NVIDIA Jetson Orin Powers the MiTAC MA1 In Industrial Machine-Vision Applications

January 08, 2024

Sponsored Story

Using machine vision, an embedded computer can interpret and understand visual information from various inputs, including cameras and other sensors. In an industrial setting, this information is combined with image-processing algorithms to help make informed and extremely valuable decisions.

Initially designed to replace human visual capabilities, the technology has advanced far beyond that point today. And some of the latest introductions, such as platforms based on the NVIDIA Jetson Orin module, advance the technology significantly.

Pivotal Advances

Machine vision plays a vital role in enhancing efficiency, accuracy, and productivity across various industries by automating visual tasks and enabling machines to "see" and comprehend the visual aspects of their environments. The technology can automate tasks, improve quality control, increase productivity, reduce costs, and provide valuable data for process optimization.

Specific to industrial applications, machine vision systems are used for tasks such as quality control, inspection, and automation. These systems can identify and analyze patterns, colors, shapes, and textures, providing valuable information for decision-making and process optimization.

In terms of quality control and inspection, machine vision systems can perform rapid and precise inspections of products on production lines, detecting defects, irregularities, or deviations from quality standards. Consistent and objective assessments reduce the likelihood of human error in quality control processes. And productivity can be maximized thanks to automated the inspection tasks and the ability to operate 24/7/365, even for high-speed production lines.

Cost reductions can be realized by automating inspection processes, thereby reducing the need for manual labor in quality control. And early defect detection helps reduces waste and associated costs.

Processes can be optimized with machine vision through, for example, real-time monitoring and adjustment of manufacturing processes, and optimizing parameters for improved efficiency. And continuously collecting data helps identify trends and patterns, facilitating proactive maintenance and minimizing downtime.

Continuous Product Tracking

Machine vision systems can be used to track and trace products throughout the production process, ensuring compliance with industry regulations and standards. This is particularly important in industries like pharmaceutical and food, where strict traceability is required.

Such systems can be programmed and adapted for various inspection tasks, making them versatile for different industrial applications. The flexibility of the technology allows it to accommodate changes in production requirements and adapt to new product designs. Another key characteristic of machine vision is the enhanced safety that it provides, including safety monitoring, and ensuring compliance with safety regulations and identifying potential hazards in real-time.

Deploy the Latest Technologies

Machine vision systems have evolved significantly, even in last few years, with improvements in hardware, software, and algorithms contributing to their improved performance and capabilities. Some of the key differences of late include higher resolution cameras, improved sensors, and of course far more compute power, including what is available in the NVIDIA Jetson Orin, which can take advantage of the latest deep-learning algorithms, while integrating machine-learning techniques.

Comparing traditional x86 systems with those based on the latest NVIDIA platform bring up a host differences when deployed in industrial machine-vision applications. Take the GPU acceleration, for example. While x86 processors may have integrated graphics capabilities, they may not be as powerful or specialized for parallel processing tasks compared to dedicated GPUs. Whereas, the NVIDIA Jetson platforms are equipped with powerful GPUs designed for parallel processing of image data and deep learning.

The same goes for deep-learning capabilities; the x86 systems can clearly handle these tasks. However, they may require additional GPUs for optimal performance in compute-intensive applications, like machine vision. The NVIDIA Jetson Orin systems, on the other hand, are specifically tailored for AI and deep learning applications, including the deep neural networks commonly used in machine-vision applications.

From a pure performance perspective, it’s hard to quibble with what’s available in the NVIDIA Jetson Orin platform, which has the chops to handle generative AI at the Edge, a requirement for most machine-vision applications. In general, Jetson offers scalable software, a modern AI stack, flexible microservices and APIs, and an application-specific AI workflow. The modules provide up to 275 trillion operations per second (TOPS) and 8X the performance of the previous generation for multiple concurrent AI inference pipelines, plus high-speed interface support for multiple sensors. Jetson Orin modules are powered by the same AI software and cloud-native workflows used across other NVIDIA platforms.

The MiTAC MA1 compact system is designed with an NVIDIA Jetson Orin Nano module, providing the horsepower for industrial machine-vision applications.

One system that fits the criteria described above for industrial machine-vision applications is the MiTAC MA1, which is designed with the NVIDIA Jetson Orin Nano or NX SOM, for a maximum performance level of 100 TOPS at 25 W. This permits a fanless, compact design over a wide operating temperature, from -25°C to +60°C. It can handle up to 4k resolution at 60 Hz from an HDMI input. Other features include 2x RJ-45 1-GbE LAN, 2x USB 3.2 Gen2 Type A, high-speed M.2 2280 PCIe x4 NVMe, and support for 5G/LTE and WiFi-6E.

The MA1 can be the Edge-based backbone needed for industrial machine-vision applications, thanks to its compute power and vast I/O. It also has the reputation of MiTAC standing behind it, a company that has built a reputation for its global tech support through an FAE network, and experienced R&D team. Note that MiTAC is also willing to white-labeled its industrial platforms for specific customers.

Examples in the Field

The MiTAC MA1 has exhibited the qualities needed for Edge-based computing in a series of real-world applications, especially those that can take advantage of the computer’s AI capabilities. Specifically, thanks to the compute power provided by the Nvidia Jetson, the MA1 is highly effective in driving automated mobile robots (AMRs) due to its compact size and low power consumption. The energy-efficient Jetson platform allows for extended AMR operation without compromising computational power, making it ideal for industries like logistics.

In all-in-one (AIO) platform scenarios, the MiTAC MA1 with the Nvidia Jetson offers additional advantages. For example, the Jetson's high-performance GPU architecture accelerates AI processing, which is especially crucial when operating as an Edge computer enhancing the original box PC’s AI efficiency. This capability empowers industries with real-time decision-making, improving efficiency in dynamic environments.

Moreover, in robotics applications, the MiTAC MA1, when fitted with the Nvidia Jetson, is the preferred platform for its efficiency in handling intricate AI algorithms. Jetson's unparalleled computational power facilitates the execution of advanced AI tasks with speed and precision. For instance, in factory robotics, the Jetson-powered MA1 ensures that inspection robots can process visual data in real-time, enhancing precision and reliability during critical procedures.