Deploy AOI With AI Capabilities Through an On-Prem Next-Gen Edge Computer

November 30, 2023

Sponsored Story

Automated optical inspection (AOI) is a quality-control process used by manufacturers to detect and identify defects in products. It plays a crucial role in enhancing the efficiency and accuracy of quality control, reducing the likelihood of defects reaching the end consumer.

Traditional AOI systems employ advanced optics, cameras, and image processing algorithms to automatically examine and analyze the visual characteristics of manufactured items. These systems can inspect intricate assemblies with high precision and speed.

During the inspection process, the AOI system captures detailed images of the product on the assembly line and compares them to predetermined standards or reference images. It can identify defects such as missing components, misalignments, soldering issues, and other anomalies that may impact product functionality and reliability.

The benefits of AOI include increased production yields, faster inspection times, and the ability to detect subtle defects that may go unnoticed by human inspectors.

By taking advantage of some of the latest technologies, namely AI at the Edge of the industrial IoT (IIoT), the effectiveness of AOI can be significantly increased by enhancing efficiency, reducing latency, and improving overall performance. In most cases, the AI algorithms can be deployed directly on the inspection equipment, near the manufacturing process, rather than relying on a centralized cloud-based system.

Deploy Directly At the Edge

By deploying directly at the Edge of the IIoT, delays between capturing images and making decisions are reduced (nearly eliminated). And by processing all the data locally, there is less (zero?) need to transfer large amounts of image data to a centralized server for analysis. In addition to saving time, this reduces costs, as transmitting through a 5G or other cellular network can be expensive. It also reduces the need for Cloud-based computing equipment, further reducing costs and increasing reliability as all computing functions are handled with on-prem equipment that’s easier to maintain.

Other key reasons to deploy at the Edge include the following:

• Increased accuracy and precision
• Adaptability and flexibility, giving AOI systems the ability to make changes dynamically
• Fault Tolerance
• Reduced false positives and negatives
• Efficiency and speed
• Advanced analytics
• Scalability

In addition, with security always front of mind these days, Edge computing allows all data to remain on-site rather than being transmitted over networks. This enhances data privacy and security, addressing concerns about the potential exposure of proprietary manufacturing information during the inspection process.

Implementing AOI with AI at the Edge of the IIoT requires the integration of a few key technologies, including:

• An Edge computer
• High-resolution cameras
• High-speed connectivity
• AI algorithms
• Storage
• Security

The heart of the system is that Edge computer. The Mitac MX1-10FEP-D embedded computer and MP1-11TGS AI accelerator card can team up to enable the perfect combination.

There are various pieces to the AOI puzzle. A good foundation starts with the MiTAC MX1-10FEP-D embedded computer.

MiTAC At the Edge

The MX1-10FEP-D, which is designed with a higher density GPU to enhance AI performance and GPU computing, contains a powerful Coffee Lake-S Refresh Xeon microprocessor at its core. It employs a discrete graphics card and supports ECC DDR4 memory. The computer supports a maximum of 208 TOPS. It also supports an optional AI accelerator card, such as the MP1-11TGS, and has the chops to transmit a high-resolution triple display output. There’s also a host of other I/O options.

The MP1-11TGS integrates a GPU for efficient GPU and AI performance. One option is to integrate a Hailo VPU, which is done using an M.2 card, for a maximum performance level of 26 TOPS. Up to 32 Gbytes of DDR4 memory can be supported. A hot-swappable SSD/HDD slot is also available. It can monitor activities through Wi-Fi, or cellular (4G/5G) if needed and contains PoE ports to connect the high-res cameras. The accelerator card combines compact size with power efficiency to maximize AI performance. Various other I/O options are supported.

As manufacturing processes continue to evolve, AOI becomes one of the most critical tools in ensuring the high quality and reliability of products, ultimately contributing to improved customer satisfaction, brand reputation, and significant cost control. Having a manufacturer like MiTAC behind you solves most of your performance and integration needs. The company, with its experienced R&D team, operates globally and offers tech support through an FAE network available throughout the world. In addition, most of the company’s embedded computers can be customized and white-labeled for specific customers. Give them a call today.