Customer Tracking Device for Retail Interaction Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Retailers face challenges in understanding how customers interact with products at physical retail locations, as existing methods lack the ability to automatically and efficiently analyze detailed customer actions, such as activities, eye gaze directions, and vocal expressions, across multiple locations.
Innovation Solution
A customer tracking device that processes video and audio data from cameras to identify customer actions, including activities, eye gaze directions, and vocal expressions, generating reports that categorize these interactions and provide insights for retailers to optimize product placement, labeling, and marketing strategies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis of customer interactions is used, then detailed customer actions can be understood, but it is time-consuming and subjective
Solution Approach 1:
The patent replaces manual mechanical analysis with an automated computer vision system that uses machine learning models to detect and analyze customer actions, eye gaze directions, and product interactions. This substitution eliminates human subjectivity and significantly reduces analysis time while maintaining or improving measurement precision through consistent algorithmic evaluation of video data from retail cameras.
Solution Approach 2:
The system enables self-service analysis by automatically processing video data without human intervention. The machine learning models autonomously identify customer actions, track eye gaze directions, and generate insights about product interactions, allowing the system to serve itself in analyzing multiple retail locations simultaneously without requiring manual review of each interaction.
2Productivity
If automated video analysis is implemented across multiple locations, then analysis efficiency improves, but computing resource consumption increases
Solution Approach 1:
The patent segments the analysis process into distinct machine learning models that handle specific tasks: customer detection, action recognition, eye gaze tracking, and product interaction identification. This segmentation allows each model to be optimized for its specific function and enables parallel processing across multiple retail locations, improving overall productivity while managing computing resource consumption through specialized rather than generalized processing.
Solution Approach 2:
The system implements partial action by focusing analysis only on relevant portions of video data - specifically when customers are near products or displaying interested behaviors. Rather than continuously analyzing all video feeds, the system activates detailed analysis only when triggered by detected customer presence or potential interest, thereby maintaining high productivity while reducing unnecessary computing resource consumption during periods of low customer activity.
3Loss of information
If detailed tracking of customer actions is performed, then granular data is obtained, but system complexity increases
Solution Approach 1:
The patent implements a universal machine learning framework that handles multiple types of customer behaviors and interactions through a single integrated system. The same core architecture detects customer presence, tracks eye gaze directions, identifies product interactions, and analyzes actions across diverse retail environments and product categories. This multi-functionality reduces system complexity compared to having separate specialized systems for each type of analysis, while still capturing granular engagement data.
Data Source
AI summary
A device receives video data concerning a plurality of customers and a product in a plurality of physical retail locations. The device processes the video data and determines actions of the plurality of customers in relation to the product in the plurality of physical retail locations, by: determining, for a customer of the plurality of customers, an activity of the customer in relation to the product, and determining, for the customer, an eye gaze direction of the customer relative to the product. The device determines categories for the actions of the plurality of customers in relation to the product. The device generates a report that indicates one or more of the categories for the actions of the plurality of customers in relation to the product and sends the report to a client device to permit the client device to display the report.