Product of the Week: ADLINK’s ROScube RQX-59 Series
July 24, 2023
Sponsored Story
Embedded robotic controllers play a critical role in edge AI applications like robotics, autonomous mobile robots (AMRs), and autonomous driving applications. Serving as the central processing unit and control hub, robotic controllers are tasked with managing communication and control, as well as directing the autonomous behavior and intelligent decision-making for these robotic systems.
Designed as a solution to further enhance the performance, reliability, and efficiency of these systems, the ROScube-X RQX-59 Series from ADLINK is a ROS (Robot Operating System) 2 robotic controller powered by the NVIDIA Jetson AGX Orin module. The NVIDIA module also enables the robotic controller to support 8 GMSL1/2 or FPD-Link III cameras for Frame Sync, precise data capture, and decision-making in the aforementioned applications.
ADLINK’s ROScube RQX-59 Series in Action
Additionally supported by an 8-core Arm Cortex-A78AE v8.2 64-bit CPU, a GPU with 1792 NVIDIA CUDA cores and 56 tensor cores, and dual deep learning accelerators, the ROScube-X RQX-59 Series is capable of supporting rugged AI computing in intelligent robotics platforms.
As part of ADLINK's ROS 2 solution, the ROScube robotic controllers support 32GB of 256-bit LPDDR5 memory and 64 GB of on module eMMC, in addition to one HDMI 2.1 display which serves as a graphical user interface (GUI) for users monitoring and interacting with robotic systems.
The I/O interfaces on the front panel of the RQX-59 Series consist of two GbE for high data transfer speeds in an Ethernet network, USB 3.0 with four USB Type A, and two USB Type A with lockable connector. The embedded system also features serial communication for COM1, which supports communication using the RS-232/485 protocols, and COM2 supporting the RS-232 protocol. The controller also supports one OTG port for changing environment image.
The side panel I/O interfaces include a DB-50 connector for UART, I2C, PWM, SPI, one CAN FD non-isolation, one CAN FD isolation 1.5KVac, GPIO, I/O for PPS / GPS Signals, LIDAR synchronization, and one audio input/output. The internal I/O features a Socket 2 and Socket 1, Key M 2280/3042 for Storage (Need NVMe SSD), Key E 1630/2230 for Wi-Fi, one Key B 3042/3052 socket for 5G LTE module, one USIM socket, and a real-time clock (RTC) supporting a 3-volt battery with a capacity of 550 milliampere-hours.
Designed for rugged AI robotics applications, especially for outdoor applications, the ROScube RQX-59 Series supports an operating temperature of 0ºC to 50°C at full CPU frequency with 0.6m/s airflow, and -20ºC to 70°C (-4°F to 158°F with 1.4GHz CPU) with 0.6m/s airflow. The operating humidity is supported at approximately 95% @40°C (non-condensing), and a storage temperature of -40ºC to 85°C.
The RQX-59 also features a fanless design and provides an ideal performance per watt with a power consumption of 40 W.
Getting Started with ADLINK’s ROScube RQX-59 Series
The embedded robotic controller solution from ADLINK provides customized BSP and drivers for top camera brands, as well as six SKU models available: RQX-590, RQX-59G/F,
RQX-590/G/F-E (with expansion box).
For software support, the ROScube family provides the NVIDIA JetPack SDK- and ADLINK's Neuron SDK-developed sources, designed to simplify and accelerate integration and development. Additionally, the ROScube family supports Ubuntu 20.04 L4T, Neuron SDK, and the NVIDIA Jetson SDK. The Jetpack 5.0.2 Beta version is available now for testing and the Jetpack 5.1.2 official MP version will be available in Mid-August.
