What is Advantech Container Catalog?
Advantech Container Catalog (ACC) brings together Advantech application projects, deployable applications and partner offerings for edge devices. Use it to find software for your use case and check the hardware and software environment that an item requires.
The Container types filter helps you choose the next action: develop and build an application, deploy an existing application image, or evaluate a partner's trial or solution.
What ACC Provides
Read the item's type before following its setup instructions:
| Catalog type | What it provides | What you do next |
|---|---|---|
| Ready-to-develop (Ready-to-Dev) | An unbuilt application project for development or customization. | Follow the project instructions, make any changes you need and build the application image. |
| Ready-to-use | A built application that you can deploy after preparing its required environment and configuration. | Select the appropriate image version and prepare the deployment settings for your device. |
| Partner container | A partner trial or an introduction to a customer solution. | Review the solution and follow its trial, access or contact instructions. |
The catalog covers application development, AI workloads, device integration and industry solutions. These examples show the range of items currently listed; open each item to check its type and requirements.
| Area | Examples in ACC | What to look for |
|---|---|---|
| Hardware-specific application development | x86 WEDA Base Container, GPU Passthrough on NVIDIA Jetson | Development resources and hardware access for your application. |
| Local LLM inference | LLM Ollama on NVIDIA Jetson, Edge LLM Base via vLLM on NVIDIA Jetson Thor | The target platform, inference runtime and model setup. |
| RAG and AI agents | Deepseek-R1 Langchain RAG, Qwen2.5 3B AI Agent | Application examples for document retrieval and tool integration. |
| Computer vision | Intel OpenVINO Retail AI Self-Checkout, Qualcomm QCS6490 Object Detection | Input sources, supported acceleration and application configuration. |
| Camera and sensor integration | Orbbec-Gemini-3DCamera, Ouster Gemini | Required peripherals, drivers and integration examples. |
| Industry and partner solutions | Neurocle No-Code AI Defect Inspection, ENERZAi Optimium VTT Whisper | The offered solution, evaluation instructions and provider-specific requirements. |
Use Chipset vendors, Chipset platforms, Advantech hardware models and Device OS to narrow the list. The catalog includes items for platforms from NVIDIA, Intel, AMD, Qualcomm, NXP and other vendors. Compatibility is specific to the selected item and version, so check its Compatible hardware and prerequisites.
When to Use ACC
- You want to evaluate an existing application. Start with Ready-to-use. For example, the Retail AI Self-Checkout listing provides a built application for evaluating a vision-based checkout workflow. Check its device and display requirements before deployment.
- You need to develop or customize an application. Start with Ready-to-develop, read the project documentation and complete its build process. The project is a development starting point; prepare a runnable application image before moving to deployment.
- You are assessing a partner solution. Use the partner listing to understand its use case, available trial and access instructions. The provider's instructions determine how you evaluate that offering.
For the Ready-to-use route, How to use WEDA Edge AI Containers shows how to select an item, review its environment and prepare its files and settings for WEDA.
How ACC Fits into WEDA
ACC supplies application and development resources. WEDA manages deployment of the application image and its runtime configuration through a Container Stack and a selected revision. The target WEDA Node pulls the required images from their registries and reports deployment and container status.
Prepare the application's host files, environment values and hardware access before deployment. A Ready-to-use application still has item-specific prerequisites; a Ready-to-develop project first needs a build that produces the application image you intend to run.
If your application needs to send measurements or inference results into WEDA as device data, use the WEDA SubNode SDK to implement that integration. Application deployment and device-data integration have separate configuration and verification steps.