Product of the Week: Syslogic's NVIDIA Jetson AGX Xavier Industrial-based Rugged Computer

November 02, 2021

Sponsored Story

Product of the Week: Syslogic's NVIDIA Jetson AGX Xavier Industrial-based Rugged Computer

A hallmark of proven technology is viability in rugged industrial applications. And a true measure of that viability is how it performs under the rigors of mobile applications.

Rugged mobile systems architects can now check both boxes while adding cutting edge AI capabilities to their designs by adopting Syslogic's NVIDIA Jetson AGX Xavier-based Rugged Computer, the RML A3.

The Syslogic RML 3 is an edge AI inferencing system suitable for transportation infrastructure and mobility use cases, including automated guided vehicles (AGVs), cars and trucks, buses, construction and agricultural machinery, and other special vehicles. It brings 32 TOPS of AI performance courtesy a ruggedized variant of the Jetson AGX Xavier Industrial module, which is generated by 512 NVIDIA CUDA cores, 64 Tensor cores, two NVIDIA deep-learning accelerators, two vision accelerators, and an eight-core NVIDIA Carmel Arm CPU, all built on the NVIDIA Volta architecture.

All of this performance empowers the RML 3 in a range of inferencing tasks from object detection, recognition, and classification to robotics and machine vision to video analytics and sensor fusion. In application terms, it can span advanced driver assistance systems and autonomous driving to traffic engineering and rail control thanks in part to its robust housing.

Syslogic combines the 10/15/30 W industrail Jetson module with its own carrier board in an IP 67 and IP 69-rated sealed system with no moving parts and passively cooled. The total solution operates at temperatures of –40°C to +70°C and meets shock and vibration requirements in accordance with standards like EN 1175-1, EN60068-2-27/64, EN 61373, ISO 15003. This thanks to features like screw-type M12 connectors, a pressure equalization membrane, and power ignition controller that support continuous operation in extreme environmental conditions.

They even comply with the UNECE-R10 (E-Mark) EMC automotive testing standard.

The Syslogic RML 3 AI Computer in Action:

In action, the 250 mm x 105 mm x 107 mm RML 3 maximizes AI performance by transferring data quickly between the Xavier AGX processor complex and 32 GB of soldered-on 256-bit LPDDR4 RAM, an additional 32 GB of eMMC 5.1 Flash onboard, and significant storage expansion cpabilities. This data can be shuttled to and from the platform over extensive I/O, inlcuding:

  • PoE+ Power over Ethernet
  • 2x Ethernet 10/100/1000Mbit
  • 2x USB version 2.0 (Type A)
  • Internal USB version 2.0 OTG
  • Passive or Active CAN
  • DisplayPort 1.2 

The industrial-based vehicle computer also supports the CAN-based SAE J1939 network protocol. It even allows for precise positioning via support for an optional GNSS module and LTE or Wi-Fi connectivity that is accessible through mini PCIe expansion.

But what really sets the RML 3's operational AI capabilities apart is software. It comes pre-installed with the NVIDIA Jetpack SDK and Linux for Tegra (L4T) Linux operating system distributions.

"Powered by the NVIDIA CUDA-X accelerated computing stack and JetPack SDK support, Jetson AGX Xavier Industrial is a fully software-defined platform enabling cloud-native capabilities," an NVIDIA blog on the use of AGX Xavier in industrial applications reads. "NVIDIA’s CUDA-X acceleration, free production-ready pretrained models on NGC and NVIDIA TAO Toolkit give developers the fastest path to build and deploy deep learning and AI training and inference systems.”

Getting Started with the Syslogic RML 3 AI Computer:

Gettint started with the Syslogic RML 3 is pretty straightforward, especially if you're already acquainted with the NVIDIA Jetson AGX Xavier developer kit. Software designed on that eval platform can easily be ported to the RML 3's AGX Xavier module, which cuts time to market and makes commercialization as streamlined as possible. This includes AI algorithms built and optimized with CUDA-friendly libraries and frameworks like TensorFlow Lite.

From there, you really just need to provide the system with an input voltage of 9 to 45 V DC. It is resistant to chemicals; comes with long-term, fixed BOM availability; and is maintenance free, so once you've set it, you can pretty much forget it.

For more information and purchasing options for Syslogic's RML 3 AI computer, visit www.syslogic.com/eng/ai-rugged-computer-rml-jetson-agx-xavier-106848.shtml or check out the resources below.

Resources: