Thin Film Transistor-liquid Crystal Display Manufacturing Research Articles

Using image analysis techniques to enhance TFT-LCD manufacturing yield through timely detection of PI alignment layer defects

This study aims to develop a solution to the manufacturing process defects of thin-film transistor liquid crystal display (TFT-LCD) panels. Specifically, it focuses on online image analysis for detecting uneven coatings of the polyimide (PI) alignment layers, which are crucial in the TFT-LCD manufacturing process. The PI solution is typically applied to a substrate surface using an array of nozzles via droplet deposition. During this process, nozzle clogging can cause a series of defects in the form of spots on the alignment layer. These defects impact the alignment of liquid crystal molecules in the TFT-LCD, leading to decreased display quality, reduced visibility, uneven brightness, and compromised image details. If suspected signs of alignment layer spots are identified, the image analysis technology assists online inspectors with prompt responses to minimize TFT-LCD production errors. The results of this study demonstrate timely detection of alignment layer spots, prevention of their progression, and significant reduction of the time required for manual inspection to improve manufacturing yield. This newly developed approach is applicable to not only TFT-LCD production monitoring but also smart display panel manufacturing, thereby providing significant value in high-volume and large-area panel fabrication processes.

Reclaimed water in Taiwan: current status and future prospects

According to the Taiwan Water Resources Agency, Ministry of Economic Affairs, the average water demand shortage is 530.6 million m3 yr−1 during the period of 2011 to 2019, and the situation will worsen in the near future due to global climate change. Therefore, reclaimed water has been an important new water source in Taiwan, particularly for industrial consumers such as high-tech industries in Science Parks. In order to meet the targeted reclaimed water supply of 1.32 million m3 d−1 (CMD) in 2031, Taiwan is focusing on two major reclaimed water sources, including reclaimed water from high water-consuming industries and municipal wastewater treatment plants. This report reviews current technologies used for reclaimed water including units for pretreatment, desalting, polishing, and reclamation. Case studies in Taiwan including reclaimed water from high water-consuming industries such as thin film transistor-liquid crystal display (TFT-LCD) and semiconductor industries, as well as from municipal wastewater treatment plants are presented. The TFT-LCD company Innolux and semiconductor company Advaned Semiconductor Engineering have implemented total recycled water system to recycle and reclaim wastewater from manufacturing processes, achieving a total recycled water of 290 million m3 yr−1 with about 97% recovery and 3.5 million m3 yr−1 with 80% recovery, respectively. The Fengshan reclaimed water treatment plant produces 40,436 CMD reclaimed water from municipal wastewater for the China Steel Cooperation’s steel-making processes, at an overall operation and maintenance cost of 11.5 NT dollars m−3. Meanwhile the Yongkang plant produces 15,500 CMD of reclaimed water for semiconductor and TFT-LCD manufacturing processes at an overall operation and maintenance costs of 25.8 NT dollars m−3, which is due to low urea and boron limits requested by the user. Finally, challenges and future prospects for promoting the use of reclaimed water to meet the targeted supply in 2031 will be discussed.

A Hybrid Method of Frequency and Spatial Domain Techniques for TFT-LCD Circuits Defect Detection

Defect detection is a crucial but challenging task in thin film transistor liquid crystal display (TFT-LCD) manufacturing. Existing vision-based methods focus on either spatial domain or frequency domain with unsatisfactory detection. In view of that, this paper proposes a hybrid template matching method by drawing benefits from both frequency and spatial domain techniques. Under proposal, frequency domain template matching method and saliency detector method are firstly adopted separately to obtain two candidate frequency components associated with defects. In template matching process, a novel selection criterion is taken to improve identifying components associated with local spatial anomaly. Subsequently, the inverse Fourier transform is applied on the intersection of the two candidates to reconstruct the defect regions. In final steps, image entropy in the spatial domain is employed to filter out false detection regions to improve accuracy. To meet industry’s real-time inspection requirement, image partition and multi-threading calculation techniques are introduced into the proposed methodology. Experimental results have shown practical viability of the proposed approach to increase the yield rate of panels through robust defect detection in TFT-LCD manufacturing.

Image-Based Defect Classification for TFT-LCD Array via Convolutional Neural Network

Defect pattern detection and classification are challenging for thin-film-transistor liquid-crystal display (TFT-LCD) manufacturing. Limitations of the existing solutions for automatic optical inspection can be traced in part to the lack of a framework within which different existing and new defect patterns can be analyzed, while integrating domain knowledge and effective technologies. This study aims to develop a framework for image-based defect classification that employs the convolution neural networks without using complex and time-consuming image-processing processes in advance. An empirical study was conducted in a leading TFT-LCD manufacturing in Taiwan for validation. The results have shown that the defect patterns can be effectively classified by the proposed convolutional neural networks that outperform the existing approaches such as Support Vector Machine and Random Forest. The developed solution is implemented to effectively support the engineers.

CNNs Combined With a Conditional GAN for Mura Defect Classification in TFT-LCDs

Mura defect classification is a critical concern for thin-film transistor liquid crystal display (TFT-LCD) manufacturers. In recent years, artificial intelligence technologies have been successfully applied in numerous areas. However, such approaches require large amounts of training image data. Simultaneously, product differentiation and customization strategies have forced the TFT-LCD manufacturing industry to shift from mass production to high-mix, low-volume, and short-life-cycle production. In this environment, collecting a large amount of training data is difficult. Moreover, images with Mura defects captured at inspection stations remain challenging because they are often contaminated with moiré patterns. Moiré patterns severely affect the visual quality of images and cause difficulty in determining Mura defects. This study proposes an approach to eliminate moiré patterns from defect images using a conditional generative adversarial network. In addition, we develop a transfer learning ensemble model that aggregates multiple convolutional neural networks based on a denoising network for defect classification in a limited training data set. The results from an industrial case study demonstrate that the proposed method provides improved accuracy for Mura defect classification. This method can therefore become a viable alternative to manual classification in the TFT-LCD manufacturing industry.

Applying Data Envelopment Analysis to Evaluate the Operation Performance of Taiwan’s TFT-LCD Industry After Post-Global Financial Crisis: A Longitudinal Study

After the global financial crisis occurred in 2008, thin-film transistor liquid-crystal display (TFT-LCD) industry, one of the major industries in Taiwan, has been subject to mass production and low-price competition by manufacturers in China. As a result, companies are facing an increasingly tough operating environment and some of them had to conduct merger/acquisition and transformation and even went bankruptcy ultimately due to unbearable long-term losses. This study aims to apply data envelopment analysis (DEA) and Malmquist Productivity Index (MPI) to evaluate changes in operation performance and productivity of listed companies operating in Taiwan TFT-LCD industry during the period between 2008 and 2017, and research results thereof could provide companies with a reference for organizing adjustment and developing strategies. According to research results: (1) among the six evaluated TFT-LCD manufacturers, TFT-LCD manufacturer F enjoys the best operation performance, and its overall technical efficiency, pure technical efficiency, and scale efficiency are above the average in the industry; therefore, TFT-LCD manufacturer F is considered as the benchmark enterprise among companies; (2) at the same technical level, both inputs of research and development (R&D) and human resources failed to be effectively utilized, which demonstrates that there is still a lot of room for improvement in operation efficiency of inefficient manufacturers.

A two-phase decoding genetic algorithm for TFT-LCD array photolithography stage scheduling problem with constrained waiting time

Abstract The thin-film-transistor liquid crystal display (TFT-LCD) array process is usually the longest process in TFT-LCD production. During the TFT-LCD array process, the scheduling problem of the photolithography stage entails complicated constraints such as photo mask availability, available machines for jobs, and limited waiting time. This study proposes a two-phase decoding genetic algorithm (TDGA) to maximize utilization in the photolithography stage, which is usually the bottleneck. An empirical study was conducted at a leading TFT-LCD manufacturing company. To compare the performance of the TDGA with other metaheuristics, eight scenarios were simulated based on the empirical data. The experimental results show the practical viability of the proposed TDGA.

Fluorescent Determination of Copper(II) in Industrial Wastewater Using Thiol-Capped Cadmium Sulfide–Zinc Sulfide Quantum Dots as the Probe

ABSTRACTIn situ detection of Cu2+ ions in the effluent from thin-film transistor-liquid crystal display (TFT-LCD) manufacturing facility has been successfully achieved using thiol-capped CdS/ZnS quantum dots (QDs) as the fluorescence probes. Three types of ligands, including L-cysteine (LC), mercaptosuccinic acid (MSA), and thioglycolic acid (TGA), were used as capping agents for stabilizing and functionalizing the QDs. Fluorescence quenching of the QDs by Cu2+ ions was well fitted by the Stern–Volmer equation, indicating a static quenching process. Interaction between the Cu2+ ions and the thiol-based capping agents to form Cu–S bonding was responsible for the nonradiative decay. The thiol-capped probes exhibited low detection limits (0.016–0.063 mg L−1), high recovery (81.7–114.5%), good precision (relative standard deviation = 0.36–4.56%), and suitable accuracy for the detection of Cu2+ ions in the field samples, even though those that contained Ca2+, Mg2 +, Na+, K+, and ions, which were 300–16,600 times higher in concentration than the target. These results clearly verify the feasibility of using the QDs as an early warning system for assisting TFT-LCD manufacturing facilities in compliance with strict effluent standards (0.15 mg L−1). Among the three types of thiol-capped QDs, the TGA-capped probes were the most sensitive and the LC-capped probes exhibited the highest errors. Substantial loading of the small TGA ligands on the surface and competitive complexation of the amino groups with Cu2+ ions are responsible for the benefits and the consequences of the TGA- and LC-capped QDs for Cu2+ sensing, respectively.

Capacity planning with ant colony optimization for TFT-LCD array manufacturing

Array manufacturing in thin film transistor-liquid crystal display (TFT-LCD) production network is characterized as a capital-intensive and capacity-constrained production system with re-entrance a...

Risk assessments of work-related musculoskeletal disorders among the TFT-LCD manufacturing operators

The thin film transistor liquid crystal display (TFT-LCD) industry is one of the major industries in Taiwan, yet few studies addressed its work-related musculoskeletal disorders (WMSDs). Thus, this study aims to evaluate the risk of WMSDs among the operators of a TFT-LCD manufacturing company in Taiwan by using the subjective questionnaire and ergonomic assessment tools. First, the NMQ survey was conducted among 393 operators, accounting for 94% of all operators in the TFT-LCD plant. The results showed that the prevalence of WMSDs was 31.8%. Among the various body parts, the highest prevalence of WMSDs was found in neck and shoulders (20.4%), while upper limbs were with the lowest prevalence (9.2%). Besides, the prevalence of WMSDs tends to be affected by both work factors and operator factors. Those who were more experienced, working in day shift and involved in the Array process tend to have a higher prevalence of WMSDs. Further, female and domestic operators reported a higher prevalence of WMSDs than male and foreign operators. Moreover, the jobs of 50 employees were analyzed using ergonomic assessment tools including OSHA MSDs, BRIEF, OWAS, 3D SSPP, and NIOSH lifting equation. The findings indicated that inadequate working height, restricted working space, awkward postures, overweight load, and high repetition were the major risk factors of WMSDs. Thus, countermeasures such as redesigning the workstations and material handling carts, proactive training of manual material handling, and providing safety clearance are recommended to enhance the occupational health and safety of the TFT-LCD operators.

An exploratory study of virtual cell design for thin-film transistor–liquid crystal display (TFT-LCD) array manufacturing

A cellular manufacturing system has several advantages over conventional production methods, including reduced cycle time, less work-in-process, and cheaper setup. The manufacture of a thin-film transistor–liquid crystal display (TFT-LCD) array is a complex process that uses large machines and automated material handling systems. Consequently, the relocation of machines for the manufacture of a TFT-LCD array is unrealistic and cost prohibitive. Virtual cells can be reconfigured using production control mechanisms without physically relocating the machines. This improves the cellular manufacturing system. This is an exploratory study for the design of a virtual cell design for TFT-LCD array manufacturing. The virtual cell design is dynamically reconfigured based on the product mix changes, in order to retain the efficiency and effectiveness of the system. A practical case is used for empirical illustration. The results show that the average improvements in terms of cycle time and the level of work-in-process are 23.73 and 23.59 %, respectively. The empirical results show that the proposed methodology is successful in a practical application.

Tutorial on Applying the VM Technology for TFT-LCD Manufacturing

In the high-tech industries, on-line quality monitoring on each workpiece under processing is required to ensure process stability and improve yield rate. However, conducting workpiece-by-workpiece actual metrology is very expensive and time-consuming. In this case, a novel idea is to use “virtual metrology” (VM) that conjectures workpiece quality based on process data collected from production equipment with a slight supplement of actual metrology data. The purpose of this tutorial paper is to select the thin film transistor-liquid crystal display (TFT-LCD) manufacturing processes as the illustrative examples for demonstrating the methodology of fab-wide implementation of the VM technology systematically. To begin with, a survey of VM-related literature is performed. Then, the features of an effective and refined VM system are presented with the automatic VM (AVM) system developed by the authors as a case study, followed by introduction of the TFT-LCD production tools and manufacturing processes. After that, the generic deployment schemes of the VM technology for the TFT-LCD tools are proposed. Finally, illustrative examples with the AVM system as a case study are presented to show how the VM technology applies to TFT-LCD manufacturing.

A hybrid MCDM approach for improving the performance of green suppliers in the TFT-LCD industry

The awareness of environmental protection/conservation and sustainable development has increased in the last two decades. In the competitive thin film transistor liquid crystal display (TFT-LCD) industry, green supplier evaluation involves environmental criteria in the traditional evaluation process and emphasises that TFT-LCD manufacturers should improve the performance of their suppliers with limited resources. The purpose of this study is to develop a hybrid multiple criteria decision-making with preference ranking organisation method for enrichment evaluations to evaluate and assist polariser suppliers for improving the performance of green suppliers in the TFT-LCD industry to achieve the aspiration level. Based on supplier audit forms, the proposed methods are as follows: to develop a decision framework by an influential network relation map (INRM); to obtain the influential weights of the criteria; to involve the concept of aspiration levels in the preference function; and to evaluate and reduce the gaps of suppliers in each criterion based on the INRM by systematically making decisions in the process of striving toward the aspirations. The results show the ranking as well as the suppliers’ strengths/weaknesses and provide an improvement report to enhance the suppliers’ competitiveness. In the evaluation procedure, the managers place importance on ‘green management’, ‘technology’ and ‘organisation’ aspects. To achieve the goal of being a cost-effective and high-performance manufacturer, the subsidiary polariser supplier should improve performance on ‘health and safety’ first and next develop technology to become the primary supplier by using the proposed methods.

A stochastic dynamic programming approach for multi-site capacity planning in TFT-LCD manufacturing under demand uncertainty

The study focuses on the dynamic multi-site capacity planning problem in the thin film transistor liquid crystal display (TFT-LCD) industry under stochastic demand. Capacity planning refers to the process of simultaneously implementing a robust capacity allocation plan and capacity expansion policy across multiple sites against stochastic demand. In addition, the demand situation in TFT-LCD manufacturing follows Markov properties, in which the correlations of the demand variations in the consecutive periods are high, and the demand status in the next period is stochastically determined by the present one. Therefore, this study constructs a stochastic dynamic programming (SDP) model with an embedded linear programming (LP) to generate a capacity planning policy as the demand in each period is revealed and updated. Using the backward induction algorithm, the SDP model considers several capacity expansion and budget constraints to determine a robust and dynamic capacity expansion policy in response to newly available demand information. The LP model then considers numerous TFT-LCD practical characteristics and constraints to decide a capacity allocation plan, and generate a one-period immediate reward used by the optimality recursion equation of the SDP model. Numerical results are also illustrated to prove the feasibility and robustness of the proposed SDP model compared to the traditional deterministic capacity planning model currently applied by the industry.

SF6 usage and emission trends in the TFT-LCD industry

Several fluorinated compounds (FCs) used in electronics manufacturing, including SF6, perfluorocarbons (PFCs), and hydrofluorocarbons (HFCs) are extremely potent greenhouse gases. These compounds are classified as global warming pollutants to be controlled by the United Nations Framework Convention on Climate Change (UNFCCC) and the Kyoto Protocol. Thin film transistor liquid crystal display (TFT-LCD) manufacturing is a major source of SF6 emissions, in which these gases are used for equipment cleaning and plasma etching. The respective Environmental Protection Administration of the Republic of China, Taiwan and the Environmental Protection Agency of the United States collaboratively strive in encouraging high technology manufacturers to improve annual reporting of FCs emissions and implement emission reduction practices and technologies.In the TFT-LCD industry, FCs account on average for 30% of total greenhouse gas emissions; 80% of emissions originate from the etching process and 20% from the chemical vapor deposition (CVD) chamber cleaning process, which is in compliance with Taiwan EPA data reports from 2005 to 2010. The leading TFT-LCD manufacturers are Japan, Korea, and Taiwan. The World LCD Industry Cooperation Committee (WLICC) subsequently established an aggressive climate protection action plan in 2001. This study focuses specifically on SF6 usage and emission factors reflecting common TFT-LCD manufacturing practices.

Fouling Removal of UF Membrane with Coated TiO2Nanoparticles under UV Irradiation for Effluent Recovery during TFT-LCD Manufacturing

An ultrafiltration (UF) membrane process was employed to treat the secondary effluent discharged from a manufacturing of thin film transistor-liquid crystal display (TFT-LCD) in this study. A bench-scale system was performed to evaluate the fouling removal of a UF membrane with coated titanium dioxide (TiO2) nanoparticles under UV irradiation. The operating pressure and feed temperature were controlled at 300 KN/m2and 25°C, respectively. It was found that the optimum operating conditions were attained with TiO2concentrations of 10 wt% for both 5 KD and 10 KD MWCO. Continuous UV irradiation of 5 KD MWCO improved the permeate flux rate from 45.0% to 59.5% after 4 hours of operation. SEM-EDS analysis also showed that the photocatalytic effect had reduced the average thickness of cake fouling on the membrane from 6.40 μm to 2.70 μm for 5 KD MWCO and from 6.70 μm to 3.1 μm for 10 KD MWCO. In addition, the membrane contact angle was reduced from 54° to 44°. The photocatalytic properties of TiO2apparently increased the hydrophilicity of the membrane surface, thereby reducing membrane fouling.

Dynamic simulation of the decision to construct next-generation TFT-LCD fabrication plants in Taiwan

Thin film transistor liquid crystal display (TFT-LCD) panels have increased in scale on an annual basis since they were adopted for use in television sets in 2003. The higher-generation TFT-LCD plant provides larger sized LCD panels. The increases in technological levels pose a serious challenge to TFT-LCD companies as the size of next-generation TFT-LCD glass substrates continually increases. The construction of next-generation TFT-LCD manufacturing plants is a complex and dynamic process, subject to the influences by interactions among many factors. This paper utilises system dynamics to examine the dynamic structure of next-generation TFT-LCD plants in an attempt to explain system behaviour and simulate relevant trends. Results show that the development of next-generation TFT-LCD plants is determined by the interaction between market demand, and technology level. In other words, these factors form a number of critical loops and a complex, dynamic development process.

An Effective Powerful Test for Supplier Selection Evaluation with Multiple Characteristics

Abstract Supplier selection in high-tech manufacturing industries has become more essential and has received considerable attention. In this paper, we consider a supplier selection problem with multiple characteristics in which we compare two-sided processes with multiple characteristics and select the better one with the higher capability. We first review an approach referred to as the division method. We then develop a new approach called the subtraction method. We compare the two methods in terms of their selection power. The results show that the proposed subtraction method is indeed more powerful than the division method. Two supplier-selection procedures are developed for practical applications. For practitioners’ convenience in applying our procedures, the various sample sizes required for designated selection powers are tabulated and discussed. For illustration purposes, a case is considered and solved for supplier selection evaluation.

Preliminary cost estimates for thin-film transistor liquid–crystal display inspection and repair equipment: A hybrid hierarchical approach

The thin-film transistor liquid–crystal display (TFT-LCD) industry has developed rapidly in recent years. Because TFT-LCD manufacturing is highly complex and requires different tools for different products, accurately estimating the cost of manufacturing TFT-LCD equipment is essential. Conventional cost estimation models include linear regression (LR), artificial neural networks (ANNs), and support vector regression (SVR). Nevertheless, in accordance with recent evidence that a hierarchical structure outperforms a flat structure, this study proposes a hierarchical classification and regression (HCR) approach for improving the accuracy of cost predictions for TFT-LCD inspection and repair equipment. Specifically, first-level analyses by HCR classify new unknown cases into specific classes. The cases are then inputted into the corresponding prediction models for the final output. In this study, experimental results based on a real world dataset containing data for TFT-LCD equipment development projects performed by a leading Taiwan provider show that three prediction models based on HCR approach are generally comparable or better than three conventional flat models (LR, ANN, and SVR) in terms of prediction accuracy. In particular, the 4-class and 5-class support vector machines in the first-level HCR combined with individual SVR obtain the lowest root mean square error (RMSE) and mean average percentage error (MAPE) rates, respectively.

Automatic Defect Detection for TFT-LCD Array Process Using Quasiconformal Kernel Support Vector Data Description

Defect detection has been considered an efficient way to increase the yield rate of panels in thin film transistor liquid crystal display (TFT-LCD) manufacturing. In this study we focus on the array process since it is the first and key process in TFT-LCD manufacturing. Various defects occur in the array process, and some of them could cause great damage to the LCD panels. Thus, how to design a method that can robustly detect defects from the images captured from the surface of LCD panels has become crucial. Previously, support vector data description (SVDD) has been successfully applied to LCD defect detection. However, its generalization performance is limited. In this paper, we propose a novel one-class machine learning method, called quasiconformal kernel SVDD (QK-SVDD) to address this issue. The QK-SVDD can significantly improve generalization performance of the traditional SVDD by introducing the quasiconformal transformation into a predefined kernel. Experimental results, carried out on real LCD images provided by an LCD manufacturer in Taiwan, indicate that the proposed QK-SVDD not only obtains a high defect detection rate of 96%, but also greatly improves generalization performance of SVDD. The improvement has shown to be over 30%. In addition, results also show that the QK-SVDD defect detector is able to accomplish the task of defect detection on an LCD image within 60 ms.