The heart of the Niagara Analytics Framework is an advanced high performance calculation engine. With it real time data can be combined with historical data using the Niagara wiresheet, a visual programming environment. Analytic algorithms are reusable, they help Niagara scale to very large systems. Unlike a traditional wire sheet logic a single analytic algorithm can run on many components in a system. Niagara 4 hierarchies are leveraged so the same calculation can produce results in a variety of contexts. Analytic algorithms can even combine the results of other analytics algorithms with raw data so that intelligence can be layered onto a system over time as new metrics are invented. Since algorithms can produce historical data this will even allow new metrics to be back tested. With the Niagara Analytics Framework you can add intelligence to your system over time to extend the life of and preserve your investment in Niagara.
The calculation engine can be leveraged for fault detection and diagnostics. Analytic rules can be implemented as analytics algorithms and can be configured to monitor components in a system. Analytic alerts can be routed to the Niagara alarm infrastructure or can be visualized using Niagara and third part tools.
The Niagara Analytics Framework produces the same kinds of data that traditional points, histories and queries produce. This means analytic calculations can integrated into existing graphics. It also means that third party tools which already integrate with Niagara automatically have access to Niagara Analytic Framework.
Analytic calculations may look like traditional Niagara logic but they're very different under the hood. The Niagara Analytics Framework uses stand proxy point to output the results of calculations which can then be injected into traditional Niagara logic. This will enable integration of historical performance into sequences and make continuous adjustments due to normal operation and maintenance.
More info refer the user documentation