AVEVA Vision AI Assistant overview

The AVEVA Vision AI Assistant application uses images from existing general-purpose cameras to provide image classification-based analytics. AVEVA Vision AI Assistant provides secure connectivity with AVEVA™ Insight and integration with AVEVA™ System Platform and AVEVA™ Vision AI Assistant Operations Management Interface (OMI), enabling alarms and notifications. This solution is designed to work in low latency industrial environments. You can select between deep learning or machine learning models, as well as train and deploy the solution from an easy-to-use web-based interface: AVEVA™ Vision AI Web Client. When deployed, the solution monitors real-time image streams and provide alerts and notifications to operators.

The following diagram displays a visualization of the high-level architecture post-installed services:

1. Input the dataset (images or video).

2. Configure skills by using AVEVA Vision AI Web Client.

3. Publish to AVEVA™ Insight Newsfeed Notification System and publish values as tags to AVEVA Insight.

4. Visualize notifications by using the AVEVA™ Vision AI Assistant OMI app.

5. Visualize skill Information with Scripting.

6. Publish values as tags to AVEVA Insight.

7. View Newsfeed in the AVEVA™ Insight OMI app.

When operators use the AVEVA Vision AI Assistant solution, they do not need to continuously monitor live video feeds from cameras, but instead visualize notifications within the AVEVA OMI (5) or AVEVA Insight (3).

You can configure AVEVA Vision AI Assistant to connect to the AVEVA Insight OMI app (3) to provide real-time alerts based on the image classification.

The AVEVA Vision AI Assistant OMI app (4) allows you to visualize the anomalies directly from an AVEVA OMI app.