ML-Based Material Package Remaining Quantity Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current just-in-time manufacturing systems face challenges in accurately tracking and managing remaining materials on packages, particularly in scenarios where traditional weight measurement methods like scales or load cells are not feasible, such as during material transportation or in disaster response situations.

Innovation Solution

Implementing a system that uses Machine Learning (ML) and image analysis to estimate the remaining amount of material on packages by analyzing images of the package's top portion, combined with data from tracking devices and databases, allowing for continuous improvement of the estimation algorithm over time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional weight measurement methods (scales or load cells) are used to track remaining materials, then measurement precision is improved, but device complexity and ease of operation worsen due to the need for physical weighing infrastructure

Engineering Contradiction:

Improveremaining material quantityVSAvoidweighing infrastructure

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces mechanical weighing systems (scales and load cells) with an optical image analysis system. Images of material packages are captured and processed by machine learning models to estimate remaining material quantities, eliminating the need for physical weighing infrastructure while maintaining measurement capability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system creates visual copies (images) of material packages and analyzes these copies to determine remaining material quantities. Instead of physically weighing the actual packages, the system works with image representations, which can be captured and analyzed remotely without requiring the packages to be placed on physical scales.

Inventive Principle:

Principle #26Copying

2Measurement precision

If traditional weight measurement methods are used, then measurement precision is improved, but ease of operation worsens due to manual data entry requirements

Engineering Contradiction:

Improveremaining material quantityVSAvoiddata entry process

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The system performs automatic image capture and analysis without requiring manual intervention for data entry. The machine learning model automatically processes images and generates remaining material estimates, making the system self-sufficient in data collection and processing, eliminating the need for operators to manually record weighing data.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Manual data entry operations are replaced by automated image analysis. The system captures images and uses computer vision algorithms to extract material quantity information automatically, substituting human-operated data entry with an automated optical measurement system.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Device complexity

If image analysis with Machine Learning is used to estimate remaining material, then device complexity is reduced by eliminating scales, but measurement precision may worsen compared to direct weight measurement

Engineering Contradiction:

Improvemeasurement systemVSAvoidremaining material quantity

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The system incorporates feedback mechanisms where machine learning models are continuously trained and refined using actual measurement data. The models learn from discrepancies between estimated and actual material quantities, progressively improving precision while maintaining the simplicity of the image-based approach.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system changes the measurement parameter from direct weight to visual characteristics of material packages. By analyzing image parameters such as package dimensions, material density patterns, and package configuration, the system estimates remaining material quantities through a different physical parameter set that can be obtained optically rather than mechanically.

Inventive Principle:

Principle #35Parameter changes

4Productivity

If automated image analysis is implemented, then productivity is improved through continuous tracking, but loss of information increases due to reliance on image data quality

Engineering Contradiction:

Improvematerial tracking efficiencyVSAvoidmaterial quantity accuracy

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The system performs preliminary actions by capturing images of material packages at various stages and establishing baseline data before material is consumed. This allows the machine learning model to compare current package states against known reference states, improving the accuracy of remaining material estimates while maintaining continuous tracking capability.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12067742B2Providing partial material package remaining material and location

Publication Date: 2024.08.20 SOUTHWIRE CO LLC

AI summary

Remaining material on a material package and its location may be determined. First, image data associated with a material package may be received by a server. Next, material package data associated with the material package may be received. A Machine Learning (ML) model may then be used to determine an estimated remaining amount of material associated with the material package based upon input derived from the image data and the material package data.