Guide
The ultimate guide to managing your PI System Download now
Why You Need Both Monitoring and Data Lineage for Industrial Data Quality
Why You Need Both Monitoring and Data Lineage for Industrial Data Quality
Monitoring and lineage answer different questions.
Monitoring tells the team that a condition is wrong or abnormal. Lineage tells the team where the data came from and what depends on it.
Industrial teams need both.
Monitoring detects the condition
Useful monitoring can identify:
Stale data
Bad states
Data gaps
Flatlines
Failed calculations
Abnormal event volume
Important configuration changes
This reduces the time between failure and detection.
Lineage provides context
When monitoring finds a problem, lineage helps answer:
Which source produced the value?
Which AF attribute references it?
Which analyses use it?
Which displays or reports are affected?
This reduces diagnosis time and supports impact analysis.
Monitoring without lineage creates a backlog
A large list of bad tags does not tell the team which findings matter most.
Without usage and dependency information, engineers must investigate each finding manually.
Lineage without monitoring is passive
A lineage graph can show relationships but does not tell the team when a tag becomes stale or an analysis fails.
Without health checks, the team still depends on users to report problems.
Combine them with change history
The strongest investigation workflow connects health, lineage, and changes.
For example:
Monitoring detects a stale calculated output.
Lineage shows the upstream input and downstream displays.
Change history shows that the input mapping changed before the failure.
The team repairs the mapping and verifies the affected displays.
This is much faster than reviewing each system separately.
Prioritize by operational use
Use lineage to identify critical consumers and use monitoring to track their health.
The goal is not to monitor every relationship with the same urgency.
The objective
Monitoring finds the problem. Lineage explains the path and the possible effect.
Together, they give PI and OT teams the visibility required to detect issues earlier, troubleshoot faster, and make changes with fewer unknown dependencies.