Tell me about a difficult PCBA problem you solved. How did you identify the real root cause?

In one case, we first reproduced the symptom and separated facts from assumptions. I reviewed machine alarms, PM, spare parts, downtime recovery, calibration, compared good and failed samples, and collected evidence before changing any process condition. We confirmed the cause with a controlled test, implemented corrective action, and monitored the next production lots to verify effectiveness.