
Critical Link’s Edge AI Camera with MitySOM-AM62A Wins Best in Show at Embedded World North America 2026
Embedded Computing Design honors TI Sitara™ AM62A-based vision platform in the Computer Boards, Systems, Components & Peripherals category
SYRACUSE, N.Y. — October 22, 2026 — Critical Link, LLC, a maker of production-ready System on Modules (SOMs) and embedded solutions, announced that its Edge AI Camera with MitySOM-AM62A won a Best in Show award at Embedded World North America 2026. Embedded Computing Design gave the award in the Computer Boards, Systems, Components & Peripherals category. The show took place September 22–24 at the Anaheim Convention Center in Anaheim, California.
Embedded Computing Design’s content team judges Best in Show entries on a 15-point rubric covering design excellence, relative performance and market impact or disruption.

The Edge AI Camera gives engineers a flexible, scalable and cost-effective way to deploy vision-based AI at the edge. It’s built on Critical Link’s production-ready MitySOM-AM62A module and an application-ready carrier board. At the module’s core is the Texas Instruments (TI) Sitara™ AM62Ax processor. It combines quad-core Arm® Cortex®-A53 processors, a Cortex-R5F real-time co-processor, and a C7™ neural processing unit which includes a floating-point DSP of up to 40 GFLOPS and a Matrix Multiply Accelerator (MMA) of up to 2 TOPS.
“Engineers building vision AI into their products have been stuck with a hard tradeoff,” said Amber Thousand, Director of Marketing at Critical Link. “They could choose off-the-shelf imagers that need redesigning every few years, or high-performance compute platforms that break the budget. We built the Edge AI Camera to remove that tradeoff. It pairs the efficient AI performance of TI’s AM62A processor with a SOM designed for long-term production and backed by Critical Link’s engineering support. Customers can get to market first and know the platform will last as long as their design does. We’re proud that Embedded Computing Design recognized it.”
Production-Ready Vision AI, from Prototype to Deployment
Many edge AI platforms are built for evaluation, not volume production. The Edge AI Camera is designed so the hardware a team prototypes on can go straight into production. Developers can start with readily available AI models for object detection, facial recognition and segmentation. They can then use TI’s CCStudio™ Edge AI Studio to train and deploy models for their own parts, materials and environments.
Key features include:
- AI acceleration: The TI AM62A7 C7™ neural processing unit which consists of aDSP with MMA-accelerated deep learning and inference engine delivers up to 2 TOPS.
- Full image signal processing: A complete ISP and tuning stack provides debayering with edge enhancement, HDR color mapping and scaling, warp transforms such as rotation and lens distortion correction, resizing and white balance.
- Flexible camera interface: Supports up to four lanes of MIPI CSI and a 22-pin Raspberry Pi Zero camera interface. It ships with an IMX219-based Raspberry Pi camera.
- Video output and streaming: Outputs 1080p HD over micro-HDMI and streams H.264 over the network (RTSP) through Gigabit Ethernet.
- Simple power: Runs on Power over Ethernet (PoE) or USB-C power, so a single cable can handle installation.
Proven in Live Demos

At embedded world North America, attendees at Critical Link’s booth watched the Edge AI Camera run live defect detection. It uses the YOLOX-nano-lite defect-detection model on the AM62A’s C7™ accelerator, which runs about 65 times faster than on the Arm Cortex-A53 cores. The demo shows live video with visual indicators and production analytics such as total parts and defect rate. These are the capabilities that matter for factory automation, production yield and quality assurance.
A second demo uses the YOLOX-S-Lite model for real-time people tracking. It reports total visitors, current occupancy and visit duration, plus a heatmap of frequently visited areas. It’s aimed at retail, building automation, and safety and security.
“AI is changing how industrial and commercial systems see and respond to the world. Our customers need solutions that balance performance, power and cost,” said Sonia Ghelani, Sitara Product Line Manager at Texas Instruments. “The AM62A SoC was designed from the ground up for deep learning and advanced image processing at the edge. Critical Link’s Edge AI Camera shows how quickly developers can go from concept to a production-ready product on TI’s edge AI platform. Congratulations to Critical Link on this well-deserved recognition.”
Availability
Demo units of the Edge AI Camera with MitySOM-AM62A are available now. For a demonstration, pricing or a quote, contact Critical Link at info@criticallink.com or reach out to your local Critical Link manufacturers’ representative. Product details and the datasheet are at criticallink.com/product/edge-ai-camera-with-mitysom-am62a.
Read the full list of embedded world North America 2026 Best in Show winners on Embedded Computing Design.
About Critical Link
Critical Link designs and manufactures CPU-, FPGA-, and DSP-based system-on-modules (SOMs) for industrial electronic applications. Its production-ready embedded solutions help customers bring products to market faster and at lower cost by reducing development complexity, risk, and time spent building core processing subsystems from scratch. With a focus on product quality, long-term availability, lifecycle support, and close customer engagement, Critical Link serves OEMs across a wide range of industrial and technically demanding applications. For more information, visit the website: criticallink.com











Join Critical Link and the local 







