Integration with OSIsoft PI Event Frames

How does the integration work

PI event frames 

Events are important process or business time periods that represent something happening that impacts your process or operations. Examples of different types of events include: asset downtime, process excursions, equipment startups and shutdowns, environmental monitoring excursions, product tracking batches or paper grades, operator shifts, or any other event type that has both a start time and end time. 

The PI System has a feature called event frames that capture critical event contexts.  The event frames feature permits customers to capture a name, start time, end time, and a series of related information (event attributes) that are useful for analysis. Typically, event frame templates are made to standardize the related data associated with different types of events.  

TrendMiner context items  

TrendMiner context items allow the user to add and classify process events on top of its time series, to create situational awareness inside of your process data. When linked to PI event frames, context items contain all the properties that event fames have, including any additional properties provided by the user. 

Linking PI event frames to TrendMiner context items 

The table below shows an overview of how properties of PI event frames are mapped to properties of context items. 

PI Event Frame  TrendMiner context item 
Template  Type 
Name  Context field: EVENT_FRAME_NAME
Description  Description 
Start time Time of start-event (EFStarted) 
End time  Time of end-event (EFEnded)
Primary element (PI AF Element) Component (asset or attribute) 
Attributes  Fields 
Other referenced elements  Additional properties 

For each PI EF template, a corresponding context item type is created in TrendMiner that will show up in the configuration screen of context item types. This means that an admin user can update the name of the type which is visible in TrendMiner, but will have no effect on the information stored in your PI system.  

The “primary element” in a PI event frame refers to an element in the related PI asset framework; the same is valid for TrendMiner: the component in a context item refers to an asset or attribute in the TrendMiner asset framework. 

Configuration

Starting from version 2021.R2.1, OSIsoft Event Frames are considered as an external data source and need to be configured in the data source section in confighub. You can add a datasource with asset capabilities (using the PI asset framework provider), or enable the asset capabilities on your current asset data source.   

IMPORTANT: As PI event frames are tightly coupled to PI Asset Framework (i.e. the reference from the “primary element” property of an event frame to an element in PI AF), the integration of TrendMiner with PI event frames is also tightly coupled to the integration of TrendMiner with PI AF, we require both TrendMiner - PI integrations to be active AND share the same configuration. If not, TrendMiner would not be able to correctly represent the component of a context item, cfr. the primary element of an event frame. As a result, you will only be able to access event frames from a PI EF data source if the coupled PI AF data source is configured in TrendMiner for asset framework integration. 

Supported versions 

To synchronise your OSIsoft PI event framework, TrendMiner needs to have access to the PI AF SDK. The SDK is included in the PI AF Client and the PI OLEDB Enterprise driver. 
 
We recommend using release 2017 R2, 2.9.0 or later of the OSIsoft PI AF Client or PI OLEDB Enterprise packages for compatibility. Syncing older versions is possible but won't achieve optimum performance and might lose events in specific cases as the PI API in the older versions don't support the necessary query parameters TrendMiner needs. Contact TrendMiner support (support@trendminer.com) for more information. 

How does the sync work? 

  • Live synchronisation 

TrendMiner offers live synchronisation (with a 2 minute interval) of ALL event frames that are modified or created. If a lot of items are edited at the same time (e.g. a bulk action in PI EF), this interval is synced in a separate thread (excessive interval thread) to avoid a delay on the live sync. The live sync can be enabled in the context properties section of the data source.  

  • Historical synchronisation 

TrendMiner offers a historical synchronisation of ALL event frames that are edited or creating within a predefined timeframe. The start and end date can be selected in the data source detail panel.   

 

blobid0.png