Acceed
4-camera controller controls autonomous systems
Acceed has announced a series of compact industrial AI platforms called 'NRU-51V+'.
The robust and fanless controllers from Acceed are equipped with Nvidia high-performance Jetson Orin or Xavier modules. The models in the series bring a low-power, fanless AI inference platform to the market. The controller in the robust aluminum chassis has four FAKRA-Z connections for GMSL2 cameras, each of which supports a resolution of 1920 x 1080 pixels at 30 FPS. Either the Jetson Orin NX or the Xavier NX from Nvidia are used as processors. One 10GbE port and one 1GbE port with screw connection are available for the network connection. The device's internal mini PCIe socket can be used for a WiFi or 4G/5G module or various storage options, among other things.
GMSL (Gigabit Multimedia Serial Link) is an asymmetrical full-duplex connection that was primarily developed for video transmission in vehicles. Power, bidirectional control data, Ethernet, bidirectional audio and several unidirectional video streams are simultaneously routed via a single line. The video transmission can be unprotected or encrypted with HDCP. The GMSL connection provides power to the camera, camera control, error messaging and video transmission. The controllers deliver an inference performance of 100 TOPS with a power consumption of 25 W. The raw camera images can be transferred via the 10GbE port to another GPU server for processing or used directly for real-time object or ROI detection. For teleoperation applications, the video codec can be used to encode the image streams from four GMSL2 cameras in real time and transmit the live images to an operator with minimal latency via 5G communication.
You might also be interested in
Does the country need new controllers?
Controller manufacturers are currently under acute pressure to innovate. Among other things, this is due to the increased requirements for cybersecurity and AI capabilities. Software updates to existing controllers are often not enough; instead,...
read more...25 years of Computer&Automation - The e-paper
For 25 years, Computer&Automation has accompanied the growing influence of IT in automation technology. Ultimately, IT and automation can no longer be clearly separated. In this anniversary issue, you can read about the importance that...
read more...The e-paper for the special issue
Sensor & Vision 2023
Intelligent sensor technology always requires smooth interaction with software and surrounding systems. The examples in this e-paper demonstrate what this can look like in individual cases. This shows that team play counts - in sensor technology as...
read more...Products of the year - The e-paper special
"Products of the Year 2023" goes interactive: the new e-paper including video news is here! The winners of the readers' poll have been chosen and in our interactive special edition you will find all the winning products at a glance and exclusive...
read more...Embedded AI agent integrated
Grossenbacher is expanding the 'Universal Controller' to include an embedded AI agent that can be executed directly in a container.
read more...Dual leadership in Shanghai
Stefan Becker has been appointed Managing Director of Eckelmann Automation Technology in Shanghai and will manage the subsidiary together with Zhongwei Huang.
read more...Complete solution for industry-specific control systems
The 'iCube Control' automation platform from Yaskawa is officially launched at the SPS: The first controllers from the control portfolio are available with the 'iC9226' series.
read more...Control multi-layered applications
Wago is launching the 'PFC300' controller on the market. According to the manufacturer, it enables universal and cross-industry programming, control and visualization. This is made possible by the Linux operating system, a dual-core 64-bit CPU and 2...
read more...Grossenbacher Systeme at the SPS 2024
Scalable 'Embedded AI' for Automation
The Universal Controller from Grossenbacher Systeme is based on the high-performance ARM-Yocto architecture and is scalable for AI applications through interchangeable System-on-Modules (SoMs).
read more...