Y. T. Tai

Capability assessment for processes with multiple characteristics: A generalization of the popular index Cpk

PU. We not only provided some tables but also presented an application example. Copyright

Measuring the Manufacturing Yield for Processes With Multiple Manufacturing Lines

Process yield is the most common criterion used in the semiconductor manufacturing industry for measuring process performance. In the globally competitive manufacturing environment, photolithography processes involving multiple manufacturing lines are quite common in the Science-Based Industrial Park in Hsinchu, Taiwan, due to economic scale considerations. In this paper, we develop an effective method for measuring the manufacturing yield for photolithography processes with multiple manufacturing lines. Exact distribution of the estimated measure is analytically intractable. We obtain a rather accurate approximation to the distribution. In addition, we tabulate the lower conference bounds based on the obtained approximated distributions for the convenience of industry applications. We also develop a decision-making method for process precision testing to determine whether a process meets the process yield requirement preset in the factory. For illustration purposes, an application example is included.

Capability Assessment for Weibull In-Cell Touch Panel Manufacturing Processes With Variance Change

Since touch panels can provide natural user-interface, including fluent multipoint touch or advance gesture recognition, recently they have been extensively applied in various portable devices, such as smart phones and tablet PCs. In-cell touch panel is the highest integration touch technology as compared to the on-cell and typical touch panel manufacturing technologies for the thinnest and lightest structure. In in-cell manufacturing processes, manufacturing yield assessment is an essential issue. However, inevitable process variance changes could arise from equipment, material, and operation, and may not be detected within a short time. In addition, the process output usually has a Weibull distribution. To circumvent the undetected variance change causing the inaccurate manufacturing yield calculation, we provide a yield measure index to avoid overestimating when the underlying distribution is Weibull with variance change. We also show that the accommodation of the process capability index would not be affected by the scale parameter of Weibull distribution. Applying this method, the magnitudes of the undetected variance change are incorporated into the evaluation of manufacturing yield. For illustration purposes, a real application in an in-cell manufacturing factory, which is located in the Science-based Industrial Park in Hsinchu, Taiwan, is presented.

Accessing Manufacturing Yield for Gamma Wafer Sawing Processes in COG Packaging

The technology of thin film transistor liquid crystal display (LCD) has become more popular due to great demand for worldwide consumer electronic products. Driver integrated circuit (IC) is a critical device that is embedded sophisticated circuits to drive panels. Since narrow border design on display products is current trend, dimensions of driver ICs are shrunken. In the high-density LCD driver ICs, the operation of wafer sawing is essential and needs accurate yield assessment. However, inevitable process variance changes could arise from sawing machine, material, operation, and workmanship, and may not be detected within short time. Conventionally, manufacturing yield is evaluated applying typical yield measure index method under the assumptions that the processes are stable and normal. To assess manufacturing yield for Gamma wafer sawing processes more accurately, we present a modified yield measure index method. Using the proposed method, the magnitudes of the undetected variance change, which are functions of the detection power of the

Measuring the Manufacturing Yield for Skewed Wire Bonding Processes

{{S}^{2}}

Group Supplier Selection for Multiple-Line Gold Bumping Processes

chart, are incorporated into the evaluation of manufacturing yield. In addition, we demonstrate, mathematically, that the accommodation would not be affected by the scale parameter of Gamma distribution. For illustration purpose, a real-world case in a wafer sawing factory which is located on the Science-based Industrial Park in Hsinchu, Taiwan, is presented.

Product mix determination and capacity allocation for heterogeneous products in thin film transistor liquid crystal display manufacturing

In semiconductor manufacturing, the technologies of stacked semiconductor packaging are important miniaturization strategies in which excellent quality is essential for a good reputation. Yield-based index Cpk has been the most popular tool for successful quality improvement activities and quality program implementation in multichip package processes. For in-plant applications, quality practitioners commonly use the nonconformities in parts per million (NCPPM) table of Cpk to obtain manufacturing yields. However, wire bonding processes are often skewed in multichip package factories. In such situations, the NCPPM table of Cpk is inappropriate to be applied directly. In order to reasonably access manufacturing yield for skewed wire bonding processes, we propose a new yield index Csk. We are the first to modify the index Csk for various skewed distributions for the consistency of NCPPM mapping between indices Cpk and Csk. In addition, bootstrap methods are applied to construct lower confidence bounds which are useful to practitioners for making reliable decisions. An approximated unbiased estimator is also provided. For the purpose of illustration, practical applications are presented.

Estimation of a Modified Capability Index for Non-Normal Distributions

Portable devices have been popularly used nowadays. Liquid crystal display driver integrated circuit is an essential component in portable devices in which gold bumping process is a critical interconnection technology. To enhance bargaining power and to reduce cost, group suppliers are commonly selected from multiple suppliers. For high-definition display devices, a very low fraction of defectives is the basic quality requirement in gold bumping processes. Unfortunately, conventional yield measurement methods for supplier selection no longer work since any sample of reasonable size probably contains no defective gold bump product items. In this paper, we propose a group supplier selection procedure for multiple suppliers with multiple-line processes that are quite common due to high demands in semiconductor industry. For various significance levels, the number of suppliers, the number of manufacturing lines, predetermined capability requirements, and sample sizes, different selected suppliers are suggested based on the multiple-line yield index CpkM. In addition, the power comparisons between different methods are also discussed. For the illustration purpose, real-world applications in a gold bumping factory that is located in the Science-Based Industrial Park in Hsinchu, Taiwan, are included.

Production Yield for Multiple Line Processes: Product Acceptance Determination

We investigate heterogeneous products involving EPD and LCD products in TFT-LCD companies.We model three mathematical programming models to obtain the appropriate product mix.Multi-stage, multi-site, and multi-generation TFT-LCD manufacturing environment is considered.We present a real-world case taken from a TFT-LCD company located in Taiwan.We perform sensitivity analysis to investigate the effect on the optimal solution. In the thin-film transistor liquid crystal display (TFT-LCD) manufacturing, the development of models to determine product mix and capacity allocation for a multi-stage, multi-site, and multi-generation company is very important. As rapid advancement of the electronic-paper display (EPD) technology, it is a trend of the coexistence of heterogeneous products involving LCD and EPD products. In the case we investigated, the TFT-LCD company merged an EPD company and incorporated their electro-phoretic technologies into their company. Notably, the manufacturing processes of the EPD products do not incorporate the process steps of liquid crystal injection and attachment of color filters. Consequently, to minimize the total cost, it is essential to decide the appropriate product mix and capacity allocation with considerations of the resource consumption of LCD and EPD products simultaneously. In this paper, we present mathematical models to determine product mix and capacity allocation, which involve three subsystems for TFT-LCD process with consideration of net demand, inventory level, yield rates, cost, margin, outsourcing allocation, cycle time, and panel conversion rate. To demonstrate the applicability of the proposed models, we present a real-world case taken from a TFT-LCD company located in the Science-Based Industrial Park at Hsinchu, Taiwan and perform sensitivity analysis to investigate the effect on the optimal solution.

Convenient Ratio Approach for Industrial Implementation in Estimating and Testing Process Yield

Process capability indices (PCIs), which are very important in quality control have been one of a numerical measure index in manufacturing processes. Index Cpk is the most popular one used in the manufacturing industry. It is applied under the assumption that the processes are normally distributed. In real-world applications, non-normal processes may occur in industries, and the index CNpk has been proposed for non-normal processes in which its exact sampling distribution is mathematically intractable. Quality practitioners commonly use the existing NCPPM (non-conformities in parts per million) table of Cpk to obtain process yields. However, the table could not be applied directly via the value of index CNpk. For the consistency of NCPPM mapping, we propose procedures to obtain the modified index C*Npk and its approximately unbiased estimator C˜*Npk for three non-normal distributions, involving Log-normal, Gamma, and Weibull distributions. The values of modified index C*Npk could be used to inquire the existing popular NCPPM table of Cpk. In addition, four bootstrap methods were used to construct the lower confidence bounds of the index C*Npk, which are useful to the practitioners for making reliable decisions regarding process performance based on process yield.