electronics-journal.com
27
'26
Written on Modified on
Edge AI Platforms for Industrial Systems
SECO and Qualcomm Technologies combine Dragonwing platforms and Clea software to deploy scalable edge computing solutions for industrial automation, smart grid, and advanced HMI applications.
www.secotools.com
SECO and Qualcomm Technologies are collaborating to deploy Qualcomm Dragonwing-based edge computing platforms within industrial automation, smart building systems, and energy infrastructure. The cooperation integrates Qualcomm’s system-on-chip (SoC) architectures with SECO’s embedded hardware, software framework, and lifecycle management tools to enable cloud-independent AI processing at the industrial edge.
Context of the Cooperation
SECO develops embedded computing modules, industrial HMIs, and IoT software frameworks for OEM customers. Qualcomm Technologies provides high-performance, energy-efficient SoCs under the Dragonwing portfolio, targeting edge AI and industrial applications.
Industrial OEMs increasingly require local AI inference for machine vision, predictive maintenance, and real-time control while maintaining deterministic behavior, cybersecurity compliance, and long product lifecycles. The integration of heterogeneous processing (CPU, GPU, NPU) with industrial form factors and operating system support requires coordinated hardware and software engineering. The partnership addresses this integration complexity.
Technical Platforms and System Architecture
At the core of the collaboration is SECO’s SOM-COMe-CT6-Dragonwing-IQ-X, a COM Express Type 6 module based on the Dragonwing IQ-X Series. The module integrates 8- or 12-core Oryon CPUs (up to 3.4 GHz single-thread performance) and delivers up to 45 TOPS of on-device AI acceleration. It supports dual ISPs, multiple camera inputs, up to 2×5K or 4×4K displays, and high-speed interfaces including USB4, PCIe, and UFS 4.0.
Designed for industrial automation environments, the module operates from –40°C to +85°C and supports Windows 11 IoT Enterprise LTSC. Fanless system designs are enabled through optimized power efficiency, reducing mechanical complexity and maintenance requirements.
SECO complements the hardware with board support packages (BSPs), integration services, and its Clea software framework. Clea provides secure device provisioning, OTA updates, data management, and remote fleet control, forming a structured digital infrastructure from edge device to cloud.
Roadmap extensions include SMARC modules based on Dragonwing IQ8 and COM-HPC Mini platforms based on IQ9, targeting higher compute density and AI workloads.
Deployment in Industrial and Energy Applications
The cooperation spans multiple deployment scenarios:
- Industrial HMIs based on Dragonwing QCS6490 integrate computer vision and voice processing for plant management and AI-based quality inspection with local inference and ultra-low latency.
- SMARC platforms using QCS6490 and QCS5430 serve as embedded building blocks for vision systems and constrained environments, including modular AI systems performing real-time object classification.
- Modular Link IQ-615, a DIN-rail gateway based on the Dragonwing IQ-615 Series, supports smart grid and industrial infrastructure applications. It integrates industrial I/O, TPM 2.0, eMMC storage, Linux Yocto-based Clea OS, operates from –40°C to +80°C, and maintains power consumption below 20 W.
End-to-End AI Enablement
SECO integrates Edge Impulse-based model training with Clea deployment tools, creating a unified AI pipeline from model optimization to secure edge execution. AI workloads are executed locally on Dragonwing NPUs, while monitoring and updates are centrally managed.
The cooperation provides OEMs with a pre-integrated hardware and software stack designed to reduce system validation time, ensure cybersecurity compliance, and enable scalable industrial edge AI deployment across automation and energy infrastructure environments.
www.seco.com
SECO integrates Edge Impulse-based model training with Clea deployment tools, creating a unified AI pipeline from model optimization to secure edge execution. AI workloads are executed locally on Dragonwing NPUs, while monitoring and updates are centrally managed.
The cooperation provides OEMs with a pre-integrated hardware and software stack designed to reduce system validation time, ensure cybersecurity compliance, and enable scalable industrial edge AI deployment across automation and energy infrastructure environments.
www.seco.com
