Carbon fiber is currently one of the most popular composite materials in the world, with its wide applications ranging from bikes to space shuttles. However, there is no comprehensive method for the total quality inspection of carbon-fiber products so far due to its feature of continuous production and destructive examination of quality inspection. Through years of development, the Automatic Virtual Metrology (AVM) technology has been applied to various industries to turn the offline sampling inspection with metrology delay into online and real-time total inspection. With AVM, every workpiece of the carbon-fiber products can be examined to ensure product quality. The biggest issue of continuous production is that it can neither accurately define each unit nor label on the products. To solve this problem, a production data traceback (PDT) mechanism is proposed in this paper. The PDT mechanism can provide a virtual label to each unit for referring back to the process parameters of the production. With PDT, the process data of each workpiece of a complete spin can be acquired so as to accomplish the work-in-process (WIP) tracking requirement of applying AVM. Also, the Advanced-Manufacturing-Cloud-of-Things (AMCoT) platform is adopted in this work to fulfill all the smart manufacturing requirements for achieving the goal of zero-defects of carbon-fiber manufacturing.

Product yield directly affects production cost. Thus manufacturers seek to quickly enhance product yield during the development and mass-production processes. In other words, when a yield loss occurs, the root causes should be found rapidly in both the development and mass-production phases. When a yield loss is encountered, the traditional yield enhancement approach is to collect all production-related data to perform big data analysis in order to find out the root causes of affecting yield and remedy them. However, production-related data are extremely large and complicated, which makes it hard to search for the root causes of a yield loss. This leads to the problem of high-dimensional variable selection. To solve this problem, a scheme of Key-variable Search Algorithms (KSAs) is proposed in this paper. The inputs of this KSA scheme include production routes, process data, inline data, defects, and final inspection results; while the outputs are search results and their corresponding reliance indices. The search results are the key stages that cause the yield loss or the key variables that are the root causes of the yield loss. The thin film transistor-liquid crystal display (TFT-LCD) process is adopted as the illustrative example to demonstrate the KSA scheme.

This paper introduces an intelligent optimization framework that integrates Digital Twin (DT) technology, deep learning, and a tailored Multi-Restart Bayesian Optimization with Random Initialization (MRBORI) to enhance parameter control and yield in semiconductor manufacturing. The proposed framework synergizes XGBoost-based feature selection, which identifies critical parameters in high-dimensional spaces, with a custom deep learning surrogate model that captures complex nonlinear interactions. Building on these insights, the MRBORI strategy leverages multiple optimization restarts, each initialized randomly, to mitigate local minima risks and systematically explore broad parameter spaces. Experimental validation using real-world data from an epitaxial silicon carbide (Epi SiC) process demonstrates notably tighter thickness control and improved yield compared to traditional methods. By unifying DT-driven real-time insights with advanced machine learning and multi-restart optimization, this framework offers a robust and precise solution for tackling the complexities of modern semiconductor manufacturing.

As semiconductor manufacturing complexity escalates, the intricacy of corresponding manufacturing systems intensifies. These extensive systems necessitate diverse engineering expertise for effective operation and analysis. For instance, yield engineers analyze yield systems, process engineers interpret FDC parameters, and equipment engineers monitor device equipment health. Traditional manufacturing systems, reliant on manual data analysis and fixed algorithms, suffer from slow decision-making and limited adaptability. They are susceptible to human error, reactive maintenance, and restricted user interaction confined to technical interfaces and business hours. Additionally, scalability and integration pose significant challenges, inflating operational costs and hampering resource efficiency. This paper introduces an Intelligent Manufacturing Virtual Assistant (IMVA) specifically designed for the semiconductor industry. By harnessing the power of Large Language Models (LLMs) and AI Agents, IMVA enhances yield analysis and seamlessly integrates with existing systems and tools. It exhibits high accuracy in defect detection through advanced data analysis and report generation. Furthermore, IMVA facilitates natural language interaction, rendering it user-friendly and accessible to non-technical personnel. Consequently, IMVA markedly improve operational efficiency and cost-effectiveness compared to traditional manufacturing systems. The efficacy of IMVA is demonstrated through the Wide-bandgap (WBG) process, showcasing its capability to simplify root cause analysis and provide comprehensive yield reports.

To ensure stable processing and high-yield production, high-tech factories (e.g., semiconductor, TFT-LCD) demand product quality total inspection. Generally speaking, sampling inspection only measures a few samples and comes with metrology delay, thus it usually cannot achieve the goal of real-time and online total inspection. Automatic Virtual Metrology (AVM) was developed to tackle such problem. It can collect the data from the process tools to conjecture the virtual metrology (VM) values in the prediction model for realizing the goal of online and real-time total inspection. With the advancement of technology, the processes become more and more precise, and better accuracy of VM value prediction is demanded. The CNN-based AVM (denoted as AVM _CNN ) scheme can not only enhance the accuracy of the original AVM prediction, but also perform better on the extreme values. Nevertheless, two advanced capabilities need to be addressed for its practical applications: 1) effective initial-model-creation approach with insufficient metrology data; and 2) intelligent self-learning capability for on-line model refreshing. To possess these two advanced capabilities, the Advanced AVM _CNN System based on convolutional autoencoder (CAE) and transfer learning (TL) is proposed in this work. It is verified that the Advanced AVM _CNN System is more feasible for the onsite applications of the actual production lines.

Despite recent advances in Digital Twin (DT) technologies for semiconductor manufacturing, no existing research convincingly demonstrates a unified, rapidly scalable, and data-efficient DT framework that can effectively handle stringent multi-objective optimization under severe data scarcity. This absence of a comprehensive approach impedes widespread industry adoption, given the pressing need for flexible, universal solutions that rapidly adapt to diverse machines and processes. This study introduces MOODFG-MLTL, an innovative algorithm that integrates Meta-Learning and Transfer Learning within a Multi-Objective Optimization using Deep-Feature Gaussian Processes (MOODFG) architecture. By harnessing a centralized meta-model repository and dynamically refining surrogate models, the proposed framework efficiently leverages limited data and transfers knowledge across heterogeneous manufacturing configurations. In our experimental validation with Epi SiC processes, MOODFG-MLTL achieves near-optimal wafer thickness and doping control using only 15 samples—reducing data requirements by approximately 50% while converging up to 40% faster compared to conventional approaches. Experimental validation in epitaxial silicon carbide (Epi SiC) scenarios demonstrates that the MOODFG-MLTL not only maintains robust performance under constrained data conditions but also accelerates the deployment of data-efficient, multi-objective DT solutions. This advancement provides a critical foundation for achieving universal, intelligent DT implementations in the semiconductor industry’s evolving landscape.