AI-Powered EN50155 Railway GPU Computers for Modern Rail Networks

AI-Powered EN50155 Railway GPU Computers for Modern Rail Networks

Neousys Technology introduces the GT Series with the GT-92RL-H and GT-92GC, two powerful EN50155-certified railway GPU computers that redefine x86 railway computing. Combining Intel 14th/13th/12th-Gen processors with an NVIDIA RTX™ 2000 Ada GPU, these fanless, ruggedised 2U 19” chassis systems are designed to meet the high-performance demands of modern railway applications. Their robust construction, wide-range power input, and extensive I/O expandability make them ideal for a variety of AI-driven railway solutions, including pantograph inspections, track monitoring, video analytics, and tunnel surveillance.

EN50155 and EN45545 Certification

Both models comply with the stringent EN50155 railway standard and in addition the GT-92RL-H also complies with the EN45545 Fire Safety Certification, ensuring they meet the reliability and fire safety requirements essential for railway operations.

EN50155: Railway Reliability Certification

EN50155 is a critical certification that ensures railway computers can withstand extreme environmental conditions. It covers aspects such as:

• Temperature Resistance: Systems must operate in a wide temperature range to endure outdoor railway environments.
• Shock and Vibration Resistance: The GT-92RL-H and GT-92GC are designed to endure continuous vibrations and sudden shocks from high-speed train movement.
• Electromagnetic Compatibility (EMC): Ensures the systems operate reliably without interference from railway electrical infrastructure.
• Power Supply Stability: The wide-range DC power input (43V to 160V for GT-92RL-H and 8V to 48V for GT-92GC) ensures uninterrupted operation even with fluctuating power sources.

EN45545: Fire Safety Certification

The EN45545 fire safety certification ensures the materials and components used in railway systems are resistant to fire hazards. This standard is essential for preventing and mitigating fire-related risks in railway environments, contributing to passenger and infrastructure safety.

Fanless GPU Design for Maximum Reliability

The integration of an NVIDIA inference accelerator in a fanless architecture eliminates a single point of failure (SPOF). By removing active cooling components, Neousys significantly enhances system longevity and reliability, reducing the risk of overheating and mechanical failures.

2U 19” Rack-Mountable and Wall-Mountable Chassis

Designed for flexibility, both feature a 2U 19” chassis, making them suitable for both rack and wall mounting. This allows easy integration into railway systems, ensuring compatibility with existing infrastructures while offering scalable deployment options.

High I/O Expandability

Both systems support up to eight GigE/IP cameras and four USB3 cameras, making them ideal for various vision-based AI applications. With high-speed storage, including an M.2 NVMe slot with PCIe Gen4x4 support, and additional full-size mPCIe sockets for expansion, these computers cater to the growing demands of AI-driven railway automation.

Pantograph Inspection System

Pantographs are essential for electric trains, enabling continuous power supply by maintaining contact with overhead wires. Traditional inspection methods often fail to detect early-stage damage, leading to costly network delays and potential safety hazards. The GT-92RL-H and GT-92GC revolutionise pantograph monitoring by enabling real-time, AI-powered inspections that:

• Detect faults before they escalate, reducing maintenance costs.
• Operate in harsh railway environments with superior shock and vibration resistance.
• Utilise high-performance computing for 3D imaging and AI analytics, ensuring accurate fault detection.

Example: A European railway operator deployed an AI-based pantograph inspection system using the GT-92RL-H, significantly reducing power failures caused by undetected wear on overhead contact lines.

Railway Vehicle Inspection System

Manual railway vehicle inspections can be time-consuming, labour-intensive, and prone to human error. By leveraging the GT Series, operators can automate the process, achieving:

• Enhanced inspection accuracy using AI and machine learning algorithms.
• Reduced human effort and risk in hazardous environments.
• Faster identification of structural defects, ensuring safer and more reliable train operations.

Example: An automated railway vehicle inspection system powered by the GT-92GC helped an Asian metro system reduce inspection times by 50%, improving maintenance efficiency and train availability.

Tunnel Inspection

Tunnel integrity is crucial for railway safety. The GT Series facilitates real-time tunnel inspections as trains pass through, using high-speed image processing and AI-driven defect detection to:

• Identify structural issues such as cracks, water ingress, and erosion.
• Minimise downtime with automated condition monitoring.
• Improve maintenance scheduling, reducing unexpected failures and costly repairs.

Example: A major European railway network used GT-92RL-H systems to detect early-stage tunnel wall deterioration, allowing proactive repairs and preventing service disruptions.

Track Ballast Inspection

Railroad ballast ensures track stability and alignment. Traditional inspections often rely on periodic manual assessments, which can miss developing issues. The GT Series provides real-time ballast condition analysis, enabling:

• Early detection of ballast wear or displacement.
• Proactive maintenance to prevent track instability and derailments.
• Improved railway operational efficiency and reduced downtime.

Example: A North American freight railway implemented GT-92GC-based ballast monitoring, reducing derailment risks and cutting maintenance costs by 20% through predictive maintenance strategies.

GT-92RL-H: High-Performance Railway AI Computing

• Intel® 14th/13th/12th-Gen LGA1700 processor (35W/65W) ensures robust computing power for AI applications.
• Fanless NVIDIA RTX 2000 Ada GPU delivers cutting-edge AI inference capabilities.
• 43V to 160V DC input via M12 K-Coded connector supports railway power standards.
• 9x GbE ports via M12 X-Coded connectors for high-speed networking and camera connectivity.
• Wide-temperature operation (-40°C to 55°C) guarantees reliability in extreme conditions.
• M.2 PCIe Gen4x4 NVMe storage provides ultra-fast data processing speeds up to 7000 MB/s.

GT-92GC: Versatile AI Platform for Industrial and Railway Applications

• Intel® 14th/13th/12th-Gen LGA1700 processor (35W/65W) supports powerful AI-driven applications.
• Fanless NVIDIA RTX 2000 Ada GPU ensures efficient deep learning performance.
• 8V to 48V wide-range DC input via M12 L-Coded connector offers flexibility for various deployments.
• 8x GbE PoE+ ports with 1x GbE port provide excellent networking and camera connectivity.
• On-board isolated CAN bus enhances in-vehicle communication capabilities.
• Wide-temperature operation (-25°C to 55°C) enables deployment in diverse industrial environments.

The Neousys GT-92RL-H and GT-92GC represent the next generation of AI-powered railway computing. With robust EN50155 certification for railway reliability, EN45545 fire safety compliance, and fanless GPU technology, these systems are engineered for high-performance applications.

By integrating AI-driven analytics, these systems provide intelligent automation in railway operations, enhancing maintenance efficiency and minimising downtime. Whether improving pantograph inspections, automating tunnel monitoring, or optimising track ballast analysis, the GT Series ensures railway infrastructure remains resilient and efficient.

With advanced real-time data processing and machine learning capabilities, these solutions enable railway operators to transition from reactive to predictive maintenance strategies. This shift not only reduces operational costs but also enhances safety, reliability, and sustainability in railway transport. The GT-92RL-H and GT-92GC redefine the future of railway AI computing, setting a new standard for intelligent railway automation.