APR 10, 2026

Proactive yield maximization in photolithography via human-in-the-loop AI on an on-premise big data platform

SPIE Advanced Lithography + Patterning

Authors

Jeng-Hun Suh, Chang Hoon An, Taekwon Jee

Abstract

Yield loss in advanced semiconductor photolithography presents an increasingly significant challenge as device features scale into the nanometer regime. Conventional reactive process control methodologies, such as Statistical Process Control (SPC), have proven insufficient for managing the growing complexity and data volume in modern fabrication environments, leading to elevated false alarm rates and delayed detection of process excursions. This paper proposes a framework that transitions from reactive fault correction toward proactive yield optimization through the integration of Artificial Intelligence (AI) with a Human-in-the-Loop (HITL) protocol. Built upon a Virtual Fab infrastructure for domain knowledge-driven synthetic data generation, the framework comprises three core components: (1) a Causal Retrieval-Augmented Generation (Causal-RAG) system for real-time root cause analysis, (2) a domain-specific embedding engine for semiconductor process data retrieval, and (3) a multi-modal multi-agent AI system serving as an intelligent decision support platform. Preliminary validation demonstrates a Top-5 retrieval accuracy of 91% for identifying relevant causal factors from historical defect databases. The architecture is designed to operate on a secure, high-performance platform ensuring data sovereignty and low-latency performance suitable for production environments.