AI System for Routing Solar Panel Images for Technical Review

The problem

Images requiring review may be identified late when panel numbers are high. When panel image, capture time and conditions, array identity, production and maintenance records are scattered across sensor, document, site or team records, the issue may not be identified in time.

This fragmented data delays preparation of the panel review queue and forces teams to search for supporting evidence again by hand. Before making a decision, the electrical engineer or authorised maintenance specialist must complete missing records and verify each inconsistency individually.

A common exception is that reflection, angle, weather or temporary shade may resemble a defect. Incomplete or incorrect automation that does not check this context could cause an unnecessary site intervention or allow a genuine fault to be missed, so no physical or operational action can be taken directly.

What we set out to improve

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

How the system works

1. 01 Define the review scope and the roles of the authorised electrical engineer or authorised maintenance specialist
2. 02 Retrieve the inputs: panel image, capture time and conditions, array identity, production and maintenance records
3. 03 Check the source identifier, date, calibration or version, and data recency
4. 04 Match the image with the array and date
5. 05 Compare the visual signal with production and capture conditions
6. 06 Check the exception: reflection, angle, weather or temporary shade may resemble a defect
7. 07 Place low-confidence or inconsistent records in a separate specialist queue
8. 08 Write the source identifier, rule, output and time to the audit trail
9. 09 Have the electrical engineer or authorised maintenance specialist approve, correct, defer or reject the output

Methods we used

• Image pre-classification with performance context
• Source, format, time and required-field checks for panel image, capture time and conditions, array identity, production and maintenance records
• A missing-data warning instead of an estimate when inputs are incomplete or inconsistent
• A separate exception check outside the main rule
• Joint retention of the source, rule result, panel review queue and human decision

Where people stay involved

Data and security

Panel image, capture time and conditions, array identity, production and maintenance records are accessed with the narrowest permissions required for the task. Commercial, location and employee information is masked where possible, and only the necessary portion of the data is sent to the model rather than the complete raw dataset. Access, outputs and the electrical engineer's or authorised maintenance specialist's decision are recorded; production does not begin until the organisation approves the retention period and data location.

Who this suits

Frequently asked questions