Roll-Fed Packaging Web Inspection for Early Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual quality control of packages produced in roll-fed packaging machines is inefficient, leading to waste and reduced confidence in package quality, as it is time-consuming and prone to errors, especially in detecting anomalies in packaging material and sealing.
Innovation Solution
Implementing a machine learning-based method that uses an image data capturing device to continuously monitor packaging material for anomalies, allowing for early detection and action to be taken to reduce waste and ensure package quality, with a machine learning model trained to identify anomalies in image data from the packaging material.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual quality control is used to evaluate batches of finished products, then quality assessment can be performed, but it is time-consuming and leads to waste of product
Solution Approach 1:
The patent replaces manual visual inspection with an automated image recognition system using machine learning models. The system captures images of packaging material and packages through cameras, processes them with trained neural networks to detect anomalies such as defects, misalignments, and sealing issues, thereby eliminating the need for time-consuming manual evaluation while maintaining or improving detection accuracy
Solution Approach 2:
The patent creates digital copies (images) of the physical packaging material and packages for analysis. Instead of manually examining physical items, the system captures visual representations and analyzes them through machine learning algorithms, enabling parallel processing of multiple items simultaneously and significantly reducing quality control time
2Reliability
If manual quality control is used to ensure package quality, then quality can be assessed, but it requires large quantities of packages to be assessed and leads to product waste
Solution Approach 1:
The patent performs quality inspection at multiple early stages in the packaging process - examining packaging material before it enters the packaging machine and inspecting packages during or immediately after formation. By detecting anomalies early, the system prevents defective packages from being produced or distributed, eliminating the need to assess large quantities of finished products and reducing waste
Solution Approach 2:
The automated machine learning-based inspection system provides more consistent and reliable quality assessment compared to manual inspection. The system can detect subtle anomalies that human inspectors might miss, reducing false negatives and ensuring higher quality assurance without requiring excessive sampling
3Reliability
If manual visual examination is used to detect anomalies in packaging material, then quality issues can be identified, but it is prone to errors and reduces confidence in package quality
Solution Approach 1:
The patent replaces human visual inspection with automated machine learning models trained to recognize various anomalies in packaging material and packages. The system consistently applies learned patterns to detect defects, misalignments, and sealing issues without fatigue or human error, significantly improving detection accuracy and reliability while maintaining operational simplicity through automated workflows
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a method (100) for quality control of packages produced in a roll-fed packaging machine, wherein the roll-fed packaging machine comprises an image data capturing device arranged at a position along a web of packaging material. The method (100) comprising: obtaining (S102) image data depicting a section of the web, from the image data capturing device; identifying (S104) an anomaly in the section of the web of packaging material, using a machine learning model trained to identify anomalies in image data of a web of packaging material; and communicating (S112), in response to the identified anomaly, a signal indicating an action to be performed, to a control system of the packaging machine.