The company’s latest embedded AI platform advances edge computing for industrial applications.

NVIDIA’s newest product in its Jetson line of small form-factor, high-performance AI computers is the Jetson AGX Xavier Industrial module. The new ruggedized module is designed explicitly for autonomous machines in industrial settings. At only 100mm x 87mm, it’s a compact and energy-efficient module that brings hardware acceleration for the entire AI applications pipeline.
Although there are many industrial settings where AI could be useful, the harsh environments of factories, agricultural operations and construction sites aren’t well-suited to tech devices working correctly. However, as some of these environments could also benefit from robotics and automation handling the dirty work, NVIDIA designed the Jetson AGX Xavier Industrial module to specifically make it possible to deploy AI at the network edge in places with harsh conditions.
Like the recently released Jetson AGX Xavier, the industrial version of the module comes with 64GB of storage versus 32GB with the standard version. Although the standard version can operate at 32 trillion operations per second (TOPS), the industrial version comes in at 30 TOPS.
The Jetson AGX Xavier Industrial module allows developers to build advanced AI systems for video analytics, optical inspection, robotics, computer vision and autonomy in conditions that call for rugged devices. NVIDIA says the module can achieve 750Gbps of high-speed I/O while also withstanding severe shock, vibration and both hot and cold conditions between ˗40 degrees Celsius and 85 degrees Celsius. When relative humidity is 95 percent or higher, it can function between ˗10 degrees Celsius and 65 degrees Celsius. In addition, while the standard module provides configurable power profiles preset for 10W, 15W, and 30W, the industrial module comes with preset profiles for 20W and 40W.
The module is targeted for use in industrial, aerospace, defense, construction, agriculture, logistics, inventory management, delivery, inspection and health care settings. Some of the applications include site safety maintenance, property access, monitoring, and inspection. For example, in the oil and gas industry, the Jetson AGX Xavier Industrial can monitor pipelines, values and equipment and deliver real-time updates as needed.
Heavy equipment that must operate in various conditions is standard on construction sites and in agricultural settings. These sites can be large areas that are challenging to monitor. However, the Jetson AGX Xavier Industrial module provides autonomy for crewless aerial vehicles and drones that can be used to monitor such places.
The Jetson AGX Xavier Industrial joins the supercomputing abilities of the Jetson AGX Xavier system-on-module with rugged features to make it more reliable for deploying AI in inhospitable environments. It’s equipped with error correction codes, single error correction, double error detection, and parity protection for internal RAM resilience. It also has address and data bus error detection and correction capabilities and IP resiliency for industrial applications.
Safety is always a concern in industrial settings, so the Jetson AGX Xavier Industrial comes with safety capabilities supervised by a Safety Cluster Engine, which contains dual Arm Cortex-R5 processors for integration with fault-detection mechanisms to be used with safety-certified industrial-grade products.
The product features a 512-core NVIDIA Volta GPU with 64 Tensor Cores, two NVIDIA deep learning accelerators, two vision accelerators, an eight-core NVIDIA Carmel Arm CPU, and an encoder and decoder. It also comes with a hardware-verified secure boot, hardware-accelerated cryptography, support for encrypted storage and memory, and other security features.