the system estimates weight by analyzing object shape and appearance from images, thereby avoiding additional mechanical components while achieving adaptive gripping force control

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent creates a digital model or representation of the object's physical properties by generating a three-dimensional model from multiple two-dimensional images. This digital copy contains shape and volume information that can be used to estimate weight without physical contact, allowing the system to adapt to varying object characteristics without additional sensors

Inventive Principle:

Principle #26Copying

2Measurement precision

If a weight sensor is added to measure object weight, then the picking accuracy is improved, but the production cost increases

Engineering Contradiction:

Improveweight measurement accuracyVSAvoidproduction cost of picking system

Core Design Contradiction:

Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent substitutes expensive mechanical weight sensors with a software-based AI estimation system using artificial neural networks and image processing algorithms. This replacement maintains measurement precision by accurately estimating weight from visual data while significantly reducing hardware costs and system complexity

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system performs self-characterization by automatically analyzing object properties from images and generating its own weight estimates without external measurement devices. The AI model learns from training data to independently determine weight based on visual features, eliminating the need for separate sensing infrastructure

Inventive Principle:

Principle #25Self-service

3Measurement precision

If the robot picks up objects one by one with manual weight checking, then the picking precision is improved, but the productivity decreases

Engineering Contradiction:

Improveobject weight detection accuracyVSAvoidpicking speed

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent performs weight estimation before the picking operation by analyzing images of objects in their original container. The AI system pre-calculates weight and identifies optimal picking targets, so that when the robot gripper approaches, all necessary information is already available, enabling high-speed picking without pause for weight measurement

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system maintains continuous operation by performing weight estimation and picking planning simultaneously through parallel processing. While the robot moves to the next object, the AI continuously analyzes images and updates weight estimates, ensuring no time is lost between picking operations and maintaining high productivity

Inventive Principle:

Principle #20Continuity of useful action

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This configuration enhances the picking system's ability to handle objects with varying shapes and weights efficiently, reducing production costs and improving the accuracy and speed of object allocation to containers.

Implementation Method 1

a weight estimator configured to estimate a weight of the object based on the shape information obtained by the shape obtainer

Methodology Applied

Scientific EffectArtificial neural network:

Data Source

PatentEP3284563B1Picking system

Publication Date: 2022.04.20 YASKAWA DENKI KK

Figure 1
Figure 2
Figure 3~4

AI summary

A picking system includes a shape obtainer, a weight estimator, a picking robot, and a controller. The shape obtainer is configured to obtain shape information of an object. The weight estimator is configured to estimate a weight of the object based on the shape information obtained by the shape obtainer. The picking robot is configured to perform a picking operation on the object. The controller is configured to control the picking operation based on the weight estimated by the weight estimator.