NORVI GSM-AE08-I-G

https://www.youtube.com/shorts/EAUiRhkJjYg

NORVI GSM-AE08-I-G: W5500 Industrial Connectivity Meets NORVI Edge AI

Industrial automation is rapidly moving beyond simple remote monitoring and control. Today, industrial devices are expected not only to collect sensor data and connect machines, but increasingly to support edge computing, machine vision, and AI-driven decision making.

NORVI provides an interesting example of this transition.

Its NORVI GSM-AE08-I-G is an industrial programmable controller built around the ESP32 and WIZnet W5500 Ethernet controller, combining wired Ethernet, cellular connectivity, RS-485, and industrial analog inputs in a DIN-rail form factor. At the same time, NORVI is expanding its portfolio toward embedded vision and Edge AI through the NORVI AI Optic, an ESP32-S3-based industrial camera platform.

Together, these products illustrate how traditional industrial connectivity and emerging Edge AI can coexist within the same automation ecosystem.

NORVI GSM-AE08-I-G: An Industrial Controller with Multiple Connectivity Options

The NORVI GSM-AE08-I-G is designed for industrial monitoring and remote-control applications where several communication methods may be required simultaneously.

At its core is an ESP32-WROOM32 MCU combined with a WIZnet W5500 for 10/100 Mbps Ethernet connectivity. The controller also includes GSM/GPRS, Wi-Fi, Bluetooth, and RS-485 interfaces.

Key specifications include:

ESP32-WROOM32

WIZnet W5500 Ethernet

10Base-T / 100Base-TX

GSM/GPRS using SIM800C

Wi-Fi and Bluetooth

RS-485

5 digital inputs

4 × 4–20 mA analog inputs

ADS1115 ADC

DS3231 RTC with battery backup

microSD card

0.96-inch OLED display

24 V DC power input

DIN-rail mounting

NORVI documents the W5500 as the Ethernet transceiver and exposes its SPI interface directly to the ESP32. The W5500 provides eight independent hardware sockets and supports TCP, UDP, ICMP, IPv4, ARP, IGMP, and PPPoE.

Why W5500 Matters in This Architecture

The GSM-AE08-I-G is particularly interesting because Ethernet is not its only network interface.

The controller already has:

Wi-Fi

Bluetooth

GSM/GPRS

RS-485

Yet NORVI still includes dedicated W5500 wired Ethernet.

This gives system integrators several possible communication paths depending on the installation.

For example:

4–20 mA Sensor
→ NORVI GSM-AE08-I-G
→ W5500 Ethernet
→ Local SCADA / Server

For a remote site:

Industrial Sensor
→ NORVI GSM-AE08-I-G
→ GSM/GPRS
→ Remote Server

And for legacy industrial equipment:

Modbus / RS-485 Device
→ NORVI Controller
→ W5500 Ethernet
→ Ethernet Network

The result is a controller that can bridge traditional field instrumentation with both local Ethernet infrastructure and wide-area cellular networks.

This is particularly useful for applications such as:

remote equipment monitoring

water and wastewater systems

environmental monitoring

energy monitoring

factory equipment

distributed sensor systems

unattended industrial installations

The industrial nature of the device is also reflected in NORVI's documented compliance with standards including EN 61131-2 and EN 61010 series requirements.

From Industrial Connectivity to Edge AI

NORVI's more recent product portfolio shows another direction: embedded vision and Edge AI.

The NORVI AI Optic is an ESP32-S3-based camera platform designed for embedded vision, machine monitoring, and TinyML development. NORVI describes it as an edge camera that can capture images, process them locally, display results, store data, and react to external trigger events.

Its hardware includes:

ESP32-S3-WROOM-1U-N16R8

16 MB flash

8 MB PSRAM

OV5640 5 MP autofocus camera

local TFT display option

microSD storage

external dry-contact trigger

dual LED illumination

USB-C programming and charging

rechargeable backup battery

This makes AI Optic very different from a conventional IP camera.

Instead of simply streaming images to a remote server, the ESP32-S3 can perform lightweight image processing or compact machine-learning inference close to the machine.

NORVI's Edge AI Workflow

NORVI documents a practical TinyML workflow using Edge Impulse.

The process is:

Capture training images using the deployed AI Optic hardware.

Build a dataset from real application images.

Train a compact vision model in Edge Impulse.

Export the trained model as an Arduino library.

Deploy the model back to the ESP32-S3.

Run classification locally on the device.

This allows AI inference to take place close to the equipment without requiring every image to be sent to a cloud AI service.

NORVI proposes applications including:

visual inspection

object detection

barcode and label workflows

meter reading

people counting

equipment monitoring

time-lapse process monitoring

For example:

External Machine Trigger
→ AI Optic Camera Capture
→ ESP32-S3 Local AI Inference
→ Classification Result

The result can then become part of a larger industrial automation system.

Connecting Industrial I/O and AI

The interesting opportunity appears when NORVI's two product directions are viewed together.

The GSM-AE08-I-G represents the conventional industrial automation layer:

Sensors / Equipment
→ 4–20 mA / Digital Input / RS-485
→ ESP32
→ WIZnet W5500
→ Industrial Ethernet

The AI Optic represents a newer machine-vision layer:

Camera / Trigger
→ ESP32-S3
→ Edge AI
→ Visual Decision

Together, they illustrate an architecture such as:

Industrial Sensors + Machine Vision
↓
NORVI Industrial Controllers / AI Optic
↓
Local Processing and Control
↓
Ethernet / Cellular Network
↓
SCADA / Server / Industrial Application

This is an important direction for industrial IoT.

AI does not eliminate traditional industrial connectivity. Instead, AI creates even greater demand for reliable field data and dependable communication between sensors, controllers, machines, and higher-level systems.

In that architecture, WIZnet W5500 provides a reliable wired Ethernet path for NORVI's industrial controller, while NORVI AI Optic introduces local vision intelligence at the edge.

From Connected Machines to Intelligent Machines

NORVI's portfolio shows how industrial automation is evolving.

The first step is connectivity:

Sensor → Controller → Ethernet

The next step adds intelligence:

Sensor / Camera → Edge Processing → Decision → Industrial Network

For WIZnet, this evolution is particularly relevant.

W5500 does not need to perform the AI inference itself. Its role is equally fundamental: providing the network path that connects field devices, controllers, SCADA systems, gateways, and AI-enabled equipment.

A useful way to describe this relationship is:

Industrial AI starts with reliable field data, and reliable field data needs dependable connectivity.

The NORVI GSM-AE08-I-G demonstrates the connectivity side of this architecture with ESP32, 4–20 mA inputs, RS-485, cellular communications, and WIZnet W5500 Ethernet.

NORVI AI Optic shows where the same industrial ecosystem is heading next: toward embedded vision, TinyML, and local machine intelligence.

Together, they provide a useful example of the transition from connected industrial devices to intelligent edge automation.

Product Status Note

NORVI currently lists the AI Optic technical documentation and Edge AI workflow publicly, but states that new orders for AI Optic are temporarily paused. Its documentation, object-detection workflow, and technical information remain available for engineering evaluation.

References

NORVI GSM-AE08-I-G official datasheet

NORVI AI Optic official product page and Edge AI workflow

NORVI AI Optic official datasheet

NORVI industrial product portfolio