Structure of event frames

Each event frame has a name, start time, end time, one or more attributes, and one or more referenced PI AF elements. As with elements, you should create event frame templates to standardize and manage the attributes for different types of events.

With event frames you can easily search the PI System for the events themselves, rather than search for events by time. You can configure event frames to return all related event data automatically in real-time so that you do not need to query multiple systems for event and process data, and then merge them together manually. You can also set up event frames to retrieve historical data.

When to use event frames

There are two categories of trackable events that would fit an event frame profile:

• "Good" events: Events that you want to track as a normal part of business, such as product tracking, shifts, and so on.

• "Bad" events: Events that are unexpected and need to be analyzed and perhaps fixed quickly if they ever occur, such as expected shutdowns or excursions. These are events that you want to track and report in aggregate, over a period of time.

Asking questions such as these can help identify events or conditions that must be tracked:

• What are all the times that event X occurred on this type of asset?

• Can I associate data from different tags for a time-range, or for a single point in time?

• What is the associated data for a particular time period when a problem occurred or may occur in the future?

• What are the critical process events that someone needs to be notified on?

• Are there digital states for PI tags that are significant when they change, and must they trigger other actions?