Reliable tool operation in semiconductor fabrication is essential for developing high-quality devices and maintaining a safe and efficient environment for researchers. However, current fabrication tools often lack systematic management and are prone to malfunctions, leading to unnecessary costs and significant research delays. To address this challenge, I propose developing a monitoring system that collects operational data from cleanroom tools through camera-based recording or integrated screen-capture functions. The collected data combined with user input will be stored in a centralized database and analyzed using computer vision techniques to detect irregularities in tool performance. The ultimate goal is to provide real-time alerts to users, actively checking for potential tool errors.

I am interested in this UROP project particularly because it is interdisciplinary, combining my interests in AI and semiconductor technologies. I hope to explore the intersection between hardware reliability and computer vision, deepening my understanding in computer vision and anomaly detection and gaining hands-on experience in developing practical monitoring systems. My goal for this project is contributing meaningfully to our research group’s efforts to improve tool reliability.