Virtual Model Deviation Detection in Packaging Containers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting deviations in packaging containers, such as defects or inconsistencies, are inefficient and disrupt production lines, especially in high-speed manufacturing of sealed containers for liquid or semi-liquid food, making it challenging to implement reliable quality control with minimal resource usage.
Innovation Solution
A method and system that create a virtual model of a packaging container with controlled deformations, generate image features representing deviations, and use a machine learning-based model for detection, allowing for efficient identification of various types of deviations without disrupting production.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional deviation detection methods are used in high-speed production lines, then production throughput is maintained, but detection accuracy and reliability deteriorate due to disruption and resource constraints
Solution Approach 1:
The system performs preliminary actions by creating a virtual model of the packaging container and pre-defining deformation zones and deviation types before actual production detection. This preparation enables the machine learning model to be trained in advance with synthetic deviation data, so that during high-speed production, the model can immediately detect deviations without disrupting the production line.
Solution Approach 2:
The invention creates virtual copies of packaging containers with controlled deformations to generate training data for the machine learning model. These virtual models replicate real container geometry and potential deviation patterns, allowing the system to learn deviation detection without needing physical defect samples or interrupting production for data collection.
2Reliability
If comprehensive quality control is implemented to detect all deviation types, then detection coverage is improved, but resource consumption and system complexity increase
Solution Approach 1:
The system segments the packaging container surface into specific deformation zones where deviations are most likely to occur. By focusing detection efforts on these predefined zones rather than analyzing the entire container surface, the system achieves comprehensive quality control for critical areas while reducing computational complexity and resource requirements.
Solution Approach 2:
The invention applies local quality by creating controlled deviations specifically within deformation zones rather than uniformly across the entire container. This allows the machine learning model to learn localized deviation patterns that are most relevant to quality control, improving detection reliability for actual defects while minimizing the complexity of analyzing all possible container variations.
3Measurement precision
If real physical deviations are used for training the detection system, then training accuracy is improved, but production disruption and time loss increase
Solution Approach 1:
The system creates virtual copies of packaging containers with programmed deviations to generate training data. These synthetic images replicate real deviation patterns including folds, wrinkles, and sealing defects, enabling the machine learning model to achieve high training accuracy without needing to interrupt production to collect physical defect samples or create test defects on real containers.
Solution Approach 2:
The invention performs preliminary action by generating and storing a comprehensive dataset of virtual deviation images before the detection system is deployed to production. This advance preparation includes creating variations of all expected deviation types in deformation zones, so that when the system goes live, it immediately has access to extensive training data without requiring production stoppage for data collection.
Data Source
AI summary
A method for detection of deviations in packaging containers is disclosed, comprising creating a virtual model of a packaging container in a virtual coordinate system (x, y, z), defining a deformation zone on a surface of the virtual model, creating a defined deviation in the deformation zone having a defined geometry and coordinates in the virtual coordinate system (x, y, z) to create a controlled deformation of the virtual model, producing an image rendering of the virtual model with said controlled deformation to generate image features representing a deviation in the packaging container, associating the image features with different categories of deviations, and inputting the image features to a machine learning-based model for subsequent detection of categories of deviations in packaging containers in a packaging machine based on the image features. A system for detection of deviations is also disclosed.