Sections: Case studies | Applied Filters | How Edge AI Cameras are changing the future of smart retail Home > ST Edge AI Suite > Case studies Case studies Discover how to improve embedded solutions using edge AI. Applied Filters Clear All You searched: {searchParam} In smart retail and warehouse management, real-time visibility of inventory levels inside refrigerators and on shelves is critical for operational efficiency. CamThink's NE301 edge AI camera, powered by the STM32N6 microcontroller, brings on-device beverage detection and counting directly to the point of need. This eliminates cloud dependency, slashes communication costs by over 99%, and keeps visual data private at the edge.

Application domains:

Smart appliances (refrigerators)
Smart retail (vending machines)
Warehouse management

Approach

Traditional inventory monitoring in refrigerators and retail shelves relies on manual inspection, weight-sensor systems, or cloud-based image recognition. Each approach carries significant limitations: manual checks are labor-intensive and prone to delayed restocking, weight sensors require costly hardware modifications and are sensitive to product placement, and cloud-based solutions demand high bandwidth, introduce latency, and raise privacy concerns. Conventional MCUs also lack the computing power to run accurate detection models locally, leading to high false alarm rates.

CamThink addressed these challenges by deploying the NE301, built on ST's STM32N6, to perform AI-based product recognition entirely on-device. When inventory changes occur, only minimal structured data (e.g., "Coke: 3 cans") is transmitted instead of images. This fundamentally changes the economics and privacy profile of smart inventory systems.

Key benefits of this edge AI approach:

Dramatic traffic reduction: Over 99% less data transmitted compared to image-upload solutions, reducing cloud costs by up to 90%
Flexible power options: Supports battery, PoE, and Type-C power, enabling deployment in diverse environments
Privacy by design: Visual data never leaves the device; only structured results are shared
Lower operational costs: Eliminates manual inspections and reduces restocking delays
Real-time awareness: Sub-50 ms inference delivers instant inventory visibility without cloud round-trips

By harnessing edge AI on the STM32N6, CamThink demonstrates how intelligent, scalable, and cost-effective inventory monitoring can be achieved in modern retail and appliance environments.

Application overview

The NE301 runs a complete detection pipeline locally, from image capture to structured result output, with no cloud processing required.

1. Image acquisition

The NE301 is periodically awakened on schedule and captures images of the refrigerator interior through its wide-angle CMOS sensor. The integrated ISP performs noise reduction and automatic exposure adjustment to ensure clear input even under complex lighting conditions such as reflections and low-light environments.

2. Edge inference

The built-in Neural-ART NPU runs a YOLOv8 Nano model to detect and count beverages directly on the device. Inference completes in under 50 ms, enabling real-time inventory awareness without uploading any images.

3. Decision and reporting

Only the recognition results, in JSON format, are transmitted to the backend system via MQTT (e.g., Home Assistant or a custom server). This keeps bandwidth usage minimal and ensures immediate actionability for restocking or alerts.

Technical details

CamThink's solution fully leverages the heterogeneous architecture of the STM32N6:

Neural-ART NPU — Accelerates YOLO model inference to under 50 ms, enabling real-time inventory detection without uploading images to any server.
CamThink AI Toolstack — Provides a complete end-to-end toolchain from data collection and labeling to training and quantized deployment (.bin). Developers can build customized product recognition models within hours, even without deep AI expertise.
Ultra-low power architecture — Combines the STM32N6 edge AI capability with an STM32U0 co-processor for power management and sleep control. The device wakes only when needed, dramatically reducing power consumption and deployment costs.

Sensor

CMOS Image Sensor: OS04C10, HFOV 137°

The NE301 performs image preprocessing through its ISP (including noise reduction and automatic exposure), ensuring that the AI model receives clear and reliable input even under complex lighting conditions inside refrigerators, such as reflections and low-light environments.

Dataset and model

Model:

YOLOv8 Nano (object detection)
Input size: 320 x 320
Optimized for beverage product recognition
Quantized for on-device deployment via CamThink AI Toolstack

Results

Inference time: ~50 ms
Model accuracy: >90% for beverage products (further optimization possible)
Memory usage: Runs entirely in on-chip SRAM, no external RAM required
Communication traffic: >99% reduction compared to image-upload-based solutions

Get started now!

Easily recreate this use-case with the following resources:

Camthink NE301 GitHub repository, including tutorials
YOLOv8n from the ST Model Zoo

Additional resources:

" data-isnew="NEW"> --> Vision | Tutorial | GitHub | Appliances | Smart city How Edge AI Cameras are changing the future of smart retail Detect and count fridge beverages fully on-device for smarter retail inventory, with Camthink's NE301

What it does

Available capabilities

Tags

Ways to use it