ADLINKジャパン List of Products and Services

ADLINK's MECS-7210 is an industry-leading multi-access edge computing server designed to meet the requirements for ultra-low latency, high bandwidth, and real-time access to 5G wireless networks. This server has an extended temperature range of -5°C to 55°C, and its slim chassis, measuring only 420mm, makes it ideal for traditional base stations. The Charles Micro-Edge Data Center SC102 enclosure is specifically designed for the MECS-7210, providing a reliable and protected environment for placing compute nodes on small cells or 5G poles. The enclosure features integrated power and cooling and is a self-contained, secure, slim unit compliant with GR-487.

In embedded devices, there is a growing demand for GPU acceleration across various applications such as industrial image processing, high-resolution multi-display, and AI. This white paper explains the optimal solutions for these needs. 【Catalog Contents】 ● Systems and boards compatible with MXM ● Systems and boards compatible with PEG ● Q&A ● Product specifications 　＜Case Studies＞ In various fields such as clinical diagnostic imaging devices, image processing radars for aerospace and defense, rugged military laptops, and gaming slot machines, industries require high resolution, multi-display capabilities, parallel computing, and AI. *For more details, please download from the link below or contact us.

The ADLINK ROScube-X, equipped with the Xavier module, features an integrated Volta GPU and dual deep learning accelerators, along with various interfaces for robot system integration. The ROScube-X supports all resources developed using the NVIDIA JetPack SDK and ADLINK's Neuron SDK, making it particularly suitable for robotic applications that demand high AI computing with minimal power consumption.

Edge Computing Platform for Realizing Decision-Making from Data 【Catalog Contents】 ● Realizing the Factory of the Future ● From Data to Decision-Making ● Edge IoT Smart Manufacturing Solutions ● Controlling Data Flow with Edge IoT ● Real-Time Collection and Management of Legacy Equipment Data ● Equipment Condition Monitoring and Predictive Maintenance ● AI-Compatible Machine Vision ● Selection Guide *For more details, please download from below or contact us.*

Edge Computing Platform for Realizing Decision-Making from Data 【Catalog Contents】 ● Realizing the Factory of the Future ● From Data to Decision-Making ● Edge IoT Smart Manufacturing Solutions ● Manipulating Data Flow with Edge IoT ● Real-Time Collection and Management of Legacy Equipment Data ● Equipment Condition Monitoring and Predictive Maintenance ● AI-Enabled Machine Vision ● Selection Guide *For more details, please download from below or contact us.*

To realize artificial intelligence (AI) at the edge, ADLINK integrates the DLAP-211-Nano edge inference platform with NVIDIA Jetson Nano to accelerate deep learning workloads for object detection, recognition, and classification. By accelerating the adoption of AI across industries, the DLAP-211-Nano analyzes traffic flow at intersections in real-time, optimizing traffic management, enhancing smart bus services, strengthening security monitoring, and enabling more smart city and smart manufacturing applications.

To realize artificial intelligence (AI) at the edge, ADLINK integrates the DLAP-211-JNX Edge Inference Platform with Jetson Xavier NX to accelerate deep learning workloads for object detection, recognition, and classification. By accelerating the adoption of AI across industries, the DLAP-211-JNX analyzes traffic flow at intersections in real-time, enabling optimization of traffic management, enhancement of smart bus services, strengthening of security monitoring, and the realization of more smart city and smart manufacturing applications.

● Smart medical robots that reduce the risk of hospital-acquired infections and alleviate the burden on healthcare staff. ● AI-based facial recognition that excels in recognition efficiency and prevents contact infections. ● Advanced remote monitoring of production facilities to reduce the risk of viral infections among workers. ● Towards a smarter future.

◆ROScube-I ADLINK's ROScube-I is a real-time ROS 2 compatible robot controller based on Intel Xeon 9th generation Intel Core i7/i3 and 8th generation Intel Core i5 processors, featuring excellent I/O connectivity that supports various sensors and actuators for unlimited robot applications. It also supports Intel VPU and NVIDIA GPU cards for AI algorithm and inference computations. ◆ROS Starter Kit - Neuron The ADLINK ROS Starter Kit features flexible connectivity with a wide range of I/O ports and support for AI computing platforms. Additionally, compatibility with open-source ROS/ROS 2 provides full access to open-source application libraries for robot control, including vision, navigation, and motion control, facilitating the rapid realization of ROS/ROS 2 functionalities.

The contact lens market in the Asia-Pacific region is expected to grow at a CAGR of 8.16% from 2020 to 2025. This rapid growth relies on increased production and improved quality inspection rates. ADLINK and its AI solution partner, LEDA Technology, provide AI-enabled Automated Optical Inspection (AOI) solutions to help Taiwanese contact lens manufacturers enhance the accuracy and speed of quality inspections.

The EGX-MXM-T1000 module features advanced NVIDIA Turing GPU technology with an MXM 3.1 Type A form factor. Its compact, slim, and reliable design makes it suitable for mission-critical environments. The EGX-MXM-T1000 enhances performance per watt. This MXM GPU module provides a flexible and easy solution for deep learning applications in fields such as healthcare, image processing, and gaming applications.

The EGX-MXM-RTX3000 module features advanced NVIDIA Turing GPU technology with an MXM 3.1 Type B form factor. Its compact, slim, and reliable design makes it suitable for mission-critical environments. With the Quadro Turing TU106 processing core, the EGX-MXM-RTX3000 supports four DP1.4 displays, providing a flexible and easy solution for medical and gaming applications.

The CM5-P1000 PCIe104 module is a compact and slim graphics module solution compliant with the PCI/104-Express & PCIe/104 version 3.0 specifications, providing the benefits of the latest and most advanced GPUs for embedded systems. Its excellent graphics performance, GPU computing, and video capabilities make it an ideal solution for systems that require performance, such as digital signage, medical imaging, defense, and aerospace applications.

Standard Operating Procedures (SOPs) in manufacturing are created to optimize product quality and cycle time, as well as to protect workers from hazards. However, the main failure factor in adhering to SOPs is human error. Therefore, continuous auditing and improvement of procedures are key to the success of SOPs. Traditionally, manufacturers have audited SOP compliance through manual monitoring. Tracking every process performed by each operator is too time-consuming to implement across all production lines. This has led manufacturers to seek automated solutions. "A high level of standardization is required in display assembly to ensure product quality. A solution that constantly monitors every movement on the assembly line is necessary. ADLINK addresses these requirements by utilizing AI smart cameras and behavior analysis deep learning algorithms," stated an assembly manager from a world-leading display manufacturer.

ADLINK provides an optimized high-availability computing platform with scalability options, including MXM cards. The DLAP-3000-CFL series platform features active cooling for applications that require a very high level of computing in limited space, supporting performance-demanding workloads such as deep learning. Features of the DLAP-3000-CFL series: - Minimal embedded GPU platform (3 liters) - Reliable low TDP MXM GPU design - Scalable graphics processing from Pascal to Turing GPUs.

Due to the aging population and the increasing demand for early diagnosis, medical imaging diagnostics are required to provide faster, more efficient, and higher diagnostic accuracy. As a result of the rapid development of deep learning algorithms and GPU-accelerated computing, AI is expected to become a catalyst to meet the growing demand for medical image processing. According to a medical market research report, the market for artificial intelligence in medical image processing is expected to grow at a CAGR of +30% from 2021 to 2026. ADLINK leverages its expertise and experience in optimizing edge computing performance for over 20 years to provide GPU solutions, helping medical imaging equipment manufacturers worldwide to utilize AI applications to achieve consistently clear and accurate medical images.

ADLINK provides an optimized high-availability computing platform with scalability options, including MXM cards. The DLAP-3100-CF series platform features active cooling for applications that require a very high level of computing in limited space, supporting performance-demanding workloads such as deep learning.

ADLINK provides an optimized high-availability computing platform with scalability options, including MXM cards. The DLAP-3200-CF series platform features active cooling for applications that require a very high level of computing in limited space, supporting performance-demanding workloads such as deep learning.

ADLINK provides an optimized high-availability computing platform with scalability options, including MXM cards. The DLAP-4000F series platform features active cooling for applications that require a very high level of computing in limited space, supporting performance-demanding workloads such as deep learning.

ADLINK provides an optimized high-availability computing platform with scalable options, including MXM cards. The DLAP-8000 series platform features active cooling for applications that require a very high level of computing in limited space, supporting performance-demanding workloads such as deep learning.

GPUs can enhance the performance of various workloads such as image processing and analysis, compute acceleration, and artificial intelligence (AI). As the demands for responsiveness and accuracy in edge systems increase, the combination of CPUs and GPUs is becoming mainstream to optimally streamline performance per watt and per dollar.

GPUs can enhance the performance of various workloads such as image processing and analysis, compute acceleration, and artificial intelligence (AI). As the demands for responsiveness and accuracy in edge systems increase, the combination of CPUs and GPUs is becoming mainstream to optimally optimize performance per watt and per dollar.

GPUs can enhance the performance of various workloads such as image processing and analysis, compute acceleration, and artificial intelligence (AI). As the demands for responsiveness and accuracy in edge systems increase, the combination of CPUs and GPUs is becoming mainstream to optimally streamline performance per watt or per dollar.

◆Introduction As ROS 1-based robotic systems become outdated or unable to meet the evolving needs of new industrial applications, the transition of robotic systems to ROS 2 is considered a practical next step in the evolution of robotic system design. The purpose of this guide is to address the following challenges related to the transition from ROS 1 to ROS 2: ● Reasons for transitioning from ROS 1 to ROS 2 ● Who needs to consider transitioning to ROS 2 ● Challenges of the transition ● Solutions and recommendations needed for the transition ● Transitioning Gazebo ROS ● Examples of migrating ROS 2 C++ source code

Tank trucks are one of the most common and efficient means for chemical manufacturers to transport liquid and gaseous chemicals to industrial production facilities. Chemicals can be corrosive, explosive, flammable, or toxic. If hazardous chemicals leak, they can cause long-term damage to people and the environment. However, being human, tank truck drivers and storage tank operators inevitably make mistakes or become distracted during unloading. An analysis of over 1,600 hazardous chemical incidents revealed that 65% of these accidents were due to human error. With the maturation of deep learning technology, AI-enabled machine vision solutions such as IP surveillance cameras and industrial smart cameras have become capable of monitoring the entire chemical unloading process and detecting anomalies before human accidents occur. Combining IP surveillance cameras with AI network video recorders (NVRs) provides installation flexibility, but transmission delays remain an unresolved issue. Industrial smart cameras integrated with machine learning algorithms have become feasible and have been introduced in manufacturing plants to prevent the occurrence of human errors.

Automakers are developing and deploying autonomous mobile robots (AMRs) that can handle and transport production materials without human intervention to enhance logistics efficiency. IDEALworks, a wholly-owned subsidiary of the BMW Group, sought support from NVIDIA and ADLINK in developing the necessary software stack and a robust edge AI computing platform. By downloading this case study, you can see how NVIDIA and ADLINK supported IDEALworks in building AMRs, freeing employees from repetitive loading tasks, and allowing them to focus on core competencies.

The EOS-JNX series is a vision system series equipped with the Jetson Xavier NX, supporting 4CH PoE cameras and digital I/O. With Smart PoE, remote control of PoE power and efficient PoE power budget control are possible, enhancing convenience on-site. It comes pre-installed with an AI inference SDK plugin, reducing the TCO for application development. The EOS-JNX-I is designed as an AI PoE switch that connects to IP cameras, supporting video stream extraction, image codecs, and an Onvif RTSP data plugin for AI inference, enabling AI safety in IP surveillance systems. The EOS-JNX-G is designed for industrial AI machine vision, providing dedicated bandwidth (1Gb/ch) with no frame drops for GigE camera connections.

GPUs can enhance the performance of various workloads such as image processing and analysis, compute acceleration, and artificial intelligence (AI). As the demands for responsiveness and accuracy in edge systems increase, the combination of CPUs and GPUs is becoming mainstream to optimally optimize performance per watt or dollar.

The camera plays an important role not only in conventional ADAS (Advanced Driver Assistance Systems) solutions but also in level 3/4 autonomous driving and many other applications. TIER IV is developing a camera that achieves optimal performance for autonomous driving applications by incorporating the extensive knowledge gained from previous autonomous driving development and demonstrations, as well as advanced software design expertise. By using this camera in conjunction with ADLINK's ROS 2-compatible robot controller, it will be possible to obtain reliable recognition results in various challenging sensing environments. This product will be on display at the booth of Marubun Corporation during the Embedded Edge Computing Expo held from April 6 to 8, 2022. We look forward to your visit.

The healthcare system is an invaluable asset to modern society, but there has been little technological effort to address the impending shortage of healthcare workers and to alleviate their physical and mental stress. Since the outbreak of the novel coronavirus in 2020, hospitals have been under excessive strain and are on the brink of collapse; however, the development of autonomous mobile robots (AMRs) to meet various medical needs is progressing. Taichung Veterans General Hospital has successfully used AMRs for the non-contact transportation of large surgical medical equipment in collaboration with New Era AI Robotic Inc. (N.ERA) and ADLINK, eliminating the delivery burden on medical staff and achieving a reduction of 5,000 hours of labor annually. Therefore, Taichung Veterans General Hospital first introduced AMRs to reduce labor costs associated with time-consuming repetitive tasks. The hospital, which was considering the introduction of AMRs, realized that the most labor-intensive task for the medical group was the delivery of heavy surgical instruments and equipment. In this regard, N.ERA examined the issue with the expectation that automating repetitive tasks in the hospital would allow healthcare workers to focus on other emergencies and responsibilities, and developed a meticulous plan.

The MECS-7211 edge server focuses on the edge of 5G networks and provides an open and white-box platform for a wide range of 5G MEC applications, including 5G Open RAN, private 5G networks, smart cities, retail, manufacturing, autonomous vehicles, augmented reality (AR), virtual reality (VR), and healthcare. The MECS-7211 is designed to meet stringent requirements, achieving high levels of deployment flexibility, reliability, scalability, and carrier-grade processing performance, while also enhancing hardware scalability. The MECS-7211 is a high-performance COTS platform that meets various application and deployment requirements, featuring a system depth of 450mm, a 2U chassis, full front access to all I/O, a wide operating temperature range (-5°C to +55°C), and reserved PCIe expansion slots for access to acceleration hardware such as GPUs, FPGAs, and QAT adapters, allowing customers to focus on differentiating their final solutions.

The ADLINK AI Camera Dev Kit, designed to facilitate PoC, integrates the NVIDIA Jetson Nano SOC, an 8MP color MIPI camera module, and verified software drivers, eliminating integration hassles and resolving compatibility issues. Users can start development with this kit right out of the box. Built on a common V4L2 video interface, users can easily enhance the kit using the included sample code and resources from NVIDIA's technical forums, allowing for convenient PoC creation with minimal effort. For AI beginners and in-house system integrators, the pre-installed EVA software provides two AI applications that include sample code for image capture, AI inference, and logical control, enabling rapid and easy development of PoCs or applications. EVA is a GUI-based development tool that offers a low-code environment with software tools for labeling and training, as well as two carefully selected open-source models, allowing users to build PoCs without AI expertise.

ADLINK's MXM-AXe offers up to 8 hardware ray tracing cores, 128 execution units, 4GB GDDR6, and a maximum TGP of 50W with 8x PCIe Gen4. It provides a 50W TGP and the industry's first full AV1 hardware encoding. This module, which significantly enhances graphics rendering, is key to adapting to next-generation graphics workloads. The MXM-AXe leverages Intel's established graphics ecosystem, which edge developers have enjoyed and trusted for years, including OpenVINO for AI and Intel OneAPI management tools. Rather than a new lineup of GPUs, it is designed to seamlessly transition from integrated graphics to discrete graphics as before. Supporting Intel Deep Link technology, the MXM AXe, when combined with 12th and 13th generation Intel Core processors, achieves even higher performance and power efficiency through automatic workload allocation between integrated GPU, discrete GPU, and CPU.

The ADLINK NVIDIA Jetson-based industrial AI camera NEON-2000-JNO series integrates Jetson Nano, image sensors, an optimized OS, and a wide range of I/O for vision applications into a compact chassis that has been verified for thermal stability, reducing TCO related to integration and troubleshooting while minimizing the space required for cables and installation. The NEON-2000-JNO series, housed in a compact enclosure with two types of image sensors and built-in COM and LAN ports, is ideal for edge AI vision applications.

Pocket AI offers ultimate flexibility and reliability on the go. This portable plug-and-play AI accelerator provides the perfect power/performance balance from NVIDIA RTX GPUs. Pocket AI is the ideal device for AI developers, professional graphics users, and embedded industrial applications, enhancing productivity by improving work efficiency.

The ADLNK NVIDIA Jetson AGX Orin-based edge AI inference platform DLAP-411-Orin is a high-performance platform designed for the deployment of AI-enabled systems at the edge, providing the computing power required for demanding AI applications. Designed in a compact size, it can be easily integrated into various edge devices and systems without limitations on connectivity. The DLAP-411-Orin features flexible connectivity options such as Gigabit Ethernet, dual USB, and M.2 interfaces. The DLAP-411-Orin is a versatile and powerful AI platform that facilitates easy integration with a wide range of devices and sensors.

The DLAP-211-Orin series is an exceptional edge AI inference platform designed to achieve unparalleled high-performance computing capabilities. Equipped with the Jetson Orin system-on-module (SoM), the DLAP-211-Orin series combines a highly optimized computing engine, multi-instance GPU, and dedicated deep learning accelerator. This device impresses with its compact size, scalability, reliability, and AI performance that is up to 8 times better than conventional products. It offers hardware acceleration for deep learning, advanced safety and security features, and support for multiple sensors and cameras. The DLAP-211-Orin series enables developers to build advanced AI-powered edge solutions with a focus on performance, flexibility, safety, and security, making it an ideal choice for cutting-edge edge computing applications that require top-tier technology.

"Pocket AI" is a compact product that combines high flexibility and reliability, allowing it to be used as an AI accelerator on the go. It supports plug-and-play and offers an excellent power-performance balance utilizing the NVIDIA RTX A500. It is suitable for the work environments of AI developers and professional graphic users, as well as for field applications in the embedded industrial sector, contributing to increased productivity through accelerated processing. The Thunderbolt interface used in this product is a high-speed communication standard that has become popular in laptops, thin clients, and compact PCs. With the evolution to version 4.0 (backward compatible with 3.0) and the adoption of USB Type-C connections, peripheral devices and gadgets utilizing this technology are continuously emerging. By leveraging the optical transfer speed of Thunderbolt (up to 40Gb/s) and the general availability of hosts, it achieves an intuitive plug-and-play UX that significantly enhances productivity.

ADLINK's NVIDIA Jetson-based industrial AI camera NEON-2000-ONO series integrates Orin Nano, image sensors, an optimized OS, and extensive I/O for vision applications into a compact housing with confirmed thermal stability, reducing the total cost of ownership in integration and troubleshooting while minimizing the cable wiring and space required for installation.

The MECS-7120 is one of the MECS (Multi-Access Edge Computing Server) series, specially designed to meet the needs of open RAN (Radio Access Network), private 5G (p5G), and industrial edge computing. The MECS-7120 features robust computing capabilities, a compact and durable design, and flexible scalability with storage slot options, addressing common challenges faced by customers such as limited space or budget constraints and the need for versatile applications in complex network environments.

ADLINK's AI edge server MEC-AI7400 series can be equipped with various acceleration cards depending on application needs. With flexible and diverse combinations such as GPU cards, motion control cards, IO cards, and image capture cards, it can adapt to various changes in smart manufacturing applications. The MEC-AI7400SR features a single 4th generation Intel Xeon Scalable processor, 8x DDR5 DIMM (4800MT/s, 8 channels, 1DPC), and supports up to 5x full-height PCIe slots for GPU/FPGA accelerator cards. Additionally, it is equipped with an AST2600-based BMC, compliant with IPMI v2.0/Redfish, has a dustproof design, and a short chassis, making it suitable for factory and automation environments.

The MLB-3000 is a medical-grade platform designed for integration into medical devices such as C-arms and surgical navigation systems, enabling real-time edge computing and rapid diagnostic testing at the point of care. This compact medical platform features ADLINK's MXM graphics module, which supports NVIDIA GPUs, accelerating image reconstruction and beamforming while significantly reducing data transmission latency. This GPU-enabled platform enhances image quality through AI, assisting doctors in detecting diseases and other abnormalities, ultimately providing better healthcare to people. The MLB-3000 complies with IEC standards regarding the safety and performance of medical devices and comes with a five-year warranty, supporting a wide range of applications from gateways to computing, as well as from PACS (Picture Archiving and Communication System) and alarm settings to image reconstruction and real-time AI inference.

The MLB-3002 is a platform compliant with medical standards, designed for integration into medical devices such as C-arms and surgical navigation systems, enabling real-time edge computing and rapid diagnostic testing at the point of care. The medical platform is equipped with ADLINK's MXM graphics module, which supports NVIDIA GPUs, accelerating image reconstruction and beamforming while significantly reducing data transmission latency. The GPU-enabled platform enhances image quality through AI, assisting physicians in detecting diseases and other abnormalities, thereby providing better healthcare to people. Compared to the MLB-3000, the MLB-3002 offers flexible expandability with two PCIe slots, allowing for the addition of functional cards tailored to specific applications, such as capture cards for video streaming and high-speed data transfer via 2.5 Gigabit Ethernet.

【High performance supporting multitasking, increased workloads, and simultaneous execution of devices】 The MLB-3100 medical box PC is equipped with Intel's 12th/13th generation Core processors for high-performance image processing. This powerful medical computing solution is designed to meet the demanding requirements of healthcare applications. Whether it's real-time image analysis, complex data integration, or operating intricate surgical navigation systems, this medical box PC delivers seamless performance. 【Compatible with ADLINK MXM GPU modules featuring NVIDIA Ampere architecture】 The MLB-3100 is compatible with ADLINK MXM modules, which accelerates image and video processing while significantly reducing data transfer latency and wait times. Additionally, it provides high-quality images that enhance AI inference diagnostics, assisting doctors in detecting diseases and other abnormalities.

【High Performance Supporting Multitasking, Increased Workloads, and Simultaneous Device Execution】 The MLB-3102 medical box PC is equipped with Intel's 12th/13th generation Core processors for high-performance image processing. This powerful medical computing solution is designed to meet the demanding requirements of healthcare applications. Whether it's real-time image analysis, complex data integration, or operating intricate surgical navigation systems, this medical box PC delivers seamless performance. 【Enhanced Functionality with Diverse Expansion Options】 The MLB-3102 medical box PC features two PCIe Gen 3 x4 expansion slots for full-height, half-length add-on cards. These expansion slots provide users with the flexibility to add various accessories, such as video grabbers. This capability allows the medical box PC to adapt to a wide range of medical applications, from image capture and processing to data collection.