Conduct a resource audit

Before implementing AVEVA Vision AI Assistant within your organization, it is important to conduct an audit on the following:

1. Available hardware:

   • Camera

   • Lighting

   • Computer with recommended hardware and software

   • Sensors

2. Identify areas of the plant floor that lend itself to detection of visual anomalies.

3. Record the URLs of IP cameras from where the live data is collected.

4. Determine a naming convention for the cameras and image/video datasets. For example, <PlantName><CameraName><Date>

5. Prepare the image or video datasets used to train the skill. Organize and label the datasets into the appropriate states making them easier to upload into the application. For example, Good and Bad, or Open and Closed.

6. If the images and videos are recorded and stored separately, organize the images and videos to make them available to the application.

7. Determine the type of skill best suitable for your use case (Discrete State Detection or Anomaly Detection).