Material Handling Flow Control Using 3D Object Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current material handling systems face challenges in accurately identifying and automating the handling of cases and pallets with diverse product trade-dresses and UPC codes, often requiring manual intervention and increasing the risk of errors in order fulfillment.
Innovation Solution
A fully or partially automated material handling system that uses 3D image capture and database comparison to identify objects, allowing for reliable operation independent of manufacturer-provided data, and applies labels to ensure correct product identification without human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine readable UPC codes are used for product identification, then automated identification is enabled, but identification accuracy deteriorates when codes are on difficult-to-access surfaces or when multiple products share the same UPC
Solution Approach 1:
The system transitions from 2D UPC code scanning to 3D image capture and analysis. By capturing images from multiple angles and using point cloud generation, the system can identify products based on their three-dimensional trade dress characteristics, overcoming the limitation of UPC codes being on bottom surfaces or shared across multiple products.
Solution Approach 2:
The system creates digital 3D copies (point clouds) of physical products and stores them in a database. These digital twins can be compared and matched without needing to physically access or scan the original product's UPC code, enabling accurate identification through visual特征 matching instead of code reading.
2Measurement precision
If human operators manually identify and label objects, then identification accuracy is maintained, but productivity and automation level decrease
Solution Approach 1:
The system enables automated self-identification of products through 3D image capture and database comparison. Objects are automatically identified and labeled without human intervention, with the system independently matching captured images against stored point clouds to determine product identity and apply appropriate labels.
Solution Approach 2:
The patent replaces manual human identification and labeling operations with an automated optical and computational system. Image capture stations, point cloud generation, and database matching algorithms substitute for human visual inspection and decision-making, while automated labeling mechanisms replace manual label application.
3Reliability
If labels are applied over manufacturer trade-dress, then product identification reliability is improved, but manufacturer objectivity and adaptability deteriorate due to decorative packaging concerns
Solution Approach 1:
The system performs 3D image capture and product identification before any labeling operation. By establishing product identity through visual特征 matching in advance, the system determines where labels should be placed without necessarily covering decorative areas, allowing label placement to be planned around the product's trade dress rather than simply overlaying it.
Solution Approach 2:
The system applies labels at specific locations determined by the 3D image analysis and database matching results. Rather than uniformly applying labels over trade-dress, the system identifies appropriate labeling zones that maintain product identification reliability while preserving decorative packaging areas where possible, based on the specific product geometry and characteristics.
Data Source
AI summary
A material handling system having object handling and transportation devices and method of receiving objects into a warehouse having a material handling system includes scanning a machine readable code of an incoming load forming at least a portion of an order and comparing the machine readable code with an order file listing objects on the incoming load. For each object on the load, object data including a 3D image and product code of the object removed from the incoming load is received with an image capture station and reconciled with the order file. Each object is identified by comparing the 3D image of the object captured with the image capture station with the 3D images in the database. The material handling system is controlled as a function of the identity of the objects being handled and transported by the material handling system.