Engineering FAQ · Wanlin
Computer Vision Locker Motherboard Manufacturer | Brazil Export
A: Yes. The 6 TOPS NPU runs YOLO-class detection frameworks directly on the board, so recognition keeps working with no internet connection and without per-transaction cloud vision fees.
Direct answer
A: Yes. The 6 TOPS NPU runs YOLO-class detection frameworks directly on the board, so recognition keeps working with no internet connection and without per-transaction cloud vision fees.
Executive Summary: Computer Vision Locker Motherboard for Brazil Market
Product Introduction: AndroidSBC Computer Vision Locker Motherboard
Core Features:
- Lock Control and Connectivity: RS485 Modbus lock buses, door-status feedback, Gigabit Ethernet, WiFi and 4G LTE options
- Display Flexibility: LVDS, eDP, HDMI and MIPI-DSI outputs from 7 to 32 inch panels, portrait or landscape
Technical Specifications
| Parameter | AndroidSBC Specification |
|---|---|
| CPU | Rockchip RK3588, octa-core (4x Cortex-A76 + 4x Cortex-A55) @ 2.4 GHz, 8nm |
| NPU | 6 TOPS, on-device object and product recognition |
| GPU | Mali-G610 MP4, 8K@30fps decode |
| Cameras | 4x MIPI-CSI channels, 8MP/13MP ISP support |
| Memory | 4GB / 8GB LPDDR4X |
| Storage | 32GB / 64GB eMMC, M.2 PCIe SSD slot |
| Vision | YOLO-class detection frameworks supported in BSP |
| Sensing | Weight sensor and door sensor interfaces for stock verification |
| Network | Dual Gigabit Ethernet, WiFi 6, 4G/5G module support |
| OS | Android 13 (Android 12-14 range, Linux available) |
Application Scenarios for Brazil Market
| Scenario | Recommended model | Key benefits |
|---|---|---|
| Smart vending coolers | AS-AI3588C (4 cameras) | Vision-based grab-and-go fridge |
| Unattended retail cabinets | AS-AI3588R (weight+vision) | Automatic checkout and stock counts |
| Fresh food cabinets | AS-AI3588F (fresh SKU) | Recognition for packaged fresh items |
| Equipment lending | AS-AI3588E (tool ID) | Tool and equipment tracking by vision |
Q: Can the board run product recognition without the cloud?
A: Yes. The 6 TOPS NPU runs YOLO-class detection frameworks directly on the board, so recognition keeps working with no internet connection and without per-transaction cloud vision fees.
Q: How many cameras can be connected?
A: Four MIPI-CSI camera channels with 8MP and 13MP ISP support cover multi-shelf cabinets, and USB cameras can extend coverage for larger units.
Q: Does it support weight sensing as well?
A: Yes. Weight sensor interfaces combine with vision for stock verification, which improves accuracy for items that are hard to distinguish visually.
Q: What development support is provided?
A: The BSP includes NPU toolchains and model conversion documentation, demo recognition apps, and our engineers support model training and deployment for partner SKUs.
Q: Which companies is this board best suited for?
A: Vending operators upgrading to smart coolers, unattended retail brands, and OEM cabinet makers building grab-and-go hardware for grocery and convenience formats.
Project Summary
For more information: www.androidsbc.com | Export Service Hotline: +8613261677119 | Email: Androidsbc@163.com
- Shenzhen HQ: Wanlin Group, Building B, Building 1, Beisida Medical Device Building, 28 Nantong Avenue, Baolong Community, Baolong Sub-district, Longgang District, Shenzhen, Guangdong, China
Application fit
Confirm the operating environment, workload, display, network and peripheral requirements.
Documents to confirm
Check the exact board revision, interfaces, firmware build and certification scope.
Commercial confirmation
Verify quantity, sample plan, lead time, warranty and delivery terms in writing.
