Computer Vision Item Tracking for Bin Placement Accuracy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional inventory systems require manual barcode scanning and data entry for item and bin identification, leading to inefficiencies and increased error rates in inventory management processes.

Innovation Solution

An automated system utilizing mobile drive units and machine learning models to track inventory movement by processing video data from cameras, eliminating the need for manual scanning and improving accuracy through bin prediction and correction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual barcode scanning and data entry is used for item and bin identification, then the system can track inventory items, but the processing efficiency decreases and error rates increase

Engineering Contradiction:

Improvetracking accuracyVSAvoidprocessing efficiency

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual mechanical barcode scanning and data entry operations with an automated computer vision system using cameras and machine learning models. The system automatically captures images of items and bins, identifies them through image recognition, and tracks inventory movements without human intervention, thereby eliminating the trade-off between accuracy and efficiency.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service tracking where the inventory management system automatically identifies items and bins through computer vision, updates inventory records, and monitors movements without requiring associate intervention. The automated detection and tracking processes perform themselves, eliminating manual data entry while maintaining high accuracy.

Inventive Principle:

Principle #25Self-service

2Loss of information

If manual data entry is required for each item interaction, then detailed tracking information can be collected, but the likelihood of errors increases

Engineering Contradiction:

Improvetracking information completenessVSAvoiderror rate

Core Design Contradiction:

Loss of informationVSReliability

Solution Approach 1:

The patent replaces error-prone manual data entry with automated computer vision-based identification. Cameras capture images of items and bins, and machine learning models automatically extract identification information, eliminating human errors while maintaining complete tracking records of all inventory interactions.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system implements continuous visual monitoring and automatic feedback loops where camera systems constantly capture inventory movements, machine learning models process the visual data in real-time, and the system automatically updates tracking records. This closed-loop feedback mechanism ensures complete information collection with high reliability by immediately detecting and recording all item interactions.

Inventive Principle:

Principle #23Feedback

3Loss of information

If associates manually enter information for items and destination locations, then inventory data can be recorded, but operational efficiency decreases

Engineering Contradiction:

Improveinventory data accuracyVSAvoiditems processed per time

Core Design Contradiction:

Loss of informationVSProductivity

Solution Approach 1:

The patent replaces manual information entry operations with automated computer vision systems that capture images of items and destination bins, automatically identify them through image recognition, and record inventory data without human intervention. This substitution dramatically increases items processed per time while maintaining data accuracy through automated machine learning-based identification.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system implements continuous automated monitoring where cameras continuously capture inventory movements, machine learning models continuously process visual data to identify items and bins, and the system continuously updates inventory records. This uninterrupted automated process eliminates the stop-start nature of manual data entry, maximizing throughput while maintaining accurate inventory data recording.

Inventive Principle:

Principle #20Continuity of useful action

Data Source

PatentUS11010903B1Computer vision and machine learning techniques for item tracking

Publication Date: 2021.05.18 AMAZON TECH INC

AI summary

Techniques are described for processing digital video data using one or more machine learning models to determine an outcome of an item placement operation within a fulfillment center environment. Video data is processed using one or more machine learning models to determine an estimated likelihood that an occurrence of a particular fulfillment center operation is depicted within the two or more instances of digital video data. Upon determining that the estimated likelihood exceeds a predefined threshold confidence level, the video data is processed using second one or more machine learning models to determine a bin placement prediction and a confidence value. A data repository for a control system for the fulfillment center environment is updated, based on the bin placement prediction and the confidence value.