From Prototype to Production

What is WEDA?

WEDA (WISE-Edge Developer Architecture) is a software framework for developing, deploying, and operating Edge AI solutions.

Combine WEDA capabilities with your application and field expertise. Select the image to enlarge it.

WEDA provides shared integration and management capabilities, so your team can focus on application logic and AI development.

Beyond the Prototype

An Edge AI prototype demonstrates that an application can work. Production adds ongoing responsibilities:

• Connect equipment and integrate it with your solution.
• Prepare applications for the target hardware.
• Deploy updates as applications change.
• Monitor devices during operation.
• Maintain AI models over time.

These responsibilities continue as your project grows across devices and sites. WEDA provides a common framework for managing that work across development, deployment, and operation.

Managing the Edge AI Lifecycle

Your training tools can read device data from WEDA Cloud through its APIs. Register trained models with WEDA Cloud to deploy them to devices. Select the image to enlarge it.

WEDA supports several parts of this lifecycle:

Your task	How WEDA helps
Connect hardware and equipment	On WEDA Node devices, WEDA SubNode supports sensor and equipment integration. A Custom Device lets an external device or application send data to WEDA Cloud without installing WEDA Node.
Prepare and deploy applications	Advantech Container Catalog (ACC) offers Ready-to-use applications and Ready-to-develop projects that you build for your target hardware. WEDA Core deploys your container stacks to target devices.
Manage AI models	Register models, maintain editions, and deploy them to devices.
Work with device data	Query DataStreams, derive values with Analyzers, and forward data to other systems with Data Stream Bridge. Use Alert Policies to raise alerts from DataStream values.
Operate deployed systems	Monitor device status and telemetry, manage configuration, send remote commands, and use remote access for maintenance.

Application deployment is one part of this lifecycle. WEDA's value extends to the devices, models, data, and ongoing operations that keep an Edge AI solution useful after its first deployment.

How WEDA Fits Your Workflow

WEDA is modular. You can adopt the components that fit your project and integrate them with your existing workflow:

• Keep your tools. Develop applications and train models with your chosen tools.
• Choose your starting point. Begin with the components needed for your first device.
• Build on that foundation. Use the same management approach for application changes, model updates, and operational maintenance as your solution grows.