Relying solely on end-of-line testing isn't enough when security, traceability, and mission reliability are vital.

The software tool developed by Stony Brook University uses self-supervised learning to detect long-term solar equipment damage weeks or years before manual inspections find it.