Sensor-Fused Shelf Frame for Cashierless Grab-and-Go Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing commercial refrigerators and cabinets lack efficient systems for tracking physical interactions with items, managing inventory, and providing real-time transaction management, especially in environments where items are taken and replenished without a cashier.

Innovation Solution

A shelf system equipped with cameras and weight sensors that communicate with a local computer, utilizing AI to detect transactions, manage inventory, and provide dynamic pricing and marketing strategies, while supporting human behavior analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If cameras and weight sensors are integrated into the shelf system to detect physical interactions, then transaction detection accuracy is improved, but device complexity increases

Engineering Contradiction:

Improvetransaction detection accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent combines multiple sensing modalities (cameras and weight sensors) into an integrated shelf system, merging their functions to detect physical interactions. This integration allows the system to cross-validate data from different sensors, improving transaction detection accuracy while managing complexity through unified system architecture.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The shelf system is designed with multi-functional capabilities, serving both as a display structure and as a detection system. The cameras and weight sensors are embedded within the shelf framework, allowing the same structure to fulfill multiple functions: supporting products, detecting weight changes, and capturing visual data for transaction analysis.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Extent of automation

If AI and machine learning are used to analyze physical interactions, then transaction detection capability is improved, but computing resources and energy consumption increase

Engineering Contradiction:

Improvetransaction detection capabilityVSAvoidenergy consumption

Core Design Contradiction:

Extent of automationVSUse of energy by moving object

Solution Approach 1:

The system pre-processes and filters sensor data before applying complex AI algorithms. By performing preliminary analysis to identify potential transactions and filter out non-relevant movements, the system reduces the computational burden on AI models, thereby lowering energy consumption while maintaining high automation capability.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The AI system applies machine learning algorithms selectively rather than continuously analyzing all sensor data. By focusing computational resources only on situations that likely represent transactions (based on preliminary filtering), the system achieves high automation capability while avoiding excessive energy consumption from processing all possible data streams.

Inventive Principle:

Principle #16Partial or excessive action

3Measurement precision

If real-time monitoring of all shelf interactions is implemented, then inventory management accuracy is improved, but data processing time and system complexity increase

Engineering Contradiction:

Improveinventory management accuracyVSAvoiddata processing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system extracts and focuses only on the critical data elements needed for inventory management, separating relevant transaction information from unnecessary sensor data. By filtering out redundant information and concentrating on key metrics such as weight changes and product removal events, the system maintains high inventory accuracy while reducing data processing time.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system implements real-time feedback loops where sensor data is continuously monitored and immediately processed to update inventory status. This feedback mechanism allows the system to maintain accurate inventory records by promptly detecting and recording transactions, reducing the lag between physical interactions and digital inventory updates.

Inventive Principle:

Principle #23Feedback

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

Enables automatic transaction detection, inventory management, and dynamic pricing, enhancing operational efficiency and customer interaction through AI-driven data analysis.

Implementation Method 1

The frame includes a weight sensor

Methodology Applied

Scientific EffectWeight sensing:

Implementation Method 2

The frame includes at least one camera

Methodology Applied

Scientific EffectImage capture: Photography

Data Source

PatentUS12567048B2Hardware system for identifying grab- and-go transactions in a cashierless store

Publication Date: 2026.03.03 KOOICK INC

AI summary

A method and system for detecting a commercial transaction through physical interactions with products, the system comprising a frame that is placed within a product display unit. The frame includes a shelf, at least one camera, a weight sensor, and display screens. The system further comprises a computing device communicatively connected to the frame. The computing device is configured to detect physical interactions with items that are placed on the shelf based on data that is received from the at least one camera and the weight sensor. The computing device is further configured to determine a commercial transaction associated with physical interactions and transmit item and price data to the display screens based on machine learning of the physical interactions with items.