wenglor sensoric - AI Lab - How do I deploy AI models to uniVision 3 via the AI Loop (weHub)?

In this tutorial, you'll learn how to deploy a trained AI Lab model directly into uniVision 3 using the AI Loop via weHub. This seamless integration allows you to synchronize your trained models and start AI-powered image classification in just a few steps.

Begin by opening uniVision 3 and adding the Image AI module to your project. Connect the module to your Input Image to prepare your inspection workflow.

Next, copy your AI Lab Dataset ID and paste it into the corresponding field in the Image AI module. This enables uniVision 3 to connect directly to your dataset stored in AI Lab through weHub.

Once connected, choose which model you want to deploy. You can load either the latest trained model or your primary model, depending on your workflow. If your dataset contains only one model, either option will load the same version.

After a few moments, the selected model is automatically transferred into your uniVision 3 project. A preview window displays important information about the model, including the input image, available class names, and image resolution.

Close the preview dialog, and the model is immediately ready for inference. Switch to Run Mode to view the live results and watch the system classify images in real time—for example, distinguishing between OK and NOK classes.

That's it! Your AI Lab model is now successfully deployed to uniVision 3 via weHub, enabling fast, reliable, and real-time AI-based image classification for your machine vision application.