UAV fault mapping turns repowering inspections into a panel-level diligence tool

What changed

Researchers from Spain developed a three-phase workflow that links UAV flight geometry, thermal fault detection and georeferenced mapping. The method was validated on two operating photovoltaic plants in England and Spain, with module-detection accuracy reported between 95% and 98.55% and overall fault-detection accuracy of 96.6%.

The study’s open-access paper reports 85 detected faults, classified as severe, moderate or low severity. It also reports an external-sensor temperature validation within approximately ±2 °C or ±2% and an estimated direct operating cost of about €202 per inspection.

Why it matters for project execution

The practical shift is from “the plant has defects” to “these specific modules, in these locations, show this level of condition risk.” That can improve maintenance prioritisation and make technical evidence easier to compare when a portfolio is being repowered, transferred or prepared for decommissioning.

This is a research workflow, not proof that every commercial inspection will deliver the same accuracy or cost. Site layout, weather, flight permissions, camera quality, data processing and independent validation still determine whether the evidence is decision-ready.

SolTraCo perspective

For large-volume panel movements, condition data matters because it affects removal planning, grading, documentation and the boundary between reusable stock and recycling. A panel-level inspection record can become a useful input to lot definition and buyer conversations, but it should remain one evidence layer alongside serial-number checks, electrical testing, warranty position and logistics constraints.