IP Library Granted Patent US 12711285
Granted Patent B2
US 12711285 · App. 19/002,240 · Granted Aug 18, 2026

Smart cart security with auditing

Inventors: Kip Oliver Morgan (Atlanta, GA); Gina Torcivia Bennett (Lawrenceville, GA); Jay Arcement (Atlanta, GA)
Assignee: NCR Voyix Corporation
G06F21/88G06Q20/12G06Q20/18G06Q20/208G06Q20/4016G06Q30/0641G06V10/70G06V20/52H04W12/12H04W12/126
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12711285
App. No.
19/002,240
Granted
Aug 18, 2026
Kind
B2
Abstract

Methods and a system for automated smart cart security and auditing are provided. A cart includes one or more cameras mounted thereon to monitor items placed in the cart. The system performs local item recognition on the items placed in the cart by a mobile shopping device, focusing particularly on high-value items to prevent shrink. When discrepancies are detected between scanned items and items placed in the cart, the system accumulates audit points based on factors including, by way of example only, customer trust level, store shrink rates, and transaction risk scores. Upon reaching configured thresholds, the system triggers automated audit interventions that must be resolved before checkouts at self-service checkout (SCO) terminals.

Claims (21)

1 . A method, comprising:

triggering an activation of at least one camera mounted to a cart during a shopping journey of a customer within a store based on a receipt of a first item code scanned by a mobile device of the customer during a self-shopping transaction;

attempting to perform a local visual item recognition on at least one image provided by the at least one camera;

determining whether the first item code matches a second item code when the performing returns a second item code;

assigning audit points to the self-shopping transaction based on at least the determining; and

initiating an audit intervention for the self-shopping transaction when the customer initiates a checkout at a self-service terminal (SST) and the audit points exceed a threshold.

2 . The method of claim 1 , wherein triggering further includes

triggering a first activation of a first camera that monitors an interior of the cart and

triggering a second activation of a second camera that monitors an exterior or a bottom of the cart.

3 . The method of claim 1 , wherein attempting to perform further includes

recognizing a particular item from the at least one image and

assigning the second item code when the particular item is associated with a cost above a predefined threshold.

4 . The method of claim 1 , wherein assigning further includes

assigning additional audit points to the self-shopping transaction based on one or more of a shrink rate associated with store, a loyalty level assigned to the customer, a shopping pattern of the customer during the self-shopping transaction, or one or more locations within the store where particular items were scanned by the customer.

5 . The method of claim 1 , wherein initiating further includes

suspending the checkout at the SST when the audit points exceed the threshold and

sending specific item information in an audit notification to an attendant to perform an audit of the self-shopping transaction at the SST.

6 . The method of claim 1 , further comprising:

initially operating in a silent data collection mode; and

using collected data during the silent data collection mode to adjust the local visual item recognition and assignment of the audit points based on actual audit results for actual audits performed on previous self-shopping transactions.

7 . The method of claim 1 , further comprising: periodically updating a local device mounted to the cart with specific item visual features for specific items of a product catalog that exceed a cost threshold to enable the local visual item recognition.