Analytic Deployment

When you create a new Visual Analytic, you don't only create a simple graph.

There are also a variety of methods you can use to deploy your analytic - to use it in other ways, across a variety of tools.

At the most basic, every analytic can be used as an API function - a web call that can get the analytic data for your choice of date range. The data is available in a variety of formats, including…

• JavaScript Object Notation (JSON) - used for web and IoT services
• CSV & Tab Delimited Files - used for spreadsheets and legacy applications
• XML - used in C#, C++ and older tools without native JSON support

Certain modules can convert your analytic to a daily report. This has several advantages…

• Key analytic information can be sent to people on a regular schedule (ie. daily),
• Reports can be analysed by our machine learning system to check if they are 'normal',
• Your cache data is regularly generated, meaning weekly or monthly reporting will be extremely quick.

Some modules can create infographics - interactive web-based dashboards to show live or recent information.

These deliver data to people in the field and operators, who need up-to-the-moment information.

Some analytics - particularly Control and Limit analytics - can be converted into an alert, to give immediate indication of issues.

Even better, you can build an analytic on a single asset and deploy it across many.

Some analytics - such as the Shear Tracker or Event Detector - can be used as a source of ARDI events.

This then allows you to easily make new analytics based on those events, or navigate through time using them.

In some cases, the analytic you create can be compiled into native code* for several different languages, including Python.

This means you can build an application that uses dynamic logic from the ARDI server. This is extremely helpful if you'd like to make an application that works across multiple sites but may have radically different logic and infrastructure.

* Some nodes - such as ARDI Point nodes or Timers - require specific functions to be present, or should be run in specially-created modules.