Detecting retail shrinkage using behavioral analytics

• US 9,218,580 B2
• Filed: 12/30/2010
• Issued: 12/22/2015
• Est. Priority Date: 12/30/2010
• Status: Active Grant

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Libal, Vit, Trojanova, Jana, Eswara, Lalitha M.
• Primary Examiner(s): Diep, Nhon
• Assistant Examiner(s): SECHSER, JILL D
• Application Number:
  US12/981,922
• Publication Number:
  US 20120169879A1
• Time in Patent Office: 1,818 Days
• US Class Current: 1/1
• CPC Class Codes: G06Q 10/063 Operations research, analysis or management
• CPC Class Codes: G06Q 20/20 Point-of-sale [POS] network systems
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions
• CPC Class Codes: G07G 3/00 Alarm indicators, e.g. bells
• CPC Class Codes: G07G 3/003 Anti-theft control theft alarms G08B13/00
• CPC Class Codes: G08B 13/1672 using sonic detecting means, e.g. a microphone operating in the audio frequency range
• CPC Class Codes: G08B 13/194 using image scanning and comparing systems
• CPC Class Codes: G08B 13/19613 Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
• CPC Class Codes: H04N 7/183 for receiving images from a single remote source

Abstract

Methods, devices, and systems for detecting retail shrinkage using behavior analytics are described herein. The retail shrinkage may be due to, for example, sweethearting, although embodiments of the present disclosure are not so limited and can be used to detect other forms of retail shrinkage as well. One or more device embodiments include a memory, and a processor coupled to the memory. The processor is configured to execute executable instructions stored in the memory to receive data associated with behavior of an individual and use the data associated with the behavior of the individual to determine whether the behavior of the individual is irregular to detect retail shrinkage.

Claims

1. A computing device for detecting retail shrinkage, comprising:

a memory; and

a processor coupled to the memory, wherein the processor is configured to execute executable instructions stored in the memory to;

receive data associated with behavior of an individual, wherein the data associated with the behavior of the individual includes;

data associated with hand motion of a customer at a point of sale site;

data associated with body motion or body part motion of the customer at the point of sale site, wherein the body motion or body part motion is in addition to the hand motion; and

data associated with a facial expression of the customer at the point of sale site;

establish a behavior model for the individual by applying a pattern recognition method to additional data associated with the behavior of the individual during additional transactions at the point of sale site and by applying the pattern recognition method to data associated with behavior of additional individuals during additional transactions at other point of sale sites; and

use the data associated with the behavior of the individual and the behavior model to determine whether the behavior of the individual is irregular to detect retail shrinkage.

2. The computing device of claim 1, wherein the processor is configured to execute executable instructions stored in the memory to:

compare the data associated with the behavior of the individual to the established behavior model to determine whether the behavior of the individual is irregular to detect retail shrinkage.

3. The computing device of claim 1, wherein the data associated with the behavior of the individual includes data associated with a number of items, scanning of a number of items, or a transaction log.

4. The computing device of claim 1, wherein the data associated with the behavior of the individual includes data associated with speech of the customer, including speech presence, content, tone, volume, words, and expression.

5. A computer implemented method for detecting retail shrinkage, comprising:

receiving, from a number of sensors, data associated with behavior of an individual during a transaction at a point of sale site, wherein the data associated with the behavior of the individual includes;

data associated with hand motion of a customer at the point of sale site, wherein the data associated with the hand motion of the customer at the point of sale site includes data associated with a handshake between the customer and an employee at the point of sale site; and

data associated with body motion or body part motion of the customer at the point of sale site, wherein the body motion or body part motion in is addition to the hand motion;

establishing a behavior model based on data received from the number of sensors for regular and irregular transactions and by applying a pattern recognition method to additional data associated with the behavior of the individual during additional transactions at the point of sale site and applying the pattern recognition method to data associated with behavior of additional individuals during additional transactions at other point of sale sites;

comparing the data associated with the behavior of the individual during the transaction with the behavior model to determine whether the behavior of the individual during the transaction is irregular; and

initiating a corrective action if the behavior of the individual during the transaction is determined to be irregular.

6. The method of claim 5, wherein the additional data is received from the number of sensors.

7. The method of claim 5, wherein the pattern recognition method is a multi-layer, hierarchal, adaptive, online learning based pattern recognition method, a discriminative pattern recognition method, or a representative pattern recognition method.

8. The method of claim 5, wherein the pattern recognition method uses the additional data to recognize a level of abstractness within an activity or behavior.

9. The method of claim 5, wherein:

the behavior model includes a behavior pattern associated with regular behavior of the individual; and

the behavior of the individual during the transaction is irregular if the data associated with the behavior of the individual during the transaction does not fit within the behavior pattern associated with the regular behavior of the individual.

10. The method of claim 5, wherein:

the behavior model includes a behavior pattern associated with irregular behavior of the individual; and

the behavior of the individual during the transaction is irregular if the data associated with the behavior of the individual during the transaction fits within the behavior pattern associated with the irregular behavior of the individual.

11. The method of claim 5, wherein the behavior model uses body part motion or body motion of the individual in conjunction with speech expression, tone, volume, content, and item pickup for detection of sweethearting or fraud.

12. The method of claim 5, wherein initiating the corrective action includes generating a report of the transaction, instructing the number of sensors to record the transaction, or providing a security alert.

13. The method of claim 5, wherein the method includes:

receiving, from the number of sensors, data associated with behavior of the individual during a number of additional transactions at the point of sale site;

comparing the data associated with the behavior of the individual during the number of additional transactions with the behavior model to determine whether the behavior of the individual during the number of additional transactions is irregular; and

initiating the corrective action if a frequency of the number of transactions during which the behavior of the individual is determined to be irregular is above a pre-defined frequency.

14. The method of claim 5, wherein the method includes assigning a confidence value to the data associated with the behavior of the individual during the transaction.

15. A system for detecting retail shrinkage, comprising:

a number of sensors configured to sense data associated with behavior of a number of individuals at a point of sale site, wherein the data associated with the behavior of the number of individuals includes;

body motion of a customer at the point of sale site;

body part motion of the customer at the point of sale site;

hand motion of the customer at the point of sale site, wherein the hand motion is in addition to the body motion and body part motion and the hand motion includes a handshake between the customer and an employee at the point of sale site; and

a facial expression of the customer at the point of sale site and

a computing device configured to;

establish a behavior model for the number of individuals by applying a pattern recognition method to additional data associated with the behavior of the number of individuals during additional transactions at the point of sale site and by applying the pattern recognition method to data associated with behavior of additional individuals during additional transactions at other point of sale sites; and

use the sensed data associated with the behavior of the number of individuals at the point of sale site and the behavior model to determine whether the behavior of the number of individuals is irregular.

16. The system of claim 15, wherein the number of sensors includes an optical sensor and an acoustic sensor.

17. The system of claim 15, wherein the number of sensors include at least one of:

a video camera located overhead of the point of sale site;

a video camera located to a side of the point of sale site; and

a microphone located at the point of sale site.

18. The system of claim 15, wherein the data associated with the behavior of the number of individuals at the point of sale site includes data associated with an interaction between the number of individuals at the point of sale site.