Edge Computing Gateway

AIHub-Z3 Edge Computing Box

A compact RK3566 edge box with ARM G52 GPU, 1TOPS NPU, Bluetooth, Wi-Fi, Ethernet, and ZigBee expansion for smart buildings, stores, restaurants, energy management, and lightweight edge intelligence.

RK3566 quad-core platform1TOPS NPUMulti-protocol device linkage
AIHub-Z3 edge computing box
Product Context

A local gateway and edge intelligence node for lightweight AIoT sites

AIHub-Z3 connects sensors, controllers, cameras, thermostats, and smart terminals, then runs local rules, light AI decisions, cache, and platform synchronization. It is best when many small sites need repeatable installation and remote maintenance.

  • Smart buildings
  • Store automation
  • Energy management
  • Lightweight security and sensing
1TOPSLocal AI inference
104mmCompact deployment
ZigBeeExpansion option
Compact AIoT edge node installed in a facility equipment room
Compact Edge Gateway

Z3 fits lightweight, distributed, always-on AIoT field nodes

AIHub-Z3 emphasizes small size, low power, and multi-protocol access. It can run local rules, edge cache, lightweight AI, and platform synchronization without placing a high-performance server at every site.

  • Fits equipment cabinets, stores, and compact field locations
  • Supports local rules, edge cache, and light inference
  • Pairs with serial terminals, thermostats, and sensors
Deployment View

What this product looks like in a real workflow

AIHub-Z3 building automation deployment sceneAIHub-Z3

Small field edge gateway

Z3 acts as the local aggregation point for communication, rules, caching, and cloud sync.

Low powerSmall footprintMulti-protocol
AIHub-Z3 port and interface detail

Interface and installation fit

Deployment is designed around field points, power, network, and maintenance cost.

Field installRemote upgradeBatch config
Core Capabilities

Product capabilities that matter in deployment

Each capability is designed around real operating conditions, not isolated hardware parameters.

RK3566 low-power platform

Balances gateway tasks, basic compute, and continuous operation.

1TOPS local AI

Supports lightweight vision, tracking, and event analysis.

Multi-protocol access

Bluetooth, Wi-Fi, Ethernet, and ZigBee expansion connect mixed field devices.

Remote maintenance

Applications and configuration can be maintained for batch field operations.

System Architecture

From field data to a managed product workflow

01

Device communication

Connect sensors, actuators, cameras, thermostats, and smart terminals.

02

Local intelligence

Run light AI, rules, and automation for fast field response.

03

Edge cache

Cache important events when the network is unstable.

04

Cloud collaboration

Sync device data, alerts, and configuration with ZedIoT or customer platforms.

Product Details

Product configuration, capabilities, and deployment details

These sections collect the practical information customers need for selection, integration, installation planning, and rollout evaluation.

Product Capability Notes

AIHub-Z3 is a compact AIoT edge box based on RK3566, ARM G52 GPU, and 1TOPS NPU. It supports Bluetooth, Wi-Fi, Ethernet, and ZigBee expansion for lightweight edge intelligence in stores, buildings, restaurants, energy management, and smart environments.

RK35661TOPS NPUBluetooth / Wi-Fi / EthernetZigBee expansion
AIHub-Z3 compact edge gateway product angle

Product Advantages

Low-power always-on node

Good fit for sites that require continuous operation with low maintenance cost.

Lightweight edge intelligence

Runs local rules, light AI inference, edge cache, and device linkage.

Multi-protocol access

Works with wireless, wired, and ZigBee expansion for thermostats, serial terminals, sensors, and actuators.

Use Scenarios

Where this product direction is a good fit

Building automation

Connect sensors and actuators for local rules and device linkage.

Security sensing

Connect cameras, door sensors, and alarm equipment for local event handling.

Energy management

Collect power, temperature, and device-state data for optimization.

Store and facility nodes

Use as a local gateway for many small repeatable sites.

Specifications

Define the business goal before finalizing the hardware list

These items help teams turn a product idea into an engineering brief that can be estimated and prototyped.

ChipsetRK3566Quad-core 64-bit edge computing platform.
AI compute1TOPS NPUFor light visual inference and local event analysis.
GPUARM G52 2EESupports basic graphics and media processing.
Storage options4G+32G / 8G+64G / 8G+128GChosen by application load.
ConnectivityBluetooth / Wi-Fi / Ethernet / ZigBee expansionCovers common AIoT device access paths.
Size104×104×17mm classFits wall, cabinet, and compact installation.
Delivery

How we turn this into a deployable product

Scene and point planning

Confirm device types, communication distance, placement, power, and network.

System image and app customization

Deploy edge apps, communication plugins, rules, and platform client.

Cloud-edge integration

Tune upload, remote configuration, logs, and exception alerts.

Batch operations

Deliver config templates, installation guide, upgrade strategy, and maintenance workflow.

Talk to ZedIoT

Build this product direction with ZedIoT

Tell us your target device, equipment scenario, current technical constraints, and the business workflow you want to support.

  • AI + IoT product architecture review
  • Hardware, firmware, cloud, and application integration
  • Prototype planning and production support