Abstract or Demonstration Description: Verifying the operation and functional stability of the semiconductor tester is required to ensure accurate results: a tester that is out of calibration would contaminate the product yield with unreliable performance. However, the tester calibration procedure normally implies a downtime scheduling, the stop of the tester operation for several hours, and the intervention of a qualified operator to undock the production load board, connect the tester to specific diagnostics and calibration hardware, and dock the tester again to the prober or handler at the end of the procedure. In this scenario, predictive maintenance and self-calibration capabilities can become a strategic initiative to minimize downtime and cost of ownership while improving quality and control. The incorporation of artificial intelligence and smart sensors inside the test equipment enables the performance of automatic diagnostic procedures, self-corrections and comprehensive data analysis, without needing any scheduling, operator’s intervention or specific hardware. The tester becomes capable of autonomously verifying the instruments’ health status, detecting any trend towards out-of-specs performance, then independently adjusting the involved parameters. Key indicators, like the number of relay commutations, give information about the instrument wear and predictive maintenance required, while operating temperature and humidity are constantly monitored to detect any dangerous anomaly. The adoption of this solution on a fab-wide scale can help semiconductor manufacturers save on scheduled and unscheduled downtime, keeping the process quality under absolute control.
.jpg "Higor Rodrigues Batagin photo")