Skip to main content Skip to search Skip to main navigation

AIB-2941M Edge AI Controller, Expandable AI Accelerator, Metal Case

Product information "AIB-2941M Edge AI Controller, Expandable AI Accelerator, Metal Case"

The AIB-2941M CR is a fanless edge AI controller built around a quad-core Cortex-A72 processor. It suits engineers who want local AI inference and edge computing in an industrial enclosure, with the option to add an accelerator module later. This variant is supplied without an accelerator, so it is the right choice when you already have a module or plan to source one separately.

The unit provides a B-key M.2 socket with USB and an M-key M.2 socket with PCIe for accelerator modules or SSD storage. Communication options include two Gigabit Ethernet ports, two isolated RS-485 ports and four USB ports, plus an HDMI output for 4K display. Two isolated digital inputs and two isolated digital outputs cover basic discrete I/O, and an I/O expansion slot accepts optional XV, XB or eXV boards.

The housing is metal and the unit mounts on a DIN rail or a wall. It operates from -20 to +70 °C and accepts a +12 to +48 VDC supply. Ubuntu OS, along with C and Python libraries, is provided for application development.

Applications

  • Edge AI inference in industrial automation
  • Machine vision and video analysis at the edge
  • Industrial IoT gateways with serial device connectivity
  • Local data acquisition and control with discrete I/O
Product type: Edge AI controller
Interface: Ethernet, HDMI, RS-485, USB
Digital inputs: 2
Digital outputs: 2
Ports: 2 Ethernet, 2 RS-485, 4 USB
Isolation: 2500 VDC (RS-485)
Power input: +12 ~ +48 VDC
Operating temperature: -20 ~ +70 °C
Mounting: DIN-rail / wall
Housing: Metal

Specifications

Software
OSUbuntu OS
Multilanguage SupportEnglish, Simplified Chinese, Traditional Chinese
SDKC / Python Library
Main Unit
CPUQuad-core Cortex-A72 (ARM v8) 64-bit SoC @ 1.5GHz
64-bit Hardware Serial NumberYes
System Memory4GB LPDDR4 SDRAM
StorageOn-board 32GB eMMC
Non-Volatile Memory16 KB FRAM
Real Time ClockProvide second, minute, hour, date, day of week, month, year
Watchdog TimerYes
AI Accelerator (Optional)
Model NumberHailo 8 Lite
AI frameworksTensorFlow, TensorFlow Lite, Keras, PyTorch & ONNX
TOPS13
RuntimeHailo RT Rev. 4.19, pre-installed
Power Consumption1.5 W
Display
GPUVideoCoreVI (OpenGL ES 3.0), H.265(up to 4Kp60fps decode), H.264 (up to 1080p60fps, 1080p30 encode)
SignalHDMI port x 1 (4Kp 60fps with Audio)
LED Indicators
Status6 LED indicators for Power and Running; 3 LED for user defined
Digital Input
Channels2
TypeWet Contact
Sink/Source (NPN/PNP)Sink/Source
ON Voltage Level+19 VDC ~ +24 VDC
OFF Voltage Level+11 VDC Max.
Digital Output
Channels2
TypeIsolated Open Collector
Load Voltage+24 VDC
Load Current100 mA/Channel
Overvoltage Protection60 VDC
Overload Protection1.1 A
Short-circuit ProtectionYes
I/O Expansion
Slots1x XV-board/XB-board/eXV-board (Optional)
1x M.2 key B
2x M.2 key M
Note: the three M.2 slots share one PCIe-2.0x1
Audio
JackMic-In x 1 and Line-Out x 1 (Optional)
Ethernet
Ports2 x RJ-45, 10/100/1000 Based-TX (Auto-negotiating, Auto MDI/MDI-X, LED indicators)
COM Ports
Ports2 x RS-485 (2500 VDC Isolated)
USB
Ports2 x USB 2.0 (host),2 x USB 3.0 (host)
Power
Input Range+12 ~ +48 VDC
Consumption4.8 W (0.2 A @ 24 VDC)
Mechanical
CasingMetal
Dimensions (mm)52 mm x 177 mm x 128 mm
InstallationDIN-Rail / Wall Mounting
Environment
Operating Temperature-20 ~ +70 °C
Storage Temperature-40 ~ +80 °C
Humidity10 ~ 90 % RH,(non-condensing)

Specifications as published by the manufacturer. Subject to change.

0 of 0 reviews

Average rating of 0 out of 5 stars

Leave a review!

Share your experiences with other customers.