US researchers say a self-supervised machine-learning tool can identify long-term physical defects in solar assets weeks or years before conventional inspections, potentially reducing operations and maintenance costs…
https://www.pv-magazine.com/2025/12/19/us-researchers-use-machine-learning-to-detect-hidden-solar-array-defects/
