Weather example: High-level procedure

We use weather as an example of a generalized application to show you how to use future PI points in predictive analysis. Predictive analysis of heavy rain, winds, dust storms, etc., can find application in various industries: it may result in additional crews sent out for various kinds of maintenance, ordering enough salt to counter icy conditions on highways, ensuring extra power in the grid for periods of severe storms, and so on.

Follow these steps to set up a sample weather system and test out the capabilities of future points; alternatively, you may use these as guidelines to create a specialized schema for your own application, depending on the specific needs of your industry.

1. Verify your PI Server installation.

   See Weather example: Verify PI Server installation for more information.

2. Create historical PI points for past weather data and future PI points for predicted weather data. Populate the PI points with weather data collected from a public source.

   Weather example: Create and populate PI points with weather data

3. Set up an AF element template and associated attribute templates that get their values from configured historical and future PI points. Then create PI AF elements for several geographical locations.

   See Weather example: Create PI AF elements and attributes

4. Using AVEVA PI Vision generate trends over a specific time period to see historical versus predicted trends for a particular weather attribute.

   Weather example: Create trends and analyze data