Consumer Packaged Goods Identification Before GTIN Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated inventory systems for consumer packaged goods (CPGs) require manual barcode scanning for new or changed products, which is time-consuming and costly, and lack efficient methods to collect data and train models for new products before GTIN identification.
Innovation Solution
A method using AI-assisted human audit procedures to identify partially-identified products through deep learning classifiers, allowing data collection and training before GTIN assignment, by associating products with GTINs and metadata, and continuously refining the classifier with corrected data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual barcode scanning is used for new or changed products, then product identification accuracy is improved, but time consumption and operational cost increase
Solution Approach 1:
The patent replaces the manual mechanical barcode scanning process with an automated computer vision system using deep learning classifiers. The system captures images of products on shelves and automatically identifies them through image processing and classification algorithms, eliminating the need for manual physical contact and barcode scanning while maintaining identification accuracy.
Solution Approach 2:
The patent creates visual copies (images) of products on shelf and uses these copies for identification purposes. Instead of requiring direct contact with the physical product to scan barcodes, the system uses captured images as representations of the products, allowing automated recognition through machine learning models trained on product visual characteristics.
2Productivity
If automated image-based classification is used for new products, then operational speed is improved, but classification reliability deteriorates for previously unseen products
Solution Approach 1:
The patent performs preliminary actions by capturing multiple images of new products at different stages before full GTIN identification is complete. The system collects visual data and metadata about new products and creates partial product descriptions that can be used for immediate classification and monitoring, with the understanding that full identification will be completed through subsequent human review and GTIN assignment.
Solution Approach 2:
The patent introduces human reviewers as an intermediary component in the classification pipeline. When the automated deep learning classifier encounters uncertainty about a new product, the system transfers the task to human reviewers who provide corrective feedback. This human-in-the-loop approach ensures reliable classification for new products while maintaining high operational speed through automated processing of known products.
3Measurement precision
If deep learning classifiers are continuously trained with new product data, then classification accuracy is improved, but system complexity increases
Solution Approach 1:
The patent implements feedback mechanisms where human reviewers correct misclassifications by providing ground truth labels for new products. These corrected data are then fed back into the training process to refine the deep learning classifiers. The system automatically identifies when retraining is needed and updates the models using the corrected data, creating a continuous improvement loop that maintains high accuracy without requiring manual reconfiguration of the entire system.
Data Source
AI summary
A method is provided for identifying consumer packaged goods (CPGs). The method includes providing to a machine learning classifier a set of images containing at least one CPG; receiving from the machine learning classifier an indication that the machine learning classifier cannot reliably identify a designated CPG in the set of images; determining whether the designated CPG is a product in a product catalog; if the designated CPG is in the product catalog, then associating the designated CPG with a Global Trade Item Number (GTIN); and if the designated product is not in the product catalog, then designating the CPG as a potentially new product. Notably, this approach allows partially identified products to be treated as full-fledged members of the product catalog, thus allowing data to be collected on these products even before they have been fully identified and their GTINs have been resolved.