Neural Network Classifier for RFID Self-Checkout Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Customer self-checkout processes are inefficient due to the need for manual alignment of product barcodes, leading to delays and frustration, as existing systems rely solely on customers to correctly scan items, resulting in lengthy checkout times and reduced satisfaction.
Innovation Solution
A point-of-sale checkout system utilizing radio-frequency identification (RFID) with a neural network classifier and strategically placed RF-absorbing walls, allowing customers to simply stand in a designated area while the system accurately identifies and decodes RFID tags without line-of-sight requirements, minimizing incorrect item identification and enhancing customer satisfaction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual barcode scanning is used for self-checkout, then customers can identify items, but checkout time increases and customer satisfaction decreases
Solution Approach 1:
The patent replaces the mechanical/manual barcode scanning process with an automated RFID detection system. RFID readers automatically detect items in the customer's possession without requiring manual alignment or line-of-sight, thereby reducing checkout time while maintaining or improving identification accuracy through automated multi-item detection.
Solution Approach 2:
The system performs preliminary item identification and classification before the actual checkout transaction. RFID readers continuously detect items in the customer's possession ahead of time, and machine learning models pre-process the data to distinguish purchased items from non-purchased items, so that when checkout is initiated, the system is already prepared with accurate item lists.
2Extent of automation
If RFID readers detect all items in customer possession, then checkout is automated, but items not purchased may be incorrectly charged
Solution Approach 1:
The patent introduces machine learning classifiers as intermediary components between RFID detection and charging decisions. These classifiers analyze RFID signal characteristics, spatial information, and temporal patterns to distinguish between items the customer intends to purchase and items merely in possession, thereby preventing incorrect charges while maintaining automation.
Solution Approach 2:
The system implements feedback mechanisms where the classifier continuously refines its predictions based on detected item patterns, customer behavior data, and correction feedback from customers when misclassifications occur. This feedback loop improves classification accuracy over time, reducing false charges while maintaining high automation levels.
3Measurement precision
If multiple RFID transceivers are deployed for coverage, then detection accuracy improves, but system complexity increases
Solution Approach 1:
The patent divides the detection task across multiple transceivers, each responsible for specific spatial zones. By segmenting the detection coverage area and assigning transceivers to specific regions, the system achieves comprehensive detection accuracy while managing complexity through distributed, modular architecture where each transceiver operates semi-independently.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system significantly reduces customer dissatisfaction by streamlining the checkout process, minimizing vendor losses, and maximizing customer satisfaction through accurate identification and charging of purchased items, while eliminating the need for continuous customer involvement and extensive human intervention.
Implementation Method 1
a plurality of radio frequency identification (RFID) transceivers within a store, and an RFID reader configured to receive an RFID code from an RFID tag activated by the plurality of radio frequency identification (RFID) transceivers
Implementation Method 2
A point-of-sale checkout system utilizing radio-frequency identification (RFID) with a neural network classifier and strategically placed RF-absorbing walls
Data Source
AI summary
Systems and methods for self-checkout at a point-of-sale are provided. The system and method includes using a plurality of radio frequency identification (RFID) transceivers within a store, and an RFID reader configured to receive an RFID code from an RFID tag activated by the plurality of radio frequency identification (RFID) transceivers. The system and method also includes using a classifier configured to determine whether the RFID tag is inside or outside a designated area, wherein the classifier is trained in a manner that a number of items incorrectly identified as being purchased is below a threshold to minimize customer dissatisfaction (CDS) determined as the ratio of the value of items charged to the customer but not purchased by a customer to the total charge to the customer.