The Importance of Tracking Tag Usage in the PI System

A PI Point is not important only because it exists. Its importance depends on where it is used.

Usage information helps PI administrators understand dependencies before they clean up tags, migrate servers, change AF models, or investigate data-quality problems.

Why tag inventory is not enough

A tag list can show name, source, engineering units, and update status. It does not show whether the tag supports an important workflow.

The same PI Point can be used by:

• AF attributes

• AF analyses

• PI Vision displays

• Reports

• Custom applications

• Data exports or cloud pipelines

A change to the tag can affect all of these consumers.

Use usage data for safe cleanup

Unused and duplicate tags can make a PI environment harder to manage. However, deletion based only on age or update frequency is risky.

Before retirement, determine whether the point is referenced by active systems. For important changes, record the known dependencies and verify downstream behavior after the change.

Use usage data during migrations

Server consolidation and control-system migrations can create many tag changes at one time.

Usage information helps the project team identify:

• Which tags must move first

• Which displays require validation

• Which AF mappings must change

• Which calculated outputs have downstream dependencies

• Which legacy points can remain read-only for historical access

This reduces manual display-by-display review.

Combine usage with data quality

A quality finding becomes more useful when the team knows who and what depends on the signal.

For example:

• A stale unused tag can be a low-priority cleanup item.

• A stale tag used by a production KPI can require immediate attention.

• A bad source used by several analyses can have a wider blast radius than one isolated point.

This context helps teams prioritize work.

Track both direct and indirect dependencies

Direct usage is easy to understand. A PI Vision symbol can reference a tag directly.

Indirect usage is more important in complex environments. A display can use an AF attribute that receives the output of an analysis that depends on several other attributes and PI Points.

A useful dependency model should follow the chain, not stop at the first reference.

Keep usage information current

A static export becomes obsolete as engineers create displays, change AF models, and retire applications.

Automate discovery where possible and refresh the inventory on a regular schedule.

Also record systems that cannot be scanned automatically. No tool can infer every external consumer unless that system exposes the relationship.

The practical outcome

Tag-usage tracking gives PI teams a better answer to a basic change-management question: what can this change affect?

That answer supports safer cleanup, faster troubleshooting, and more controlled migrations across the PI environment.