Rugged Fanless PC Powers Real-Time AI Inference on Smart Factory Floor
A leading provider of material handling equipment approached Things Embedded to integrate advanced technology into their operations. Seeking a rugged fanless PC with built-in GPU, they aimed to enhance the intelligence and efficiency of their solutions. These embedded systems were required to support demanding real-time AI applications, enabling smarter and more reliable material handling across diverse industries.
Driven by a commitment to innovation, they provide comprehensive material handling solutions customised to meet each customer’s unique needs. Their end-to-end approach encompasses warehouse optimisation, fleet management, and operator training, ensuring enhanced efficiency, productivity, and safety across various industries. With expertise in electric forklifts, pallet jacks, reach trucks, and AGVs, this solutions provider empowers customers in sectors such as retail, warehousing, and distribution to transform their material handling operations.
Transitioning to a Smart Factory Floor with Real-Time AI Inference
The transition to smart factory floors introduces new opportunities and challenges, especially with the integration of autonomous vehicles into material handling operations. One key requirement is the need for advanced pedestrian avoidance systems and precise route navigation to ensure the safe and efficient movement of autonomous vehicles. Unlike traditional operations, smart factories demand systems that can adapt to dynamic layouts and fluctuating traffic patterns. Legacy detection systems often struggle in these environments, resulting in false alarms or missed hazards, which can lead to safety risks, operational disruptions, and increased costs.
Autonomous vehicles play a crucial role in optimising material handling on smart factory floors. However, their successful deployment depends on real-time AI inference and advanced data analytics to monitor vehicle movements, prevent collisions, and adapt to real-time changes. The harsh conditions of factory floors, including shock, vibration, and dust, further necessitate the use of rugged computing systems that can withstand these challenges. By leveraging embedded systems with integrated GPUs or AI Accelerators, smart factories can enhance safety, streamline operations, and make better, data-driven decisions to achieve greater productivity and efficiency.
Deploying discrete GPUs in harsh industrial environments presents several technical challenges. Discrete GPUs typically require high power input and active cooling with onboard fans to manage heat dissipation. The additional exposure to constant shock, vibration, dust, and debris can also degrade system performance and reliability, increasing the risk of hardware failure and maintenance costs.
NVIDIA, renowned for designing and manufacturing cutting-edge discrete GPUs for industrial applications, has expanded its innovation to edge computing with the introduction of NVIDIA Jetson, a powerful platform tailored for AI-driven edge applications. Part of the latest Jetson series, the Orin NX system-on-module provides an efficient alternative. Capable of delivering up to 100 TOPS of AI performance with a low-power draw of 10-25W, these system-on-modules can be integrated inside ruggedised PCs. This makes the latest embedded system architecture from NVIDIA ideal for real-time AI inference in harsh environments without the need for active cooling or high energy consumption.
Enabling Edge AI with Ruggedised Fanless PCs
Integrating a rugged fanless PC powered by the NVIDIA Jetson Orin NX architecture delivered ground-breaking advancements for the customer’s autonomous vehicles being deployed in smart factory environments. Equipped with powerful AI capabilities, the JCO-3000-ORN-B by C&T Solution enhances real-time object detection and pedestrian avoidance, optimising route navigation to minimize risks and promote workplace safety. Their ability to process and analyse imaging data instantaneously ensures seamless adaptation to shifting layouts, traffic patterns, and environmental conditions, making it an indispensable solution for dynamic material handling and evolving industrial workflows.
Rugged Chassis with Fanless Cooling
The rugged design features of the JCO-3000-ORN-B met the customer's criteria for industrial deployments across a wide range of applications, ensuring reliable performance even in challenging harsh environments, minimising disruptions and maintaining operational efficiency. This includes a fanless design with dustproof enclosure, shock and vibration resistance with a solid state design, and low power draw with reliable thermal management. By preventing collisions and minimising downtime, the JCO-3000-ORN-B rugged fanless PC helps reduce operational costs, offering customers greater peace of mind and a more secure investment.
Multiple PoE for Device Management
Featuring four GbE RJ45 LAN ports with support for Power over Ethernet (PoE) provided simplified installation and power management for onboard devices. These PoE ports enable high-speed data transmission for connected devices onboards the autonomous vehicles, such as cameras and LiDAR sensors. With reliable, high-bandwidth connectivity, the system ensures real-time data processing and decision-making, critical for improving the efficiency, safety, and productivity of material handling operations, and offering seamless communication across the smart factory network.
Remote Management with Out-of-Band Capabilities
Supporting Allxon out-of-band (OOB) capabilities into rugged fanless PCs deployed across remote locations provides a transformative solution for managing and maintaining distributed systems. OOB technology allows administrators to remotely monitor, diagnose, and resolve issues even when the primary system is offline, significantly reducing the need for on-site maintenance. This not only minimises operational downtime but also cuts maintenance costs by eliminating travel expenses and reducing the time required to address technical problems.
See How We Can Help With Edge AI Computing
Tell us about your application and a member of the team will get right back to you.