AI System for Reviewing Grid and Transformer Maintenance Signals

The problem

In day-to-day operations, maintenance signals held in different systems may not be viewed together. When transformer identity, load, temperature, alarms, oil analysis, relay events, work orders and calibration records are scattered across sensors, documents, field records or team files, issues may not become visible in time.

This fragmented data delays preparation of the transformer maintenance review file and forces teams to search manually for supporting evidence again. The authorised electrical or maintenance engineer must complete missing records and verify each conflict before making a decision.

A common exception is that load testing, commissioning or sensor calibration may create a temporary signal. Incomplete or incorrect automation that fails to check this context could take a sound transformer out of service unnecessarily or leave risky equipment in operation, so no physical or operational action can be carried out directly.

What we set out to improve

• Proportion of maintenance reviews verified by the electrical engineer
• Proportion of outputs corrected or rejected by the authorised electrical or maintenance engineer
• Time between the source event and the review output

How the system works

1. 01 Define the review scope and the role of the authorised electrical or maintenance engineer
2. 02 Retrieve the inputs: transformer identity, load, temperature, alarms, oil analysis, relay events, work orders and calibration records
3. 03 Check source identity, date, calibration or version, and data recency
4. 04 Combine signals by transformer and time
5. 05 Check the work-order, load and calibration context
6. 06 Check for exceptions: load testing, commissioning or sensor calibration may create a temporary signal
7. 07 Place low-confidence or conflicting records in a separate specialist queue
8. 08 Write the source identity, rule, output and timestamp to the audit trail
9. 09 Have the authorised electrical or maintenance engineer approve, correct, defer or reject the output

Methods we used

• Threshold and time-series signal combination without producing a fault determination
• Source, format, timing and required-field checks for transformer identity, load, temperature, alarms, oil analysis, relay events, work orders and calibration records
• A missing-data warning instead of an estimate when inputs are incomplete or conflicting
• A separate exception check outside the main rule
• Combined retention of the source, rule result, transformer maintenance review file and human decision

Where people stay involved

Data and security

Access to transformer identity, load, temperature, alarms, oil analysis, relay events, work orders and calibration records is restricted to the minimum permissions required for the task. Commercial, location and employee information is masked where possible, and only the necessary portion of the dataset is sent to the model. Access, outputs and the authorised electrical or maintenance engineer's decision are logged; production use does not begin until the organisation has approved the retention period and data location.

Who this suits

Frequently asked questions