Create a guided analytics model in InSight

This article shows how to create a guided analytics model in InSight (cloud).

​locate the asset you wish to add a guided analytics model on. In this case browsed from the Asset hierarchy, but there are other ways to get to this screen:

clipboard_e837029a61fa03f3ca15e69287a76b2c8.png

 

Next select the "add guided analytics model" selection from the asset actions menu as shown:

clipboard_e3f595f7c3858eeb4896524129119a727.png

 

select the tags you wish to add to the model:

clipboard_e9cef9dd24328ae60ffc363ff3bef42d7.png

Now give the guided analytics model a name and a training window:

clipboard_efe78f0bb42dd661b8bd2d708ea0bdc3a.png

 

Next we will add a filter to tell the model to ignore monitoring the variables when the asset is not running.  In my case I assumed that when the current was below 35 amps I want to assume my asset is not under load and I do no want the data to be considered in the model.

clipboard_e640b7a1599e47101dc4b548e4b0f7af6.png

You can also add an operational mode indicator at the next step, but I did not add one in this example so now we see the summary screen:

clipboard_e48349637517bd29cbdf949118a59a2a8.png

 

The end result that gets created shows an anomaly score with all of the contributing variables and the score ranges from 0-100 based on how much that variable contributed to the anomaly score.  The higher score indicates likely where the problem came from:

clipboard_e9109362757040c360d46f8578212c749.png