Event frames: way to track and capture events and find related data.
Big data: real challenge is how to filter down to problem that you’re trying to solve, how to find all the related data to make that happen. Graph of electrical grid (tags tracking faster than humans have seen before – 10x/sec, 10 Hz, 24 hours, 46M values. Zoom into two minutes of data, drop in frequency over several substations (in this example).
Over time, it’s difficult to remember where the events are in all the data (like finding a 2 minute period somewhere over 20 years). Other events you might want to track: downtime. What happened in batch of highest yield ever. What was root cause for $200K excursion event.
Needs: simplify data analysis, perform asset comparisons, event comparisons, or discover event interrelationships. PI event frames are a core capability of PI server. Many types of events in different industries (different use cases).
Roadmap: wave 1 (2011): partner and early adopters. Wave 2 – event frames for mainstream (2012): first end-to-end frames experience. Wave 3: move PI batch customers forward (future). Now several companies/partners using event frames.
PI Batch database: PI server migrate and link (part of wave 3), Generation of event frames: updates to batch interfaces, batch generator. (This is an upgrade discussion.) Using any tag: Abacus works with generic events at sometime TBD. Event frame analysis rules for element templates, work in progress.
PI DataLink (reporting and table-based analytics) and PI Processbook (authoring and process monitoring) upgrades. Event integration: data access with timie series data and assets (and events).
Demo. It’s hard to describe these live action demos when I’m trying to understand exactly what’s going on. Getting a list of assets, measurements, queries. Microsoft Power Pivot allows integration of data from disparate sources.
Data Coresight related assets: automatically discover them, access other data that might be useful. Drag onto other assets, trends, etc.