Open search

Edge AI computer

For visual automotive applications

The NRU-110V computer by Neousys (Taiwan) supports eight automotive cameras and one isolated CAN interface for in-vehicle communication.

The solution includes on-board GMSL deserializers, and a pre-installed board support package for selected cameras (Source: Neousys)

The device integrates the Nvidia Jetson AGX Xavier system-on-module comprising the Volta GPU (graphic processing unit) and the Carmel CPU (central processing unit). By incorporating a 10-Gbit/s Ethernet interface, the computer can act as a camera sensor hub by streaming real-time images to another similar computer. In addition to the CAN interface, the device supports one EIA-232 port for GPS (global positioning system) input and isolated digital I/Os (3/4) for connection of sensors and actuators. The GPS pulse-per-second (PPS) input supports time synchronization among multiple platforms. Three USB interfaces and one mini PCIe socket for the Wifi or the 4 G mobile network module are offered. For data storage, a 32-GiB eMMC (on the Xavier module) and an M.2 2280 NVMe socket for SSD (solid state drive) read/write performance are available.

The video input of the device supports commercial off-the-shelf (COTS) automotive cameras. It features the water-proof ability, high dynamic range, auto white balance, and LED flickering mitigation. Thus, the image quality does not depend on the lighting conditions, from rainy or bright sunny days to pitch-black nights. The computer requires a power supply of 8 VDC to 35 VDC and operates at temperatures from -25 °C to +70 °C. The design with 30 W of power consumption is suited to extend operation time for battery-powered machines, such as autonomous mobile robots (AMR). The PC is dedicated for AI-based vision applications that require continuous interaction with surroundings, such as AMR, ADAS (advanced driver assistance systems), V2X (vehicle-to-any communication), etc.


Publish date