Configuration overview

Configuring the a data model for the first time should begin with a planning phase in which you decide on the information to analyze.

The Niagara Framework® maintains a hierarchy of components, devices, and points that reflect the physical network to which each object belongs. While information from each device and point is useful for tracking real-time values and raising alarms, Niagara 4 provides separate hierarchies, tags and relations with which to set up more meaningful relationships that may have nothing to do with the physical arrangement of devices on a network. The Niagara Analytics Framework (referred to as the framework in this documentation) builds on these standard Niagara 4 features to collect and analyze real-time and historical data in a variety of ways.

For example, your campus may include many buildings. Each building has its own AHU unit for which the data model monitors the values generated by three points: Cool Setpoint, Heat Setpoint and Supply Temp.

The following list summarizes the tasks involved in configuring the data model in this environment:

If you are using a Supervisor computer, you need to add the framework components to the station. A Supervisor computer provides the necessary resources to process large quantities of data.
You add the a tag to each device point that will be part of the data model.
Tagging data is the key to setting up the model. The easiest way to tag data is by using a tag dictionary with a tagging rule.
Each tag may be accompanied by a data definition, which identifies the type of data (cooling capacity, temperature, voltage, etc.) the tag represents. You can view the associated data definitions at AnalyticService > Definitions.
After tagging each point, you mayset up an optional hierarchy by geographical location or, perhaps, by the person responsible for maintaining the AHU unit(s).
Next, you use a pre-defined algorithm (formula) or create your own algorithm to perform calculations. These calculations define how to combine historical trend data, and maximum and minimum acceptable values.
Each algorithm includes a Data Source Block. The Data property for this component contains the same tag as that used to tag the device points. Data collection happens by virtue of the assigned tags and hierarchy without requiring complicated programming.
To visualize the collected data, you bind an algorithm to a Px or Web chart that can take the form of a dashboard or report. To run an algorithm you open the chart that references the algorithm or set up a poller to run the algorithm at regular intervals.
Alerts use algorithms that yield a binary result (true or false). A true result can generate an alarm, which appears on the standard alarm console.
A control point with an Analytic Proxy Extension stores the result of processing an algorithm and can serve as an input to standard components for the purpose of controlling device performance based on logic.

NOTE: You can start by working with a few tags, hierarchies, and algorithms, learning how to visualize and manage the results a little at a time.

Configuration overview

Related Links