AEC200
Atomrock Edge Computing AEC200
The AEC200 series is an AIoT edge gateway built for enterprise video and audio AI at the network edge. It pairs software-defined perception with native AtomCloud connectivity, so video streams, audio anomalies, and operational data flow into one centrally managed loop. Compact, fanless, and AtomOS-native, it is the practical entry point into Atomrock's edge-to-cloud stack.
Video AI at the edge
Audio AI anomaly detection
128-channel processing
AtomCloud-native
Software-Defined Video AI
The AEC200 analyzes live video streams in real time, surfacing patterns, anomalies, and inefficiencies that human operators miss. Facial detection and matching run on-device, with photo search across uploaded and stored profiles. Up to 128 channels of multi-stream processing on a single unit, with HD/FHD at 24×10fps and UHD at 6×10fps.
Powered by an ARM Cortex-A72/A53 octa-core CPU, Mali-G52 MC3 GPU, and a dedicated 6 TOPS NPU, all in a fanless metal chassis. Custom video models can be loaded for scenario-specific inference, alongside Atomrock's standard VisionSense pipelines.
Audio AI for Predictive Maintenance
Atomrock's Audio AI runs continuously on the AEC200, listening for the subtle deviations that precede equipment failure, intrusion, or process drift. Custom audio models can be deployed alongside video models on the same device, giving operators a multi-modal early-warning layer without adding hardware.
Real-time inference on the device, anomalies and snapshots streamed to AtomCloud for retention and review. The same multi-channel architecture handles audio inputs, so a single AEC200 can cover dozens of monitored points across a facility.
AtomCloud Native, Quick Onboarding
Built on AtomOS 4.0+ with native AtomCloud connectivity, the AEC200 lets enterprise users adopt AIoT without a custom integration project. Cloud recording with 7 to 366 days of retention, AI event logs with one-year retention, and a 1 GB on-device buffer to keep AI analysis flowing through cloud disconnects.
Compact fanless aluminum chassis at 188 × 88 × 51 mm and 790 g. Dual Gigabit LAN, WiFi 6, Bluetooth 5.2, and 4G via SIM, with HDMI 2.1, USB 3.0, USB-C, and a Phoenix connector exposing RS232 / RS485 / CAN 2.0. Wide -20°C to 60°C operating range; CE / FCC / RoHS certified, ISO 9001 manufactured.
Application Scenarios
Enterprise AIoT at the edge: video and audio AI across retail, facilities, transit, and industrial operations.
Facial Recognition Access
Enterprise access control with on-device facial detection and matching against managed profiles.
Retail & Branch Analytics
Footfall, dwell, and incident detection across retail floors and branch networks.
Facility Safety
Multi-channel video monitoring across offices, campuses, and managed buildings.
Audio Anomaly Detection
Continuous sound monitoring for early failure indicators and unusual events.
Predictive Maintenance
Audio AI catches drifts in machinery sound signatures before failure events.
Logistics & Warehousing
Loading-zone monitoring, package tracking, and operational anomaly alerts.
Transit & Mobility
Stations, depots, and mobile assets with real-time perception and cloud archival.
Education & Public Spaces
Campus and public-facility safety with multi-channel video and event-driven alerts.
AIoT Pilot & Rollout
Quick AtomCloud onboarding for enterprise teams piloting AI at the edge.
Specifications
| SYSTEM | |
|---|---|
| Operating System | AtomOS 4.0+ |
| Internal Flash | 64 GB / 256 GB (optional) |
| Memory | 8 GB |
| CPU | Quad-Core Cortex-A72 + Quad-Core Cortex-A53 |
| GPU | ARM Mali-G52 MC3 |
| NPU | 6 TOPS computing power |
| Ethernet | 1000M Ethernet (×2 RJ45) |
| WiFi | 2.4 GHz / 5 GHz, 802.11 a/b/g/n/ac/ax (WiFi 6) |
| Bluetooth | BT 5.2 |
| AI FUNCTIONS | |
|---|---|
| Camera Stream | Video streaming: HD / FHD 24×10fps · UHD 6×10fps |
| Facial AI | Facial detection, facial matching, real-time identification · Photo search across uploaded / stored profiles |
| Advanced AI | Audio AI, Video AI · Supports customized audio / video models |
| AUDIO / VIDEO | |
|---|---|
| Cloud Disconnect Buffer | 1 GB AI analysis data |
| Multi-channel Processing | 128 channels |
| Cloud Recording | 7 to 366 days cloud data retention |
| Cloud Recording Archive | 1 to 15 years cloud data retention |
| AI Events | Data, Snapshots · 1 year cloud data retention |
| CERTIFICATION | |
|---|---|
| Hardware | CE · FCC · RoHS |
| Manufacturing | ISO 9001 |
| CONNECTIONS | |
|---|---|
| USB | 2 × USB 3.0 · 1 × USB Type-C |
| Serial | 1 × RS232 · 1 × RS485 · 1 × CAN Bus 2.0 |
| HDMI | HDMI 2.1 (4K @ 60fps) |
| Network | 2 × RJ45 · WiFi · 4G |
| Storage | 1 × TF Card Slot · 1 × SIM Card Slot |
| Audio | 3.5 mm Audio Jack |
| Power | DC 12V / 2A input |
| Indicator | LED — System Status |
| ELECTRICAL | |
|---|---|
| Power Supply | DC 12V / 2A |
| Operating Temp. | -20°C to 60°C (-4°F to 140°F) |
| Operating Humidity | 10% to 90% (non-condensing) |
| Storage Temp. | -20°C to 70°C (-4°F to 158°F) |
| Storage Humidity | 10% to 90% (non-condensing) |
| MECHANICAL | |
|---|---|
| Color / Material | Black / Metal |
| Size | 188 × 88.44 × 50.65 mm (7.40" × 3.48" × 1.99") |
| Weight | 790 g ± 10 g (net) |