Step 1: Select Responses

Responses are the sensors to be monitored.
In multi-variate models all responses together form the object to be monitored. In data analytics this is called the Y vector, consisting of multiple sensors being monitored all together like one single object. In univariate models each single sensor or response represent an individual Y, or single objects to be monitored separately.

Step 2: Select Features

Features are the sensors used as input for the monitored object.
In data analytics this is called the X vector, consisting of multiple sensors. The selected features must not contain sensors already use as response within this monitored object.

Step 3: Select Training Period

By default, please select the general training periods.
For some plants we figured out the major plant events and subtract them from the total time range.
The other options consist of special period settings for the respective asset.

Select Periods	View Periods

Finally: Summary

Finally, check your design of the monitored object and save it.

If you want to remove one of the entries, please go back to the respective section, e.g. "Responses".

Name for the monitored object:

 

Responses

Features

Periods

Data Config

Confirm your actions: