• testing
• semiconductors
• accelerometer
• artificial intelligence
• vibrations
• z-score
• maintenance
• predictive

The thesis project focuses on the implementation of a Predictive Maintenance system on semiconductor testing equipment. In the thesis project the concept of Predictive Maintenance takes two different paths: Mechanical Predictive Maintenance and Electrical Predictive Maintenance. The purpose of Mechanical Predictive Maintenance is to analyze the vibration signals produced by the testing equipment in order to implement an artificial intelligence model which, by analyzing the real-time data produced by the equipment, is able to understand when the vibrations are anomalous and therefore predict the future equipment breakdown. Electrical Predictive Maintenance, on the other hand, is based on the analysis of electrical signals produced by the testing equipment. The z-score algorithm implements a mechanism that can tell when electrical signals are out of bounds.