Automated Order Checking System for Retail Surveillance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High traffic of persons and goods in retail environments with warehouses poses challenges in tracking individual movements, preventing unauthorized entry, pilferage, and ensuring accurate receipt/dispatch of goods, which traditional manual methods are inefficient in handling.
Innovation Solution
An automated order checking system utilizing video sensors to capture footage, a processing unit for event analysis and entity detection, and a database for storing face and product images, enabling the identification of third-party suppliers, validation of delivered products, and automated check-in/check-out processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual security personnel are deployed to track persons and goods, then surveillance can be performed, but the process becomes tedious and inconsistent under high traffic conditions
Solution Approach 1:
The patent replaces manual mechanical surveillance by security personnel with an automated video-based surveillance system. The system uses video sensors to capture footage and a processing unit with machine learning models to automatically detect entities (persons, goods, vehicles), track their movements, and analyze events such as door openings and ingress/egress activities. This substitution eliminates human limitations in handling high traffic while maintaining consistent and reliable surveillance.
Solution Approach 2:
The surveillance system performs self-service by automatically detecting, tracking, and analyzing all surveillance tasks without human intervention. The processing unit autonomously processes video footage, identifies entities, tracks their movements across multiple zones, detects events, and generates alerts when necessary. This self-service capability allows the system to handle high traffic volumes consistently without the tedious manual effort required by security personnel.
2Measurement precision
If manual tracking of individual movements is performed, then surveillance accuracy can be maintained, but the system cannot handle high traffic efficiently
Solution Approach 1:
The patent segments the surveillance area into multiple zones (e.g., first zone, second zone, third zone) and uses separate tracking mechanisms for different entity types (persons, goods, vehicles). Each entity is assigned a unique identifier and tracked independently through the zones. This segmentation allows the system to maintain precise tracking of individual movements while efficiently processing high traffic volumes by dividing the complex tracking task into manageable segments.
Solution Approach 2:
The system creates digital copies of entities through video frames and tracking data structures. Each detected entity is represented by a digital model containing its position, velocity, zone information, and identification data. These digital copies are updated continuously as entities move through the surveillance area, enabling precise tracking without physical intervention and allowing parallel processing of multiple entities simultaneously.
3Productivity
If automated video-based surveillance is implemented, then productivity and consistency are improved, but device complexity increases
Solution Approach 1:
The processing unit serves multiple functions within a single integrated system: it detects entities, tracks their movements, analyzes events (door openings, ingress/egress), identifies objects using machine learning models, and generates alerts. The video sensors also serve dual purposes by capturing both security surveillance footage and inventory monitoring data. This multi-functionality reduces the need for separate dedicated systems for each task, thereby managing complexity while maintaining high productivity.
Solution Approach 2:
The patent introduces an intermediary layer of machine learning models and event analysis algorithms between the video sensors and the surveillance output. These intermediaries process raw video data, extract meaningful information (entity detection, tracking, event recognition), and present processed results in a usable format. This intermediary layer simplifies the overall system architecture by handling complex processing tasks in a modular manner, improving productivity without proportionally increasing overall system complexity.
Data Source
AI summary
An order checking system includes video sensors configured to capture video footage of a monitored area located in proximity to a receipt/dispatch portal. A processing unit performs event analysis on the captured video footage, detects an entity and from a door opening event, an incoming delivery from a third-party supplier, identifies the third-party supplier and implements a check-in process for delivery persons associated therewith, detects an ingress/egress of merchandise through the receipt/dispatch portal and validates that detected delivered products matches with data regarding products that the third-party supplier should be delivering. A database stores at least a dataset of face images/logos for detecting faces/brands and a dataset of product images for identification of products. The database records an outcome of an order checking process and a check out of a delivery person at an end of a delivery.