Skip to main content

Industrial Edge-AI Camera

The Alp Lab Industrial Edge-AI Camera is a production-ready vision system built on the E1M-X V2N-M1 module, delivering 25 TOPS of on-device inference in a ruggedised enclosure with flexible lens-mount and connectivity options.

At a glance

ParameterValue
Compute moduleE1M-X V2N-M1 (RZ/V2N + DEEPX DX-M1)
AI performance25 TOPS (dense)
Lens mountsC-Mount, CS-Mount, M12
Ingress protectionIP67 (sealed) / IP20 (open)
ConnectivityUSB-C, PoE (IEEE 802.3af/at)
OSYocto Linux
Indicative price$599

Key features

  • 25 TOPS on-device inference for real-time quality inspection, object detection, and classification
  • Multiple lens-mount options — C-Mount, CS-Mount, and M12 — to suit a wide range of industrial optics
  • IP67-rated enclosure for harsh environments (dust, water jets, washdowns); IP20 variant for clean indoor installations
  • Dual connectivity — USB-C for development and local integration; PoE for single-cable deployment with power + Ethernet
  • Full Linux stack with DRP-AI3 and DEEPX M1 accelerator support

Enclosure variants

VariantIngress protectionUse case
IP67Dust-tight, water-jet proofFactory floors, outdoor, washdown areas
IP20Finger-safeClean rooms, indoor inspection stations

Lens compatibility

The camera accepts standard industrial lenses via three interchangeable mount options:

  • C-Mount — 1" flange distance, widely used in machine vision
  • CS-Mount — 12.5 mm flange distance, compact form factor
  • M12 — Board-level mount for miniaturised lens assemblies

Connectivity

USB-C

  • High-speed data transfer for model deployment and log retrieval
  • 5 V power input for bench-top development
  • Serial console access for debugging

Power over Ethernet (PoE)

  • IEEE 802.3af/at compliant
  • Power + Gigabit Ethernet over a single Cat5e/Cat6 cable
  • Ideal for distributed camera networks in factory settings

Software

The Industrial Camera runs the same Linux BSP and AI toolchain as the standalone E1M-X V2N-M1 module — same Yocto image, same <alp/inference.h> dispatcher, same DEEPX userland (libdxrt.so). Pre-built application images are available for common machine-vision use cases.

Getting started

  1. Mount the camera and attach a compatible lens.
  2. Connect via USB-C (development) or PoE (deployment).
  3. Access the camera's web interface or SSH terminal.
  4. Deploy an AI model via the <alp/inference.h> dispatcher or a pre-built application image.

Resources

Questions about this page? Discuss in Community Forum