Papers by Author: Po Tsang B. Huang

Paper TitlePage

Authors: Po Tsang B. Huang, James C. Chen, Yuan Tsan Jou
Abstract: The key element of the in-process surface roughness monitoring system is the decision-making model, which is utilized to analyze the input factors and then to generate a proper output. The success of the in-process monitoring system depends on the accuracy of the decision-making model. To increase the accuracy and reliability of model, it is important to reduce the variation of the inputs. To achieve this objective, an integration of regression and neural network was developed as a decision-making model in this research. In this integrated model, the regression model was applied as a filter to sort the input variables into groups. Furthermore, the grouped data was implemented to train and to generate different neural networks models to reduce the affection of input variation and increase the accuracy of the monitoring system. The input variables was first filtered by the threshold of regression model, and then analyzed by different neural networks model based on the filtered result. Finally, to evaluate the performance of the integrated model, the regression neural network and traditional neural networks were both developed for surface roughness monitoring system in an end milling operation to compare the accuracy of systems.
369
Authors: Wei Jung Shiang, Hsin Rau, Po Tsang B. Huang
Abstract: The advantages of applying Internet in E-manufacturing are lower cost, enhancing distances in remote monitoring and control, but it may have time delay problem and it could decrease system performance and stability. In order to keep network control system more stable and effective, a human-centered automation approach was proposed in the network control system. Experiments were designed to identify significant factors of these network control systems, and the task was to command a mobile robot to complete the maze exploration with five kinds of human-centered network control systems. Total elapsed time and subjective evaluation are performance indexes and defined as dependent variables. Furthermore, personal hobby and level of automation for the network control system are independent variables. According to ANOVA results, level of automation indeed has a significant impact on the network control performance. The manual control obtained highest performance in total elapsed time of completing the task. It also showed that it is feasible to apply human centered automation in E-manufacturing.
1467
Authors: Po Tsang B. Huang, James C. Chen, Chiuhsiang Joe Lin, Peng Hua Lyu, Bo Chen Lai
Abstract: The use of CNC machines is one of the successful factors in the computer integrated manufacturing (CIM). Even though the CNC machine can automatically perform the machining processes, some of the situations that may significantly influence the quality of product such as a cutting tool breakage. Therefore, to prevent the machine from damaging and ensure the quality of product, it is important to develop a system that can monitor the tool conditions. The purpose of this study is to develop a Taguchi-neural-based in-process tool breakage monitoring system in end milling operations that can monitor the tool conditions and immediately response a proper action. For an in-process tool breakage monitoring system, a neural network was applied to making decisions of monitoring. One of the disadvantages of neural network is the training processes. It is difficult to determine an optimal combination of training parameters of neural networks. Traditionally, the try-and-error method is time-consuming and without systematic base. Therefore, the optimization of training parameters for neural networks using Taguchi design was applied to training the neural network model and to enhance the accuracy of the tool breakage monitoring system.
251
Authors: James C. Chen, Chen Huan Cheng, Kung Jeng Wang, Po Tsang B. Huang, Chih Chiang Chang, Chien Jung Huang, Ti Chen Ting
Abstract: This study presented the application of Radio Frequency Identification (RFID) technology to improve the efficiency and effectiveness of warehouse management. Fixed RFID readers and antenna were installed at the receiving/shipping dock and passive tags were mounted on either storage box or pallet. RFID system can quickly and simultaneously read multiply tags, compared to the sequential reading of barcodes by handy barcode scanner in manual operations due to the inconsistency in box sizes and the locations of barcode for items of various types. Significant improvements were observed in preliminary experiments. The numbers of pallets and items processed by each operator per day were increased by 425% and 438%, respectively. The data processing time at receiving and shipping docks was reduced by more than 90%. With RFID technology, the number of operators can be reduced while maintaining current service capacity at the studied warehouse. The benefit using RFID in the warehouse management is analyzed and promoted.
297
Authors: James C. Chen, Shu Jen Hu, Po Tsang B. Huang, Chiuhsiang Joe Lin, Kuo Jung Chao, Chih Cheng Chen
Abstract: A Finite Capacity Planning Policy (FCPP) is developed for multiple color filter (CF) fabs where each fab has several identical production lines. FCPP assigns customer orders to multiple fabs and multiple lines by taking into account each fab’s available capacity and Work-In-Process (WIP) level, as well as each order’s batch size, product type, process routing, and processing time. After all orders are assigned to fabs and lines, order release time and finish time are determined by the implementation of several modules. FCPP is developed in Microsoft Visual Basic for Application (VBA), and an AutoMod simulation model is also developed. A company with two CF fabs and two lines at each fab is treated as a case study, and industrial data from these fabs are collected and used to evaluate the performance of FCPP based on the design of experiments. Preliminary simulation results show that FCPP can effectively and efficiently balance the loading between fabs and also balance the loading between lines in each fab.
637
Authors: James C. Chen, Cheng Ju Sun, Po Tsang B. Huang
Abstract: Integrated Circuit (IC) manufacturing consists of four major phases - wafer fabrication, wafer probe, IC packaging, and final test. Wafer fabrication and wafer probe belong to front-end process, while IC packaging and final test belong to backend process. In this research, integrated capacity planning system (ICPS) is proposed for IC backend manufacturing by taking into account the capacity and capability of resources. This study develops integrated infinite capacity planning integrating the capacity planning of both IC packaging and final test processes to minimize the standard deviations of machine utilization and kit utilization as well as the total extra capacity requirement (as a percentage) exceeding the capacity limit of machine and kit. ICPS applies the concept of workload leveling and infinite capacity planning on dual resources. ICPS consists of five modules: Work-In-Process (WIP) Pulling Module, Lot Release Module, Resource Selection Module, Workload Accumulation Module, and Workload Balance Module. WIP-Pulling Module pulls WIP from the end of the process route to meet the master production schedule (MPS). If WIP cannot meet the MPS requirement, Lot Release Module is executed. ICPS will use Resource Selection Module to accumulate the required resource capacity along its routine at different time buckets. Workload Accumulation Module simultaneously accumulates the workload of the major resources. Workload Balance Module levels the loading of resources by adjusting the lot’s start time using these resources. ICPS is developed in Microsoft Visual Basic and industrial data are used to test its performance. The results show that ICPS is effective and efficient to balance the workload of resources for IC backend factories including packaging and final test.
369
Authors: James C. Chen, Po Tsang B. Huang, Hui Yu Peng, Ming Chien Hung, Chih Cheng Chen, Tzu Wei Peng, Shih Wu Li, Shiuan Ming Huang
Abstract: An AutoMod simulation model was constructed for real-world 4.5th-generation (4.5G) Thin Film Transistor - Liquid Crystal Display (TFT-LCD) Color Filter (CF) fabrication plants in this study. The characteristics and data of customer orders, production facility, material handling tools, and warehouse (also called stocker) were collected from real plants. Based on these, a simulation model was developed and validated. As the simulation model was complex and had different parameters representing the operation details in the plants, it is of interest to justify the effects of these parameters to the key performance indicators such as system throughput and equipment utilization. Using fishbone analysis, sensitivity analysis and experimental design, a series of simulation experiments were conducted and the resulting simulation model was representative to the CF plants. On the basis of the simulation model with 3-dimension animation, decision makers can conduct what-if analysis for the evaluation of production control policies such as in-process capacity and tact time without disturbing the operations in the real plant. Better alternatives leading to higher system throughput can be identified in system simulation before its real implementation in the fab.
119
Showing 1 to 7 of 7 Paper Titles