．uilt-in Cortex-A72 octa-core processor with a 6 TOPS NPU — ample compute for real-time on-site AI video recognition ．AI inference runs locally in real time, unaffected by cloud network latency ．Web-based control console; easy to review alarm logs and set recognition parameters ．Connects to existing IP cameras (RTSP stream); no need to replace on-site cameras ．Video is processed on-site, lowering bandwidth demand while safeguarding data privacy ．Industrial-grade hardware, supporting 24-hour continuous operation
- Application areas -
Construction sites, quarries, parking lots, road traffic, plant entrances, and aquaculture farms
Download
Center
Optional Code & Specifications
AIM Code1
| Code1 |
Description |
|---|---|
| N | Bare board (no enclosure; integrable into an IEC enclosure) |
| 175 | PCW175 waterproof enclosure, 252 × 175 × 60 mm (W×H×D, same as CB enclosure) |
| 300 | Waterproof enclosure 300, 300 × 300 × 180 mm (W×H×D, ref. IEC-H enclosure) |
| Power Input | C 12V / 2A (accepts DC 9~26V) |
|---|---|
| Power Consumption | Approx. 10W (varies with AI load) |
| Operating Temp. | -20°C ~ 60°C |
| Processor | Quad Cortex-A72 + Quad Cortex-A53, ≤2.2GHz |
|
AI Compute Unit |
Dedicated NPU, 6 TOPS; INT4/8/16, FP16, BF16 mixed |
| Memory | DDR4 8GB |
| Storage | eMMC 64GB (64G USB included) |
| OS | Dedicated Linux OS |
| Video Input | IP cameras via network (RTSP); built-in HW decoding |
| Video Output | Recognition results streamed via RTMP; viewable with JNC WEE |
| Network Interface | 10/100/1000M auto-sensing Gigabit Ethernet ×1 |
| Recording/Snapshot | To external USB for recording and snapshot storage |
ACHIEVEMENT