1. Product Overview

Product Name: ROScube-X
Model Series: ROScube-X Series
Article Number: ROScube-X
Manufacturer: ADLINK Technology Inc.
Product Type: Embedded Robotic Controller
Processor Platform: NVIDIA® Jetson AGX Xavier™
Target Application: AI-enabled robotics, autonomous systems, and intelligent edge computing

The ROScube-X is an advanced embedded robotic controller designed and manufactured by ADLINK Technology Inc. It integrates the NVIDIA® Jetson AGX Xavier™ module, providing a high-performance AI computing platform optimized for robotics and autonomous machine applications. The controller is fully ROS 2-enabled, offering seamless compatibility with the Robot Operating System (ROS) ecosystem. It is engineered for developers and integrators building intelligent robotic systems that require powerful AI inference, real-time sensor fusion, and efficient power consumption in a ruggedized form factor.

---

2. Key Features

- High-Performance AI Computing: Powered by the NVIDIA® Jetson AGX Xavier™ module featuring an integrated NVIDIA Volta GPU with 512 CUDA cores and dual deep learning accelerators (DLAs) for high-throughput AI inference.
- Energy Efficiency: Exceptional performance-per-watt ratio with total system power consumption as low as 30 W, ideal for mobile and embedded robotic platforms.
- Comprehensive I/O Connectivity: Multiple high-speed interfaces including USB 3.1 Gen2, Gigabit Ethernet, HDMI 2.0a, serial ports, and GMSL2 camera connectors for advanced sensor integration.
- Ruggedized Design: Lockable USB connectors, wide DC input range (9–36 V), and industrial-grade environmental tolerance ensure reliable operation in demanding field conditions.
- Time Synchronization: Built-in synchronization support for GMSL2 cameras and IMU sensors, enabling precise multi-sensor data alignment for robotics and vision applications.
- Software Ecosystem: Full support for NVIDIA JetPack SDK and ADLINK Neuron SDK, providing a complete development environment for AI, computer vision, and robotics applications.
- ROS 2 Integration: Preconfigured for ROS 2, facilitating rapid deployment of robotic frameworks and middleware for autonomous navigation, manipulation, and perception.
- Flexible Expansion: M.2 and Mini PCIe slots for storage, Wi-Fi, LTE, and other communication modules.
- Industrial Reliability: Compliance with CE, FCC Class A, UL, and cUL standards, ensuring electromagnetic and safety robustness.

---

3. Technical Specifications

| Component | Specification |
|----------------|-------------------|
| Processor | NVIDIA® Jetson AGX Xavier™ |
| CPU | Carmel ARMv8.2 8-core @ 2.26 GHz |
| GPU | 512-core NVIDIA Volta GPU @ 1.37 GHz |
| Memory | 32 GB LPDDR4x on-module |
| Storage (eMMC) | 32 GB on-module |
| Display Output | 1 × HDMI 2.0a |
| Ethernet | 2 × GbE ports (IEEE 1588 and 802.1AS support) |
| USB 3.1 Gen2 | 1 × Type-A (lockable) |
| USB 3.1 Gen1 | 4 × Type-A + 1 × Type-A (lockable) |
| Serial Ports | COM1: RS-232/485; COM2: RS-232 |
| OTG Port | 1 × USB OTG (for environment image change) |
| DB-50 Connector | UART, SPI, CAN, I²C, PWM, ADC (isolated), 20-bit GPIO |
| Audio | 1 × Audio In/Out |
| M.2 Expansion | 1 × Socket 2 (Key B+M, 2280) for NVMe SSD; 1 × Socket 1 (Key A+E, 1630/2230) for Wi-Fi |
| Mini PCIe | 1 × Mini PCIe socket for LTE module |
| USIM Slot | 1 × USIM socket |
| RTC | 2-pin 1.25 mm wafer (optional) |
| User LEDs | 5 × user-defined LEDs (Green: U1–U3; Amber: U5; Yellow: U4) |
| Power LED | 1 × Power ON indicator |
| Power Input | 9–36 V DC (±5% tolerance, reverse polarity protection) |
| Power Adapters | 160 W / 220 W AC/DC adapters (optional) |
| Power Switch | 1 × Power button |
| Reset Controls | 1 × Recovery button; 1 × Hardware Reset button |
| Dimensions (Main Unit) | 190 (W) × 210 (D) × 80 (H) mm (7.48 × 8.27 × 3.15 in) |
| Dimensions (With Expansion) | 322 (W) × 210 (D) × 80 (H) mm (12.68 × 8.27 × 3.15 in) |
| Mounting | Wall-mount kit included |

---

4. Functionality

The ROScube-X functions as a high-performance embedded controller for robotic and AI-driven systems. It integrates compute, sensor, and communication interfaces into a single compact platform. The NVIDIA Jetson AGX Xavier module provides the computational backbone, enabling real-time AI inference, computer vision, and sensor fusion tasks.

Use Cases and Applications:
- Autonomous mobile robots (AMRs) and AGVs
- Industrial automation and collaborative robots (cobots)
- Drones and unmanned ground vehicles
- Smart surveillance and inspection systems
- AI-based vision and perception platforms
- Research and development in robotics and AI

Compatibility:
- Fully compatible with ROS 2 and ROS 1 frameworks
- Supports Ubuntu 18.04 LTS with Linux for Tegra (L4T)
- Integrates with NVIDIA JetPack SDK for CUDA, cuDNN, TensorRT, and DeepStream
- Works with ADLINK Neuron SDK for AI model deployment and sensor integration

The controller’s GMSL2 camera connectors and IMU synchronization capabilities make it ideal for robotics applications requiring precise timing and multi-sensor data fusion. Its ruggedized design ensures stable operation in industrial and outdoor environments.

---

5. Components & Accessories

Included Components:
- ROScube-X main unit (RQX-58G or RQX-580 variant)
- Wall-mount kit
- Power button and reset controls
- Lockable USB connectors for secure peripheral attachment

Optional Accessories:
- Wireless Module: Intel® Wireless-AC 9260 M.2 2230 Dual-Band 2×2 Wi-Fi + Bluetooth 5 kit (P/N: 91-95266-0010)
- FAKRA Cable: For GMSL2 camera connection
- AC/DC Power Adapters:
- 220 W adapter (P/N: 31-62149-0000)
- 160 W adapter (P/N: 31-62120-0010)

Ordering Variants:
- RQX-58G / RQX-58G-E: With FAKRA GMSL2 connectors / with expansion box
- RQX-580 / RQX-580-E: Standard model / with expansion box

---

6. Installation & Setup

Power Requirements:
The system operates on a 9–36 V DC input with ±5% tolerance and reverse polarity protection. It can be powered using the optional 160 W or 220 W AC/DC adapters, which connect via a Molex lockable connector to a Phoenix terminal block.

Mounting Options:
The ROScube-X supports wall mounting using the included kit, allowing flexible installation in industrial cabinets, mobile platforms, or laboratory setups.

Configuration and Setup:
- The system runs Ubuntu 18.04 LTS (L4T) as the base OS.
- Developers can install and configure ROS 2, Neuron SDK, and JetPack SDK for application development.
- The OTG port allows environment image updates or recovery.
- LED indicators provide system status and user-defined signaling.
- Reset and recovery buttons facilitate system maintenance and firmware updates.

---

7. Performance & Capabilities

Computational Performance:
The NVIDIA Jetson AGX Xavier module delivers up to 32 TOPS (Tera Operations Per Second) of AI performance, enabling real-time inference for deep learning models. The 512-core Volta GPU with Tensor Cores accelerates neural network processing, while the dual DLAs offload AI workloads for optimized efficiency.

Power Efficiency:
With a power draw as low as 30 W, the ROScube-X achieves an excellent balance between performance and energy consumption, making it suitable for battery-powered or mobile robotic systems.

Connectivity and Expansion:
Dual Gigabit Ethernet ports with IEEE 1588 and 802.1AS support ensure precise time synchronization across distributed robotic systems. The multiple USB 3.1 ports and DB-50 connector provide extensive peripheral connectivity for sensors, actuators, and communication modules.

Environmental Performance:
- Operating Temperature: 0–50°C at full CPU frequency; -20–70°C at reduced frequency
- Storage Temperature: -40–85°C
- Humidity: Up to 95% @ 40°C (non-condensing)
- Vibration Resistance: IEC 60068-2-64, 5 Grms (5–500 Hz, 3 axes)
- Shock Resistance: MIL-STD-202G Method 213B, 100 G half-sine, 11 ms duration (without expansion)

Limitations:
While optimized for embedded and industrial environments, the ROScube-X is not designed for extreme outdoor exposure without additional enclosure protection. The operating temperature range may require thermal management in high-load or high-ambient conditions.

---

8. Standards & Compliance

Electromagnetic Compatibility (EMC):
- CE & FCC Class A (EN61000-6-4 / EN61000-6-2)
- IEC 61000-4-2: ESD protection (±8 kV contact, ±15 kV air with expansion)
- IEC 61000-4-3: Radiated susceptibility (10 V/m from 80–1000 MHz; 3 V/m from 1400–2000 MHz; 1 V/m from 2000–2700 MHz)
- IEC 61000-4-4: Electrical fast transient (±2 kV on power port, ±1 kV on signal port)
- IEC 61000-4-5: Surge immunity (±2 kV line-to-earth on power port; ±1 kV on signal port)
- IEC 61000-4-6: Conducted susceptibility (10 Vrms, 0.15–80 MHz)
- IEC 61000-4-8: Power-frequency magnetic field immunity
- IEC 61000-4-11: Voltage dips and interruptions

Safety Certifications:
- UL and cUL certified for electrical safety

Environmental and Mechanical Standards:
- Vibration: IEC 60068-2-64
- Shock: MIL-STD-202G Method 213B

These certifications ensure that the ROScube-X meets industrial-grade reliability and safety requirements for deployment in mission-critical robotic systems.

---

9. Additional Information

Software Environment:
- Operating System: Ubuntu 18.04 LTS (Linux for Tegra)
- SDK Support: ADLINK Neuron SDK and NVIDIA Jetson SDK (JetPack)
- Development Tools: CUDA, cuDNN, TensorRT, DeepStream, ROS 2 middleware

Warranty and Support:
- Standard manufacturer warranty provided by ADLINK Technology Inc.
- Technical support includes documentation, SDK updates, and developer resources via ADLINK’s support portal.
- Firmware and software updates are available through the Neuron SDK and JetPack SDK ecosystems.

Maintenance:
- Regular firmware updates recommended for optimal performance and security.
- System diagnostics and recovery can be performed via the OTG port and recovery button.
- Lockable connectors and ruggedized design minimize maintenance frequency in industrial environments.

Product Lifecycle and Updates:
- Specifications subject to change without notice (latest revision: May 12, 2020).
- Continuous software support for ROS 2 and JetPack SDK ensures long-term compatibility with evolving AI and robotics frameworks.

---

Summary

The ADLINK ROScube-X is a robust, high-performance embedded robotic controller purpose-built for AI-driven robotics and autonomous systems. By integrating the NVIDIA Jetson AGX Xavier module, it delivers exceptional computational power for deep learning, computer vision, and sensor fusion tasks while maintaining low power consumption. Its comprehensive I/O, ruggedized design, and ROS 2 compatibility make it an ideal platform for industrial, research, and commercial robotics applications.

With extensive connectivity options, environmental resilience, and compliance with international standards, the ROScube-X provides a reliable foundation for next-generation intelligent machines. Its support for the Neuron SDK and JetPack SDK ensures developers have access to a complete AI and robotics development environment, enabling rapid prototyping and deployment of advanced autonomous solutions.