5G has become one of the most anticipated technologies for manufacturers in the two recent years. This technology is regarded as being especially crucial for the coming era of enterprise private networks and is also considered to be a key factor for the next-generation manufacturing systems. Chia-Wei Yang, director of the IoT Solution & Technology Business Unit at ADLINK Technology, said that manufacturing systems will take on a variety of additional functions by integrating existing smart infrastructure with enterprise private networks. This will help companies reduce costs, enhance their efficiency, and increase the safety of manufacturing.Unlike the traditional focus on the consumer market of the 3G and 4G standards in the past, 3GPP took note to integrate corporate scenarios along with consumer use in setting the 5G standard, approving three sets of specifications for 5G communications: eMBB, URLLC, and mMTC. Ultra-reliable low-latency communications (URLLC) and massive machine-type communications (mMTC) are considered to be most relevant to manufacturing environments. ADLINK has recently started working with partners in different fields to aid companies in building a smart system based on enterprise private networks.Yang further analyzed that 5G enterprise private networks can be divided into two types. One relies on infrastructure built by telecommunication companies, while the other is based on custom-made infrastructure built by a corporation. These two each have their unique pros and cons: the private network relying on 5G infrastructure from telecommunication companies is likely to be fast to deploy and easy to use, but comes with a hefty price tag. On the other hand, a corporation's own custom-built private network will take more time to deploy, but with lower costs and higher customizability, allowing the network to be tailored to the needs of a corporation. ADLINK has established itself in both types of networks. For the first one, the company has been working with major telecommunication providers for many years; with regards to the second one, ADLINK has also started working with partners in different fields. Recently, ADLINK teamed up with AU Optronics, Foxconn Technology Group, and prominent industry equipment vendor Fair Friend Group to build a 5G private network smart manufacturing system.ADLINK's 5G product strategy is based on the idea of "P5G+X," meaning that products will be based on private 5G with the addition of various smart solutions to customize products for different scenarios. This strategy directed ADLINK's cooperation with AU Optronics and Foxconn. According to Chia-Wei Yang, while AU Optronics provided the testing site and Foxconn built the 5G private network infrastructure, ADLINK was in charge of the various edge devices for smart manufacturing. This includes Internet of things (IoT) devices that retrieve data from terminal equipment, as well as AI platforms that perform image recognition, all of which are connected and controlled via the wide area connectivity and low latency of 5G networks.Yang pointed out that 5G private networks are undoubtedly, one of the key components for realizing the future vision of smart manufacturing. However, since the manufacturing industry is vast and complicated, there will be a large number of system builders coming from different backgrounds and possessing different expertise. These teams of builders will each require varying levels and types of assistance. Chia-Wei Yang divided the teams of system builders into three categories: teams with expertise in a specific field, teams in charge of AI system integration, and teams made up of internal personnel from the enterprise. ADLINK will be able to provide software and hardware assistance services to all three types of system builder teams.Automated optical inspection (AOI) companies are representative examples of teams with expertise in a specific area. They are highly capable in terms of both software and hardware deployment and ADLINK's complete product lines will be the perfect one-stop shop for their hardware supply. AI system integrators are usually focused on algorithms and lack knowledge in hardware, as well as specific expertise areas needed for deployment. ADLINK has years of experience in vertical industry integration and a thorough product lines, capable of making up for their shortfalls in those areas. As for teams made up of internal personnel, they are usually tasked with solving pain points within the production lines. Problems in the production lines tend to involve a wide range of issues and usually have no apparent starting point where the team in charge can start to solve the problem. ADLINK can help by clarifying the problem and guiding the establishment of SOPs and processes to gradually lead clients to integrate the equipment needed to construct an optimized manufacturing system.ADLINK's software and hardware products have already found great success in the industry. Recently, a client managed to apply ADLINK's solutions to ensure worker safety on the production line by integrating AI machine vision platforms with deep learning algorithms. The result was a system that could recognize whether the staff is wearing the appropriate clothing and accessories, check whether product assembly is following SOPs, and enforce virtual barriers that protect the safety of both machines and men in environments in which both are present. ADLINK has also worked with an AI system integrator to build an AOI solution based on artificial intelligence that helped raise the detection rate of flaws on contact lenses to 95%. Chia-Wei Yang concluded that the above mentioned cases are only a few examples of the many projects that ADLINK has worked on. As 5G private networks further accelerate the trend for "P5G+X" products, many more smart manufacturing applications are sure to appear. ADLINK will continue to focus on the R&D of related technologies, as well as seek cooperation with partners in different fields to provide manufacturers with the most optimized high-efficiency systems.Chia-Wei Yang, director of IoT Solution & Technology Business Unit, IST-Connected Factory Business Center at ADLINK Technology

ADLINK Technology Inc, global leader in edge computing, launched the NEON-2000-JNX series, the industry's first industrial AI smart camera that integrates the new NVIDIA Jetson Xavier NX module. The new camera's high performance, small form factor and ease of development open the door for innovative AI vision solutions in manufacturing, logistics, retail, service, agriculture, smart city, healthcare and life sciences, and other edge applications. The camera is an all-in-one solution, eliminating the traditional need for complex integration of the image sensor module, cables, and AI box PC.The NVIDIA Jetson Xavier NX provides more than 10X the performance of its widely adopted predecessor, NVIDIA® Jetson™ TX2. The NEON-2000-JNX series integrates this performance in an all-in-one rugged and compact device, designed to simplify the deployment process and speed up the time to market."Until now, a typical AI vision solution required a complex integration of the image sensor module, cables and GPU modules. This ready-to-develop edge AI smart camera reduces the effort of software/hardware integration and reliability validation, allowing AI vision developers to focus on application development. The NEON-2000-JNX series is a hassle-free, compact, reliable and powerful product for edge AI applications, and also the best match for AI software providers," said Kevin Hsu, Senior Product Manager of ADLINK's IoT Solutions and Technology business unit, ADLINK."AI-driven machine vision is reshaping entire industries from robotics and retail to healthcare, manufacturing and more," said Murali Gopalakrishna, Head of Product Management, Autonomous Machines and General Manager for Robotics at NVIDIA. "With the ADLINK smart camera kits powered by the Jetson edge AI platform, developers get an integrated, easy to deploy solution for AI vision-based embedded and industrial AIoT applications."The NEON-2000-JNX Series comes with all necessary components and an optimized OS already integrated and well validated.*Supports six sensor configurations in total between 1.2M-8M to deliver raw data and complete image detail for machine vision, including four image sensors with Basler.*Two new MIPI image sensors reduce CPU loading and support a higher operating temperature range.*An embedded Image signal Processor (ISP) provides enhanced and environment-adaptive imaging to improve AI accuracy.*Integration overcomes EMC/EDS/vibration/thermal problems, interface compatibility, image drops caused by fault camera and OS settings, and other common reliability issues.The NEON-2000-JNX series AI smart camera is pre-installed with ADLINK's new edge vision analytics software, EVA SDK (Edge Vision Analytics Software Development Kit), creating an optimized platform that reduces proof-of-concept (PoC) schedules and speeds time to market.*Wide selection of field-ready application plug-ins and ADLINK-optimized AI models guarantee AI vision quality and simplify building AI vision applications with limited coding required.*Pre-view function makes the verification of AI Inference flow and results quick and intuitive. *AI developers, even newcomers to AI, can focus on the application and training and build a proof of concept in as few as two weeks.Adlink launches industry-first Nvidia Jetson Xavier NX-based industrial AI smart camera

ADLINK Technology Inc, a global leader in edge computing, introduces two new modules in the 6-CH EU Series digital I/O – the ECAT-4XMO motion control and trigger module, and the ECAT-TRG4 trigger modules. The new modules are a dynamic extension to ADLINK's EtherCAT System and are designed to perfectly complement the EU Series, enabling higher performance in automatic assembly, test, and inspection equipment across a multitude of applications including cell phone glass inspection, battery cell assembly, camera lens assembly and test and glue dispensing machinery, as well as inspection equipment.The ECAT-4XMO motion control and trigger module is a 4-channel control and high-speed trigger module that connects to both EtherCAT and non-EtherCAT motors, including stepper motors, linear motors, direct drive and AC. Advanced functions of the ECAT-4XMO include high precision motion interpolation, continuous contouring, axes synchronization, point table for circuit limit and 2D compensation. The superior performance of the ECAT-TRG4 includes high speed position comparison trigger (4 channels, 10 MHz linear/table), pulse outputs (12 MHZ), encode inputs (20 MHz) and support for encoder re-driver operation. The flexible expansion interface of the ECAT-TRG4 allows easy connection to cameras and air guns."Industrial automation equipment, depending on different usage conditions, needs to control pulse-type motors and communication-type motors at the same time, as well as trigger control based on different signals," said Simon You, Director, Smart Machine Business Product Center, ADLINK. "Generally, it takes separate systems to control different types of motors, which consumes larger floor space, and increases maintenance efforts and costs. With ADLINK's highly integrated and compatible EtherCAT solutions, users are allowed to control pulse motors and communication motors via a single system at the same time, fulfilling the demands of many trigger control applications."In addition, ADLINK offers an all-in-one automation software (APS SDK), which serves as a uniform interface to access all ADLINK machine automation products. With APS SDK, users can operate ADLINK motion products in a simple and consistent way, minimizing the reprogramming hassles of adding or removing devices. APS Function Library also allows existing motion control users to easily upgrade to EtherCAT-based solutions without complicated program modifications or application redevelopment.The new modules feature sliding side-by-side installation and screwless design, to reduce installation costs, and support nearly all field applications in a smart factory. In test applications, the EtherCAT system delivered total cost of ownership (TCO) savings of up to 10% and up to 30% increases in productivity.ADLINK launches new EtherCAT modules, completing the EtherCAT solution for industrial automation

The back-end semiconductor manufacturing process refers to the IC packaging and testing that people often hear about. Specifically, the process known as chip probing (CP) is conducted to test the electrical characteristics of each die on the wafer, so that dies with faulty electrical performance can be vetted before packaging, effectively reducing unnecessary manufacturing costs from faulty products.Once probing is complete, the second process is called sorting, which is a quality inspection process. The goal of sorting is to quickly distinguish the dies that meet the desired specifications from those that don't. For products that pass the sorting process, the manufacturer may have them further binned into levels 1, 2, 3, or even more levels to match the specifications needed in the production of various electronic components, while defective ones can then be subjected to follow-up repairs and adjustments.Ray Lin, ADLINK Smart Machine Product Center Manager, said that increasing sorting efficiency is without a doubt the main goal for semiconductor manufactures looking to speed up packaging and testing to quickly and accurately distinguish good dies from defective ones. The reality of the situation, however, is that speed and accuracy are often opposing goals, in which meeting one means sacrificing the other. This is the biggest conundrum that many manufactures have to face.To solve this, ADLINK developed a die sorting solution that integrates both image acquisition cards and motion control cards. This solution allows end users to easily achieve capabilities like multiaxial integration and rapid feedback response through a single coding process, striking the optimal balance between speed and accuracy.More units to be tested create demand for higher motion control efficiencyLin pointed out that several pain points exist in semiconductor sorting, with the first being the complexity of simultaneous multiaxial control. During normal sorting operations, a robotic arm will pick the die up, rotate to the destination zone, then unload the die down to the tray. This process not only involves x-axis and y-axis movements between the dies and the tray, but also z-axis movement owing to the lifting and unloading motion of the arm. Consequently, the sorting machine will have to control 6 to 8 motion axes to guide the movement of arm, the suction head, video input, and other sub-systems to integrate into different work areas. To have a traditional PLC-based control system perform these tasks would involve an extremely high level of coding complexity.Second is the fact that the commonly seen traditional PLC-based infrastructures lack flexibility and usually do not support video input. If an end user wishes to achieve fast and accurate positioning on such system, then a separate video system needs to be purchased and arduously integrated into the system, greatly ramping up development costs.Yet another issue will be caused by the demand for higher testing efficiency due to the increase in the number of units to be tested. Take this year (2021) for example, as smaller and smaller dies become more in demand, wafer manufacturing processes have also evolved, resulting in 3 to 4 times the number of dies that can be cut in comparison with that of the past. This has in turn translated into a much higher number of units to be tested. Motion control efficiency must therefore also increase to match previous production speeds, presenting a challenging task.Fusing video and motion control to easily achieve multi-work station integrationFeynman Lin, Technical Director of the ADLINK Smart Machine Product Center, said that ADLINK can provide components like image capture cards and motion control cards under a unified PC-based infrastructure to give them exceptional compatibility and usability. Furthermore, since the cards share a unified interface, end users are also much less likely to make errors during operation. More importantly, ADLINK's sorting solution is built ready-to-use and based on the premise of integrating machine vision and motion control into a single system, meaning that end users can easily perform further customization to create a single program that can coordinate between visual input and simultaneous multiaxial motion control. Not only does this help reduce development costs, but having the steps of visual location, suction loading, arm rotation, and unloading be performed as a continuous motion can contribute towards further raising packaging and testing efficiency.It is also worth noting that as the industry moves towards the age of the Internet of things (IoT), factory operations are also starting to rely on more types of ports for greater connectivity. International interface standards like USB and Ethernet jacks, are being used for connectivity to various sub-systems like machine vision, temperature sensing, and pressure detection. On this front, PC-based solutions undoubtedly provide superior flexibility and compatibility compared to traditional PLC-based industrial controllers, making PC-based solutions better at meeting clients' various needs for precision equipment.Lin stressed that ADLINK's sorting solution includes three other major functions that make it the optimal choice for solving pain points in the sorting process. First, the solution can quickly adjust itself based on real-time feedback of the detected die surface (e.g., unevenness of the surface) to reduce sorting time by 20% to 30%.Second, the solution has an "instant trigger" function that combines machine vision and multiaxial precision positioning compensation technology to swiftly sort out faulty products from non-faulty ones with a precision of up to 30nm, and separate the dies according to their levels and place them into corresponding trays. Lastly, ADLINK's sorting solution supports automated sorting processes that can be quickly adapted to different pre- and post-processing needs. For example, this solution can be made to transfer an object in the "NG die" tray to the repair station or send an object from the "Level 1 qualified die" tray to the corresponding testing machine to verify that the die truly meets Level 1 specifications. This is the main reason why ADLINK's sorting solution has been well received by end users.For the past 25 years, ADLINK has been pushing the boundaries on new solutions for the semiconductor industry as many clients have become reliant on ADLINK technologies. The Company has become the best equipment partner for major semiconductor players in the Greater China region. ADLINK's comprehensive sorting solution combines machine vision and motion control for faster, more accurate, and more productive solutions in preparation for future market demands.Ray Lin, ADLINK Smart Machinery Product Department Manager, said that increasing sorting efficiency is without a doubt the main goal for semiconductor manufactures looking to speed up packaging and testing to quickly and accurately distinguish good dies from defective ones.ADLINK can provide components like image capture cards and motion control cards under a unified PC-based infrastructure to give semiconductor exceptional compatibility and usability.

Semiconductor chips have been the hottest trending topic recently due to their indispensability in all types of applications, including automobiles, mobile phones, military weapons, and space technology, etc. We can't help but wonder, how do semiconductor plants produce chips with such a large variety of functions out of large silicon wafers? The entire procedure involves an extremely important back-end packaging process known as "die bonding," which is to remove the die from the silicon wafer, and then attaching the die to the IC substrate using a conductive medium such as epoxy or gold wire (with epoxy the most common).With the evolution of the semiconductor miniaturization process, an increasing number of die can be cut from a silicon wafer. The LED die, for example, has evolved from an original die size of about 1mm to micro LED dies of just 100μm or even only 30μm. This means the number of pick-and-place motions in the die bonding process have also undergone a sharp increase, posing severe challenges for manufacturers seeking to maintain operating efficiency and positioning accuracy. To address this issue, ADLINK continues to introduce die bond solutions that are faster, more accurate, and offer easier setting of the dispensing trajectory to fulfill the needs of customers in the semiconductor industry.Supports multiple I/O channels to meet various needs for motion controlFeynman Lin, Technical Director of the ADLINK Smart Machine Product Center, stated that ADLINK's die bond solutions feature an extremely convenient development interface which allows users to easily set the dispensing trajectory and fully supports linear, circular, spiral or mixed trajectories of various shapes while also providing extremely efficient error compensation. More importantly, once the trajectory has been set, users can perform "trigger" control using check list mode, which means the computer will confirm that the correct position has been reached based on sensor information such as temperature, position and weight supplemented by the camera's precise positioning comparison. The adhesive dispensing valve will only be triggered after all conditions are met, enabling precise amounts of adhesive to be dispensed according to user settings. In other words, ADLINK's die bond solutions can be used with a great variety of I/O channels to fully meet the check list requirements that users anticipate.Ruey Wang, Product Manager at ADLINK's Smart Machine Product Center, added that with the advancement of semiconductor manufacturing processes, die size has shrunken exponentially, requiring an increasing number of pick-and-place motions. As such, users as a matter of course have come to expect accelerated pick-and-place sequences that offer the same positioning accuracy. To this end, ADLINK's die bond solution supports "high-speed P2P" and uses closed-loop motion control technology to speed up the tuning speed, suppress resonance or jitter effects in the shortest time, and ensure that adhesive can be dispensed as soon as possible once positioning has been completed.As for why it is necessary to use a camera to assist with positioning, Ching-pang Lin explained that even with more advanced general sensors, positioning accuracy can reach only about 2mm while the process takes 10ms to complete, making it difficult to meet user requirements in terms of both accuracy and speed. In contrast, cameras can reach an accuracy of 0.1mm, and the position can be obtained immediately after shooting, offering a more immediate and convenient positioning method. To facilitate camera applications, ADLINK offers PCIe-GIE74 frame grabbers that can be used with all renowned camera brands. In addition to excellent image capture quality that guarantees no frame drop, this product also comes with real-time power control capability that provides outstanding energy saving efficiency. In addition, ADLINK also offers PCI-8258 DSP-based 4-axis advanced motion controllers, which are suitable for mounting on the motion axes that require high accuracy and immediate response.One particular challenge that users often encounter when performing die bonding is the inability to distinguish driving axles from driven axles, which can easily affect the control ability of the machine due to the resonance effect of the mechanism. ADLINK's die bond solution offers master-slave synchronous control, thus effectively avoiding interference between machines and mechanisms through frequency-domain response analysis technology, a feature that is becoming increasingly important. Over the past couple of years, an increasing number of manufacturers have been attaching two different ICs into the same chip. Apart from the need for increased alignment accuracy, this approach must moreover be supplemented with evasion functions or related processing mechanisms to avoid interference between heterogeneous components. As the perfect solution to this issue, ADLINK's high-speed P2P fast tuning technology and master-slave synchronization control capabilities not only help users achieve higher alignment accuracy and speed, but also completely eliminate interference variables.As the packaging process becomes more and more miniaturized and high-precision, ADLINK's die bond solution continues to evolve and advance, ensuring that the correct positioning can be quickly achieved and the precise amount of adhesive can be dispensed so that users no longer have to worry about short circuits and defects due to misaligned contraposition or adhesive dispensing.ADLINK's die bond solutions feature an extremely convenient development interface which allows users to easily set the dispensing trajectory and fully supports linear, circular, spiral or mixed trajectories of various shapes while also providing extremely efficient error compensationPhoto: CompanyADLINK's die bond solution continues to evolve and advance, ensuring that the correct positioning can be quickly achieved and the precise amount of adhesive can be dispensed so that users no longer have to worry about short circuits and defects due to misaligned contraposition or adhesive dispensingPhoto: Company

The recent global shortage of automotive chips has brought Taiwan's semiconductor industry into the focus of worldwide attention. The real cause of why Taiwan's semiconductor chips are in such great demand is due to their trusted, unwavering quality that's supported and assured by a series of rigorous verification procedures. The electrical measurement performed before integrated circuit packaging is one of the most important steps of verification. To complete this task, the engineering probe station system must be used in conjunction with a variety of test instruments, automatic test equipment and specially designed engineering verification system.The challenges of electrical measurement lies in the fact that the input of the silicon wafer measurement signals cannot be seen directly by the naked eye, but only through simulation. As such, how does the probe station generate a real-time response based on the analog signal generated from contact between the probe and the silicon wafer, thus directing the probe to continue performing the subsequent motion (such as a slight pause, or horizontal movement, etc.) while also ensuring that the probe head accurately and effectively touches the silicon wafer? Although this indeed constitutes a major challenge, it's something that can be overcome with good motion control.To address this issue, ADLINK has been closely following the evolution of semiconductor manufacturing processes over the years, and has developed a series of probe test solutions that offer the advantages of both high precision and high speed. These solutions can be used to simulate and control at the same time, thus enabling clients such as wafer manufacturing plants or IC packaging plants to directly begin probe test operations without needing to purchase additional servo controllers with related functions.Enhancing the efficiency of error compensation through closed-loop motion controlFeynman Lin, Technical Director at ADLINK's Smart Machine Product Center, states that as ADLINK has been working closely with test equipment factories and semiconductor factories for many years, the company is well versed in the challenges that may arise in the probe test process. Therefore, it has continued to refine the probe test solution and already perfected many key technologies. In addition to the above, the analog input control function has been integrated into the motion control card so that it can fully correspond to the corresponding instantaneous command control, thereby preventing the probe from moving abnormally. Lin further explains that as the silicon wafer is often not a neat, level plane, it is necessary to constantly adjust the position of the probe to avoid damage to the silicon wafer or probe head due to excessive contact.Ruey Wang, Product Manager at ADLINK's Smart Machine Product Center, goes on to add that whether before or after the positioning of the probe, it will need to perform various motion controls, including moving up and down continuously, shifting horizontally, or moving to the next point. Now, through its profound knowledge of motion control technology, ADLINK is able to efficiently meet these diverse and constantly changing needs.Furthermore, each silicon wafer has hundreds or thousands of points to be tested on which testing must be performed. Therefore, in order to achieve optimal timeliness, the probe must be able to continue moving quickly and efficiently. Resonance and jitter however are likely to occur at the moment of positioning, and measurement cannot be performed until after tuning has been completed. To minimize the lag resulting from time spent waiting for tuning to be completed, ADLINK has developed rapid tuning technology which allows the probe to move before the Z axis of the workbench under test (for silicon wafers) is stabilized, thereby increasing tuning speed and achieving the effect of resonance suppression. This is to ensure that the probe will no longer vibrate as soon as it is in place, and is able to immediately execute testing.Feynman Lin additionally indicated that the ADLINK probe test solution supports "closed-loop motion control" by continuously calculating on a rolling basis the average value of the error between the number of commanded steps and the actual number of steps. When the next command is issued, it can directly combine the number of commanded steps with that of error steps (if the original command is to move 5mm and the average error is 0.01mm, then the subsequent command is directly adjusted to 5.01mm). By directly carrying out error compensation within the same motion, error distance can be reduced even if the compensation is not 100% complete, and is thus greatly conducive to improving testing speed.Other technical features of the solution include "mixed use of multi-axis linearity and arc angle compensation" and "process interruption protection mechanism." In the previous usage scenario, the main reason is that the wafer carrier may move in X-axis, Y-axis or Z-axis interchangeably. The path of travel is a mixture of straight lines and curves. In complex situations like this, the ADLINK probe test solution can continue to perform route detection and error compensation step by step, regardless of whether each segment is a straight line or an arc without the need to pause for switching or causing breakpoints in the testing process. Process interruption protection, on the other hand, usually comes into play in scenarios of sudden power glitches or power failure. The state immediately before the interruption will be recorded, and the process will continue to be executed after power is restored without need to waste time by starting the process all over again.Also worthy of mention is the fact that ADLINK has developed an "Automation Product Software Development Kit" (APS SDK) based on the experience and product functions the company has accumulated over the years. The APS SDK can also be used in probe test applications to help semiconductor clients significantly accelerate the development and deployment of motion control programs.In conclusion, ADLINK will continue to closely monitor the evolutionary trends of Mini LED or MicroiLED backlight technology, 2.5D or 3D packaging technology, and even 2nm or 1nm process technology, as it continues to strive for excellence by launching probe test solutions that fulfill customer expectations with faster execution speed while maintaining high-precision performance.ADLINK has been closely following the evolution of semiconductor manufacturing processes over the years, and has developed a series of probe test solutions that offer the advantages of both high precision and high speedPhoto: CompanyHow does the probe station generate a real-time response based on the analog signal generated from contact between the probe and the silicon wafer, thus directing the probe to continue performing the subsequent motion while also ensuring that the probe head accurately and effectively touches the silicon waferPhoto: Company

ADLINK Technology, a global leader in edge computing, announced that the MECS-6110 edge server has been verified as an Intel Select Solution for Universal Customer Premises Equipment (uCPE) on CentOS, for deployment of a wide range of edge-based communication, networking and hosted services.For service providers, general purpose multi-access edge computing (MEC) edge servers are the platform of choice when deploying virtual network functions, such as SD-WAN. MEC systems process data at the network edge rather than a centralized data center, cutting latency, shortening response times and helping remove range and compute limitations.Paired with virtualization technologies such as network function virtualization (NFV), MEC edge servers such as the MECS-6110 eliminate the need for providers to employ multiple proprietary communication platforms to deliver network services, enabling those functions to be delivered by software. MEC systems can also reduce capital expenditure (CapEx) spending, as they operate on white-box, general-purpose processing platforms, and offer lower operating expense (OpEx) via cloud-based automation and management features."ADLINK has launched its MEC Server MECS-6110 edge server based on the Intel Select Solutions for uCPE reference design, offering network and communication service providers a verified networking platform for network function virtualization," said Julian Ye, director of networking and communications at ADLINK. "ADLINK's verified, open-standards-based MECS-6110 platform enables our customers to speed time-to-market by allowing them to focus their development efforts on unique next-generation applications for the edge of cloud and mobile networks. They also benefit from the work we have done with Intel to integrate the latest technologies into our edge server."Intel Select Solutions are pre-defined, workload-optimized solutions designed to minimize the challenges of infrastructure evaluation and deployment by providing a defined configuration foundation that delivers optimized, predictable results. These reference designs are validated by original design manufacturers (OEMs), certified by independent software vendors (ISVs) and verified by Intel. The MECS-6110 combines a powerful Intel Xeon D processor, an optimized software stack, and a high-performance white-box edge server designed for virtualized networking and other communication applications.MEC edge server systems sit on the extremities of a network, and must be able to operate in unfriendly, often challenging, deployment environments. The MECS-6110 is designed and built to withstand dust, shock, and vibration while also supporting an operating temperature range from -5 degree C to +55 degree C. The MECS-6110 has a 420mm system depth in a small footprint 1U chassis, with all I/O front access and dual FHFL PCIe expansion slots reserved for access to acceleration hardware such as GPU, FGPA, and QAT adapters.For more information about ADLINK's MECS-6110, visit the product page here. For more information about Intel Select Solutions for uCPE, visit the solutions webpage here.ADLINK MECS-6110 edge server verified as an Intel select solution for uCPE

ADLINK Technology, a global leader in edge computing, announced its edge server, MECS-7210, is officially validated as an "NGC-Ready" computing platform. In a configuration with two Nvidia Tesla V100 Tensor Core GPUs, MECS-7210 has been tested for functionality and performance of easy to deploy AI frameworks for machine learning and deep learning."NCG-Ready validation of the MECS-7210 helps our customers across industries accelerate their AI-enabled application deployments at the edge," says Julian Ye, ADLINK's director of networking and communications. "In the case of telecom for instance, customers can bring AI to advance an array of applications - from 5G Open RAN to edge datacenters, private networks, and MEC - they will benefit from continuously refined AI frameworks from Nvidia to address new opportunities and challenges with optimal performance, maximum system utility and a great ROI."The NGC-Ready validation process included extensive testing on the MECS-7210 for a wide range of applications for on-premise, cloud and edge deployments. ADLINK's validated MECS-7210 can help customers leverage the extensive range of GPU-accelerated software available for real-time intelligent decision making. Customers can easily migrate workflow compute environments including both hybrid and multi-cloud implementations, run software on bare metal servers or on virtualized environments, and maximize utilization of GPUs and portability. The validation also gives customers direct access to enterprise-grade support from Nvidia where they can leverage continuously optimized frameworks with the latest featuresADLINK's validated MECS-7210 can also help speed time-to-market of AI-based applications and services across many industries including telecommunications, smart cities, retail, manufacturing, autonomous vehicles, virtual reality and healthcare.Nvidia NGC-Ready-for-Edge servers

5G achieves a 20Gbps downlink peak data rate (20 times faster than 4G), a latency as low as 1ms (one tenth of 4G) and a connection density of one million devices per square kilometer (10 times that of 4G) with enhanced mobile broadband (eMBB), ultra-reliable and low latency communications (URLLC) and massive machine type communications (mMTC). On top of that, it supports network slicing, which allows different levels of services to be provided based on user requirements. These advantages make 5G an enabler of cloud computing, IoT, big data analytics, AI, IoV/autonomous driving, AR/VR and a diversity of other innovative technologies or applications.Changes in the radio access network (RAN) are a major item among what make the 4G to 5G transition different from the 3G to 4G transition, which only brought faster data transfer speeds. 5G communication has refined and separated multiple network functions to enable a more flexible network structure so that different levels of services or functions can be provided and network virtualization and slicing can be effectively implemented. Moreover, a baseband unit (BBU) is split into a centralized unit (CU) and a distributed unit (DU), which drives the miniaturization of base stations and allows the use of white-box equipment, helping telecom operators reduce capital expenditure.Multi-access edge computing (MEC) will play an instrumental role maximizing all the above-mentioned 5G benefits. In view of this, ADLINK, with long-standing leadership in edge computing, industrial PC, data acquisition (DAQ), machine vision and motion control technologies and long-term devotion into vertical OT solutions including smart factory equipment, is actively developing MEC platforms while engaging in partnerships with AI, autonomous mobile robots (AMR), autonomous driving, automated optical inspection (AOI) and smart healthcare solution providers to jointly form a complete 5G ecosystem and expedite advances of 5G-enabled smart applications across vertical industries.Qianqian Shao, product marketing manager of ADLINK's networking, communication and public business unit, emphasized that MEC is a critical element of 5G communication. It ensures that latency-sensitive applications such as industrial robots and self-driving vehicles can operate efficiently by performing instant data processing and analysis close to the source to allow decision-making in real time. It is no longer required to upload data to the cloud platform or data center so as to prevent risks resulting from data transmission latency and largely save network bandwidth.Joining forces with 5G ecosystem partners to accelerate private network applications in diverse use scenariosFeaturing small footprint and wide operating temperature range, ADLINK's MEC platform is capable of supporting core computing needs within the 5G network architecture while enabling GPU and FPGA hardware acceleration. It also implements the IEEE 1588 Precision Time Protocol to support high-efficiency AI inferencing and clock synchronization throughout IP networks. System integrators can build applications/services to address all kinds of critical tasks on top of ADLINK's MEC platform.Shao added that ADLINK collaborates with ecosystem partners targeting a variety of application fields by providing hardware equipment and middleware. They have jointly built multiple application-ready platforms to help enterprises bring their private network applications to reality. For example, the Robot Operating System (ROS) running on optimized MEC platforms coupled with edge AI inference and low latency 5G communication accelerates AMR development and deployment and even achieves swarm autonomy. The AMR can carry out smart logistics tasks in factories, warehouses, hospitals, retail stores, restaurants and a slew of other application scenarios.Furthermore, leveraging its strength in edge computing and AI accelerators, ADLINK engages in close collaborations with multiple autonomous driving solution providers including autonomous driving software developer Tier IV and intelligent vehicle platform AutoCore to make use of 5G communication's low latency and massive connectivity features to develop high-performance and high-reliability autonomous driving applications.Making use of 5G communication's high-bandwidth, low latency and massive connectivity features, ADLINK's MEC edge computing platform will also expedite the advent of fully automated and unmanned factories as it helps connect diverse processes and workstations in the factory including AI-enabled AOI, AI-based process monitoring (operator behavior detection and analysis as well as object recognition), production equipment monitoring and diagnostics as well as material transportation by AMR.Apart from being used on 5G private networks, ADLINK's MEC platform can also be connected to public networks as MEC is needed on both types of networks but only with differences in how the platform is set up. For private networks, ADLINK's MEC platform serves as a DU to support communication access while acting as an application-ready platform. ADLINK and partners have worked together to help customers quickly build applications for specific use scenarios. For example, multiple smart factory proof-of-concept (PoC) projects in collaboration with several solution providers have been completed. For public networks, ADLINK's MEC platform also functions as a high-performance DU and provide edge computing capability. More than that, it fully supports data acquisition, transmission and processing under the premise that information security is guaranteed. ADLINK's MEC platform will accelerate the development of smart applications, bringing the futuristic scenes in sci-fi films to real life.More information on ADLINK's MEC platform: https://www.adlinktech.com/en/Edge_ServerQianqian Shao, product marketing manager of ADLINK's networking, communication and public business unit

Digital technologies such as 5G, big data, and AIoT have driven the continuous update of displays and smart applications with reach extending to many vertical markets, one of which is the key market of transportation. During "AUO Tech Forum 2020" hosted by AU Optronics Corp, a dedicated subforum was held for "smart transportation."Huge potential in rail transportation is a new blue lake for digital commercial displaysEric Su, Director of the Public Information Display Division at AUO stated that the company has invested heavily in transportation over the years and has gradually identified a "blue lake" in rail transportation. The company will continue to develop business opportunities in relevant infrastructure and focus on working with companies dealing in passenger information systems (PIS) in the future.Eric Su emphasized that PIS is one of the fields that best exemplifies the commercial value of displays. Using the example of light boxes which were previously dominated by mono color LED, the company is seeing more and more LCD products being implemented. However, traditional LCD products attempting to enter the field of transportation face many challenges such as space restrictions, harsh environments, and high customization requirements.Despite the many challenges, these hurdles are perfect for showcasing the capabilities of AUO's display technologies such as maximum support of 2,500 nits, readability under direct sunlight, and compliance with new trends such as wide temperature range and green energy. More importantly, AUO has developed the industry leading and revolutionary technology TARTAN and the ability to offer stretched display panels as well those with special aspect ratios such as square and round shapes that can meet client requirements with diverse customizations.Dr Andy Cheng, Industrial Technology Research Institute Deputy Director of the Vehicle Information and Control Systems Group, dissected the current status and outlook of self-driving buses. He mentioned while many believe that displays are not necessary in self-driving vehicles, the opposite is true; as self-driving vehicles cannot be entirely driver-less when initially entering roads, accompanying staff require display panels to keep up to date with dynamic vehicle information.In the field of self-driving vehicles, buses have great potential due to their fixed routes, which can reduce environmental complexity. Therefore, the Industrial Technology Research Institute delved into self-driving buses in specific Level 4 areas and began closed testing in the first half of the year; Soon, they will work with the Hsinchu County Government in Taiwan to launch self-driving shuttle buses from Hsinchu THSR station to Sheraton Hsinchu Hotel and perform the nation's first self-driving tests on open road environments.Smart Internet of Vehicles greatly enhances operation efficiency of fleetsADLINK Director of Internet Communications and Infrastructure Business Dr. Henry Hu gave a talk about "Self-driving Vehicles are Simple." He mentioned that based on U.S. statistics, traffic consumes the amount of time equal to the lifespans of 162 people making it clear that traffic has a significant impact on a country's competitiveness and that self-driving cars can ease this issue; he points out that self-driving technology is not far away in the future but already present in everyday life. Farmers in the U.S. and Japan use driver-less vehicles to harvest crops, spray pesticides, and turn soil while closer to home, the MRT Wenhu line in Taipei has also been self-driving since 2016.Many in the public believe that self-driving technology is challenging mainly because while Waymo, a self-driving technology company obtained their license for driver-less vehicles in 2012, there has been no massive commercial conversion; on the other hand, the driver of a renowned electric vehicle brand activated the autopilot function on the freeway and crashed into a truck that was laying on its side, showing that even with Level 2 self-driving there are many challenging aspects. Dr Henry Hu believes that if we choose "limited areas" and reduce complicated variables, such as dedicated lanes for self-driving buses, it's possible to massively reduce the obstacles to viability.ADLINK purchased a company specializing in IoT data distribution service (DDS) and made the technology open source. This became the communication architecture for the robot operating system ROS 2.0 while Autoware, the world's largest open source software for self-driving vehicles is also based on ROS architecture. ADLINK has forged a series of ecosystem partners in self-driving technology and can provide comprehensive hardware/software integration to help clients compete for self-driving opportunities in limited areas.Sintrones Sales Manager Derek Ho introduced in-vehicle computing solutions. The development of in-vehicle computing is closely related to internet technology. The 3G era saw computers integrated with 3G, GPS, and Wi-Fi with support for diverse applications; during the 4G era, we began to receive video that satisfied requirements for surveillance and recording; as we enter the 5G era, we can not only capture video but also combine them with AI modules for inferential analysis.The key to self-driving technology, whether we're between Levels 2~5, is data collection. When looking at the 3 major fields of application in C-V2X, this defines the data collection requirements before, during, and after an accident; this requires in-vehicle computing to connect with radar, cameras, GPS, or 4/5G data to ensure there are no communication obstacles between the driver, vehicle, networks, and mobile base stations.AUO Senior Director of Transportation Solutions Jackson Weng provided details on the value of AUO's internet-of-vehicles solutions. He pointed out that network connectivity makes vehicles smarter and increases operational benefits. AUO is currently developing an AIoV total solution package for internet-of-vehicles that encompasses cloud-to-ground, software and hardware, AI and vertical applications that provides the level of customization that satisfies the management requirements of vehicle operators.The AIoV cloud platform provides a wealth of management tools and AI added-value functionality that can connect the vehicle, dashboard, PIS, sensors, cameras, and CAN bus. The system is supported by high brightness AUO vehicle displays with support for TARTAN technology, providing operators with 5 major benefits including monitoring and optimization, fuel or battery consumption, and reducing operating costs; the platform collects driving records for different periods and routes to use as a basis for analysis for load efficiency or fuel consumption; monitor emergency braking activity performed by the driver and activate remote control at any time; monitor health status of sensors and parts for early repair; enhance safety by monitoring driver activities such as use of smartphones or falling asleep to reduce the costs of insurance and traffic tickets.Looking towards the future, the opportunities of self-driving vehicles, buses, rail transportation, and internet-of-vehicles will inevitably grow with limitless potential; For this end, AUO has provided advanced vehicle display technologies, forged alliances, and called on our partners in various fields to collaborate and satisfy the demand for application development in smart transportation.AUO TARTAN displays are available in diverse ratios and sizes and can be applied in the display of transportation informationAUO and ADLINK demonstrate AIoV and self-driving solutionsAUO Director of Public Information Display Division Eric Su explains how display technology is one of the keys of smart transportation systems