Aetina Ships Jetson Orin In-Vehicle Edge AI Box
Aetina Corporation has made its NVIDIA Jetson AGX Orin powered DeviceEdge AIE-VX28/18-IP available in mass production for in-vehicle video analytics.
Updated on Nov. 5, 2026 in Robotics
Aetina Corporation has made the NVIDIA Jetson AGX Orin powered DeviceEdge AIE-VX28/18-IP available in mass production. The fanless, IP66-rated in-vehicle edge AI system is designed for aftermarket video analytics and supports up to 12 camera inputs.
Why it matters
The system's rugged design, with MIL-STD-810H durability and a -25°C to 70°C operating range, targets vehicles that need AI video analysis outside factory-installed systems. Its availability in mass production gives fleet and aftermarket integrators a ready hardware option.
The DeviceEdge AIE-VX28/18-IP supports 12 camera inputs, with four PoE cameras and eight GMSL2 cameras running simultaneously, and delivers up to 275 TOPS of AI performance. Its integrated M12 PoE ports supply a combined 60W for IP cameras, though the source provides no independent benchmark results.
The players
Aetina Corporation
Aetina Corporation is the company that developed and is bringing the DeviceEdge AIE-VX28/18-IP into mass production.
NVIDIA
NVIDIA is the chipmaker whose Jetson AGX Orin modules power the DeviceEdge AIE-VX28/18-IP.
The details
The system uses NVIDIA Jetson AGX Orin modules and includes M12 10GbE, isolated CAN FD, and RS-232/422/485 interfaces. GMSL2 cameras connect through mini-FAKRA connectors, while the fanless design carries an IP66 ingress rating.
Timeline
Mass-production availability was announced in an article dated November 5, 2026.
Under the Hood
The DeviceEdge AIE-VX28/18-IP is built on NVIDIA Jetson AGX Orin modules, extending that platform into a ruggedized, fanless in-vehicle form factor. Its M12 10GbE, isolated CAN FD, and multi-protocol serial interfaces show how edge AI hardware is being adapted for vehicle networks.
Fleet operators and aftermarket integrators can now source a rugged, IP66-rated system for running multi-camera video analytics in vehicles. The device's 9-48 VDC input and ignition power control are designed to fit standard vehicle electrical setups.
The takeaway
Edge AI hardware is moving out of the lab and into rugged, vehicle-ready packages. Organizations evaluating in-vehicle video analytics should compare thermal, power, and camera-interface requirements against a system like this before committing to a platform.
Further reading
For more on the hardware behind systems like this, explore Robotics coverage.
Source note: This article includes information reported by Embedded Computing Design.
