Perpetual Inventory Reconciliation via Image Recognition

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Perpetual inventory systems face errors due to inventory overstatements, understatements, shrinkage, missing items, and incorrect scanning, leading to discrepancies between physical and system inventories, which can result in inadequate replenishment and are labor-intensive to correct.

Innovation Solution

A computer-implemented method for perpetual inventory reconciliation that analyzes item data using purge criteria and confidence rules to identify inactive seasonal items, prioritizes them, and uses sensor data to verify physical presence, updating system records accurately.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual scanning of physical inventory items is performed to update system inventory, then inventory accuracy can be improved, but labor time and operational complexity increase significantly

Engineering Contradiction:
Improveinventory accuracyVSAvoidlabor time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical scanning operations with an automated image recognition system using cameras and machine learning algorithms to detect, identify, and count physical inventory items, thereby eliminating labor-intensive manual scanning while maintaining or improving inventory accuracy

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables the inventory to essentially count itself by capturing images of physical inventory and using automated image analysis to update system inventory records without human intervention, allowing the inventory management system to self-update based on visual detection of actual stock levels

Inventive Principle:
Principle #25Self-service

2Reliability

If comprehensive manual inventory verification is performed to eliminate discrepancies, then inventory record accuracy improves, but operational complexity and resource requirements increase

Engineering Contradiction:
Improveinventory record accuracyVSAvoidoperational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent substitutes complex manual inventory verification procedures with an automated computer vision system that captures images and uses machine learning to automatically reconcile physical inventory with system records, reducing operational complexity while improving reliability

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system introduces an intermediary image analysis component that acts as a mediator between physical inventory and system records, using captured images as intermediate data to automatically detect discrepancies and update inventory levels without requiring direct manual comparison

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If system inventory is maintained with high precision to prevent replenishment errors, then inventory management reliability improves, but the system becomes more sensitive to updating errors and omissions

Engineering Contradiction:
Improvereplenishment accuracyVSAvoidsensitivity to errors
Core Design Contradiction:
ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The system performs preliminary inventory verification by capturing images of physical inventory before discrepancies can affect replenishment decisions, proactively identifying and correcting inventory errors before they lead to replenishment mistakes

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements continuous feedback loops where image-based inventory detection results are automatically fed back to update system inventory records, creating a self-correcting mechanism that reduces the impact of errors and maintains high replenishment accuracy

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS10846654B2Perpetual inventory reconciliation
Publication Date: 2020.11.24 WALMART APOLLO LLC
  • US10846654B2 patent drawing
  • US10846654B2 patent drawing
  • US10846654B2 patent drawing

AI summary

Examples provide a perpetual inventory (PI) reconciliation system. A PI controller analyzes item data using a set of PI purge criteria and a set of weighted prioritization variables to select a highest priority seasonal inventory item having a positive PI value a per-item threshold time after an end-of-season (EOS) date for PI purge analysis. A PI controller determines if the selected item is a high confidence item or a low confidence item based on a set of confidence rules and item data. If the selected item is a high confidence item, a PI purge component sets an inventory value for the selected item to zero to eliminate the positive PI. If the selected item is a low confidence item, a verification component verifies the number of physical instances of the selected item. The inventory value is updated using the verified number of physical instances.