Title Supply Chain Management for PC _ PC Peripheral Assembly

Document Sample
Title Supply Chain Management for PC _ PC Peripheral Assembly Powered By Docstoc
					         Department of Industrial Engineering and Management

Title: Classification of Punched Parts Based on Global and Local Shape Information
Principal Investigator: Muh-Cherng Wu
Sponsor: National Science Council
Keywords:Simplified Skeleton, Group Technology, Neural Network, Part

       This research presents an automatic classification scheme for punched parts.
 The classification is performed based on part’s global shape descriptor modeled by
 their simplified skeletons. In the first stage, the simplified skeleton of a punched
 part is represented by a sequentially coded tree structure. Based on this tree
 representation, punched parts with similar tree structure can be clustered using a
 tree matching method. In the second stage, the parts within a cluster can be further
 classified by a back-propagation neural network classifier. The two-stage
 classification scheme can be used in a punched part retrieval system aiming to
 facilitate the design and manufacturing tasks, which would greatly reduce the
 time-to-market and material inventory.

Title: An Empirical Study and Modeling of Bullwhip Effect in Electronic and
Information Industry
Principal Investigator : Muh-Cherng Wu
Sponsor: National Science Council
Keywords: Bullwhip Effect, Supply Chain, Demand Management

      Bullwhip effect is a significant phenomenon appearing in a supply chain (a
 system consisting of several units interfaced with upstream and downstream
 relationships). This effect claims that demand order variabilities are amplified as
 one moves up the supply chain. That is, a small fluctuation of demand in retail
 stores would cause higher demand fluctuation in its upstream units--distributors,
 and the demand fluctuation would become more adverse as it moves up the next
 upstream units--manufacturers. This phenomenon has been found in several
 industries such as computer, grocery, and pharmaceutical industries.
      The major impact of the bullwhip effect would cause tremendous amount of
 inventories. The inventories at stake in a supply chain typically would be in the
 order of billion US dollars. Therefore developing a model for analyzing the
 bullwhip effect and proposing the counter-actions is very important to industries.
 However, the development of such models in literature is still at an infant stage.
      We therefore propose a 3-year research plan aiming to study the bullwhip effect
 in the supply chain of our electronic and information industry. The supply chain
 includes semiconductor industry, electronic component industry, computer assembly
 industry, and the computer distribution industry in our country. The main objective
 of this research is developing an effective model, through empirical studies, for
 analyzing the bullwhip effect. Based on such a model, counter-actions for our
 electronic and information industry then can be quantitatively justified. The research
 results would be helpful to enhance the integrated productivity of our electronic and
 information industry.
       The objectives of the first year are to finish a comprehensive literature review
 and develop analytical models for the bullwhip effects. The literature review task
 has been done, and a model for modeling the scenario of a supply chain has been
       The objective of the second year is to investigative the bull-whip effects in
 computer and its peripheral industries. That is, empirical data collected from this
 industry will be used to justify and modify the analytical model developed in the
 first year.

Title:The Study of Lead-Time Management for IC Industry
Principle Investigator:Shu-Hsing Chung
Sponsor:National Science Council
Keywords:Testing Factories, Wafer Probing, Delivery Performance, Rush Lot,
            Sequence-Dependent Setup Time

      This research is the third sub-plan of the team project for “A study of Supply
Chain Management for Electronic and Information Industries”. The manufacturing
processes for IC products can be divided into four stages. This sub-plan will spend
three years developing an effective lead time management system for IC industries to
improve on time delivery performance. This research is the second-year plan of the
sub-plan, and the research scope will focus on the lead time management of
wafer-probing factories.
      Reducing manufacturing lead time of IC products is one of the competitive
factors for the IC industries. Manufacturing lead time is related to production
planning and scheduling methods. Two of the four stage processes of IC products are
testing related processes. Thus, the developing of production planning and scheduling
model for probing and testing factories to control the testing lead time of IC products
is very important.
      On time delivery is the major objective of wafer-probing factories. The
scheduling characteristics of wafer-probing factories are dynamic arrival of testing
lots, the allowance of the inserting rush orders in current dispatching list, the
sequence-dependent setup time, etc. Because the testers are expensive, the reduction
of setup times to increase capacity utilization rate is also an important performance
index. To meet both requirements of pre-determined length of lead time and reducing
setup times, this research regards the scheduling problem as vehicle routing problem
with time window constraint (VRPTW) to develop the production planning and
scheduling model.
        The production planning and scheduling model consists of three modules.
Firstly, the rough-cut capacity planning module evaluates the suitability of tester
allocation and the necessity of adjusting backup machines. Secondly, the scheduling
module decides the probing sequence for normal lots. Finally, the dispatching module
executes the insertion of rush lots in the dispatching list when the rush lots arrive.

Title: Supply Chain Management for PC & PC Peripheral Assembly Industry(II)
Principal Investigator: Hsi-Mei Hsu
Sponsor: National Science Council
Keywords: Assembly-to-Order, Safety-Stock, Advance Purchase, Discount-Pricing

       In an assembly-to-order production system, manufacturers usually perform two
strategies to reduce the impacts of customers’ demand uncertainty. One is by setting
certain levels of safety stocks, and the other is by adjusting the demand forecast. In
this study, we propose the additional strategy that is to stimulate customers to make
advanced purchase.
       In this study, we develop a discount-pricing model for customers’ advance
purchase. First, we developed two discount-pricing models for manufacturers and
customers, respectively. Then, the two discount-pricing models are combined to a
single model for determining the optimal advanced purchase ratio and the feasible
discount price.
NSC 88-2213-E-009-024

Title: Benchmarking Total Quality Systems of Taiwan’s Notebook Industry:
       Customer Satisfaction and Feedback (the second year)
Principal Investigator: Lee-Ing Tong
Sponsor: National Science Council
Keywords: Service Quality, Customer Satisfaction, Bench marking, Total Quality,
            Management, Development Level

      Notebook computer manufacturing is an important high-tech industry for
Taiwan’s economy. Statistics showed that the notebook computers from Taiwan has
constituted about 32% of the world volume. Therefore, implementing total quality
management (TQM)system to notebook manufactures may provide a key competitive
edge to increase their product’s quality, reduce cycle time and the cost.
      As the sub-project 4 of the overall project, ”Benchmarking Total Quality
Management System of Taiwan’s Notebook Industry,” this project is planned to
completed in three years. In the first year, we reviewed the literatures and construct
satisfaction model specially for notebook industry. In the second year, we surveyed
the manufacturer’ customer satisfactory information based on the model developed in
the first year. We also benchmark all notebook manufacturers at Taiwan. The study in
the second year is to

1. develop a customer satisfaction (SC) model specially for ODM or OEM notebook
   industry. Establish the CS index from the CS model and benchmark the
   CS ”development level” of Taiwan’s notebook manufacturing.
2. establish a feedback system to pass the customer satisfaction information back to
     the manufacturer’s total quality management system. The feedback system can
     help manufacturer establish a strategy or a roadmap for improving their customer

Title:The Design of Constraint-based Target Planning System for Wafer Fabrication
Principle Investigator:Shu-Hsing Chung
Sponsor:National Science Council
Keywords:Performance Measurements, Target Planning, Wafer
               Fabrication,Theory of Constraint

       The manufacturing style of semiconductor industry is gradually converting to
foundry in Taiwan nowadays. How to improve the overall competition of the
company to match the request of the customers - low cost, high yield, short cycle
time, and on-time delivery, becomes considerable and meaningful. Thus, the effective
and efficiency target planning system dominates the success of short-term production
planning and control system.
       This project will consider constraint resource management and target planning
procedures with a global viewpoint and develop a constraint-oriented target planning
systems including the management of constraint resources, the evaluation criteria of
production performance and the algorithm of defining production target levels, etc.
By its integral functions, the system will successfully achieve its capability and
improve the efficiency of manufacturing and management in wafer fabrication
       The management of capacity constraint resources(CCRs) plays a key role on
production performance due to the high complexity of wafer fabrication. Therefore,
identifying the bottleneck resources correctly in every planning level and defining the
performance criteria related to the CCRs, used as the basis of performance control and
measurement, is one of the major works of this research in the first year. Besides,
finding proper production performance measurement and designing index calculation
algorithm based on the overall considerations about performance-related CCRs, is the
another primary goal in the foremost year.
       The plan of the second year basically follows the first year. According to the
result of first year, we will keep on defining the criteria of production target control
and algorithm, analyzing the relationship, intensity, and trade-off between these
production performance criteria. The constraint-oriented concept is adopted to
develop an iteratively improved procedure to keep the production work under
reasonable target condition so as to get the better comprehensive production

Title: Evaluating the Manufacturing Performance of Multiple Production Lines in IC
Principal Investigator: Lee-Ing Tong
Sponsor: National Science Council
Keywords: Production Performance, Integrated Circuit, Yield Model, Process Defect,
           Multiple Production Lines

       Evaluating the manufacturing performance among different production lines in
IC fabrication is a relevant task. The yield is a conventional index to indicate which
production line has a better manufacturing performance. However, although the yield
index can be used to compare several lines producing the same IC product, it can not
be utilized to compare the performances of lines under circumstances such as one line
producing different products at different time intervals and different lines producing
different products. The fact that the production parameters (e.g. chip area, mask levels,
and minimum dimension) vary when fabricating various types of chips accounts for
why defects with the same number, size, or distribution may lead to different degrees
of yield loss. Therefore, in this study, we present a novel procedure to evaluate the
manufacturing performance of multiple production lines. A manufacturing
performance index (MPI) is also derived on the basis of a cluster-modified Poisson
yield model to measure the degrees of maturity of the products. In addition, two
methods, i.e. generic method, regression method, for estimating the parameters in the
yield model are also proposed. Moreover, a case study demonstrates the effectiveness
of the proposed manufacturing performance index in terms of its ability to evaluate
the multiple production lines in IC fabrication.

Title:Effect of Control-Display Gain on Remote Pointing Performance
Principal Investigator:Shang Hwa Hsu
Sponsor:National Science Council
Keywords : Human Factors, Remote Pointing, Control-Display                                     Gain,
            Human-Computer Interface

      Control-display gain of pointing devices is an important factor in determining the
effectiveness of performance. The control-display gain (C-D gain) is usually defined
as ether the relationship between control displacement and display displacement (i.e.,
first order) or the relationship between control velocity and display velocity (i.e.,
second order). The purpose of this study is to investigate the effects of the velocity
control-display gain on the performance of discrete pointing task based hand input
      Seventy-five students of National Chiao Tung University were randomly
assigned to one of 5(velocity C-D gain)×3(remote distance) experimental situations to
perform cursor positioning tasks using a remote pointing device. The results showed
that velocity C-D gain, movement amplitude, target width, and remote distance had a
significant effect on total completion time. And target width and remote distance had
a significant effect on selection error rate. High velocity C-D gain made tasks faster
and inaccurate. The optimal level of velocity C-D gain depended on the
characteristics of the task required. Velocity C-D gain also had a significant effect on
mechanism of remote pointing.

Title : A Production Inventory Model Which Considers the Dependence of
      Production Rate on Demand and Inventory Level
Principal Investigator : Chao-Ton Su
Sponsor: National Science Council
Keywords: Production Inventory;Declining Demand;Backlog;Deteriorating Items

      This project will present a production inventory model for deteriorating products
in which the production rate at any instant depends on the demand and the inventory
level. While the demand rate is assumed to decrease exponentially, shortages are
allowed and excess demand is backlogged as well. Approximate expressions will be
obtained for the production scheduling period, maximum inventory level, unfilled
order backlog and the total average cost. Special cases of the model will be discussed.
In addition, a numerical example will be illustrated to show how the theory works.
Finally, the sensitivity of these solutions to changes in underlying parameter values is
also discussed.

Title:A Systematic Approach for Developing a Foot Size Information System for Shoe
       Last Design
Principal Investigator: Der-Baau Perng,
Sponsor:National Science Council
Keywords:Shoe Last, Bi-Yariate Normal Distribution, Foot Size Grading System,
            Foot Size Information System (FSIS)

      The appropriateness of a shoe last design significantly impacts the fitness of the
shoes thus produced. Traditionally, a shoe last is designed by using numerous foot
measure data. Among all the items of foot measure information, the foot length and
joint girth are further identified as principle factors affecting shoe last design. In this
paper, foot length and joint girth were analyzed by using a bi-variate normal
distribution to obtain a more efficient foot size grading system. A set of 2486 adult
male samples collected in Taiwan area is used in this analysis to establish the norm for
the foot length and joint girth. A Foot Size Information System (FSIS) including the
percentage of population that a last can fit in is established from the analysis results.
This study provides designers the grading information that is helpful for shoe last

Title:On the Generalizations of the Capability Index Cpmk for Asymmetric Tolerances
Principal Investigator: W. L. Pearn
Sponsor:National Science Council
Keywords:Process Capability Index, Target Value, Process Yield, Process Centering,
         Process Loss, Asymmetric Tolerances, Departure Ratio

      Boyles (1994) presented a comprehensive study of several proposed indices for
processes with asymmetric tolerances, and provided a comparison between the
proposed and the existing indices on the accuracy in measuring process potential and
performance. However, none of those proposed indices reflect process capability
accurately. In fact, for fixed process standard deviation
                                                  which is between the target value T and the
center of the specification interval. Recently, Vännman (1997) considered several
generalizations of the existing indices for processes with asymmetric tolerances. But,
those generalizations either understates the process capability, or obtain their maximal
                                              ännman’s generalizations are inappropriate for
cases with asymmetric tolerances. In this paper, we consider a new generalization of
C pm k which we refer to as C k . We show that the new generalization C k is
                                      pm                                                         pm

superior to the existing generalizations. We also investigate the statistical properties
of the natural estimator of C k assuming the process is normally distributed.


Title: Operation Path Risk and Sensiivity Analyses of a Satellite Mission
Principal Investigator : Chao-Ton Su
Sponsor: National Science Council
Keywords: Risk Analysis, Failure probability, Sensitivity of Failure

      Risk associated with all possible operation path influence the success of a
sattellite mission. Optimizing the selection of operation path and reducing the
operation failure probability depend on performing operation path risk and sensitivity
analyses. Based on the operational procedures and failure data provided by National
Space Program Office, Taiwan, this project examines the feasibility of applying
appropriate models for the risk and sensitivity analyses of a satellite mission. In
addition, the risk and sensitivity for each operation path ROCSAT-1 are analyzed.
Finally, a decision tree is also proposed to ensure the success of the satellite mission.

Title:Human Factors Implementation Plan for ZPRL Human-System Interface
Principal Investigator:Shang Hwa Hsu
Sponsor:Institute of Nuclear Energy Research, Atomic Energy Council
Keywords:Human-Machine Interface, Advanced Control Room

     Previous accident/incident data showed that operator performance is critical to
system safety. Human-machine interface is one of the factors that will affect human
performance. Currently, the I&C system of ZPRL is going through a upgrading effort.
This I&C modernization program encompasses replacement of analog control with
digital control. Computer-based human-machine interface for nuclear systems, which
consolidates data from many indicators and controls into a confined display screen
estate, poses a new challenge to human-machine interface designers. Therefore, the
development of computer-based human-machine interface requires a through
procedure of human factors analysis, design, and evaluation in order to ensure that
operability and maintainability requirements are met.
      The purpose of this project is to develop the human-machine interface of ZPRL
system according to the human factors program prescribed by NRC. Through this
effort, human factors requirements can be incorporated into the design of the
human-machine interface of ZPRL. In addition, the experience gained from this
development effort can be fed into TPC Lungmen Project.

Title: The Design and Simulation of Material Transformation System for the
            Modularized Clean Production Environment
Principle Investigator:Shu-Hsing Chung
Sponsor:Industrial Technology Research Institute
Keywords : Semi-Conductor Factory, Plant Layout, Material Handling System,

      Recently, semi-conductor industry becomes the promising and booming industry
in Taiwan and contributes a lot to GNP value of our country. Since all the wafer
fabrication factories are working for creating or improving process technologies,
owning the good process capability does not guarantee to have competitive power. In
such environments, enhance the throughput and delivery performance is the only way
to keep the competitive superiority.
      On the throughput enhancement, the 300 mm wafer fab factory owns throughput
advantage since the number of dice on a 300 mm wafer is more than twice of that on
200 mm wafer and the size of each die is reducing under the new process technology.
On the delivery performance, Wu [8, 9] proposes an architecture for modularized and
economical flexible automated semi-conductor factory. It adopts the flow line layout
instead of traditional functional layout. Thus, the production lead time and the
complexity of production management can be reduced effectively. In the same time,
the delivery performance improvement can be guaranteed.
       In this research, we propose a methodology to design the plant layout and the
corresponding material handling system for Wu’s modularized wafer fab. A
simulation model will be built to simulate the effects of all related design variables on
system performance. The results of this research will provide useful information for
the decision maker during the construction of 300 mm wafer fab

Title: Developing and Implementation of the Electronic Commerce                                   and
            Network-intergrated Auction Technology for Agricultural Produces
Principal Investigator: G. R. Liang
Sponsor: Council of Agriculture
Keywords: Electronic Commerce, Bidding Strategy, Auction Theory

      The goal of this project is to implement a secrete-sharing double auction system
on Internet for egg trading. Essentially suppliers and consumers can get the free
software from the Internet for submitting bids. Then the encoded bids are distributed
to different sites immediately. Before opening the bids, major part of encoded bids are
collected and decoded to decide the winners. In this research, various secrete-sharing
algorithms are compared in order to find the best one for trading agricultural produces.
Also the underlying trading system has been implemented through Java technology.
The prototype system shows the feasibility of this new approach.

Title:A Construct of The Human Performance Safety Audit System
Principal Investigator:Shang Hwa Hsu
Sponsor:National Science Council, Executive Yuan of the Republic of China
Keywords:Human Error, Human Performance Safety Audit, Nuclear Safety

      Human performance is very important to nuclear safety. Most of accidents and
incidents were related to human error. Therefore, it is imperative that human
performance be improved. The purpose of this study has two folds: (1) to develop an
audit system for safe performance, and (2) to transfer the human performance audit
system to Taipower personnel.
      The human performance safety audit system was developed on the basis of
human error theories, human error data of nuclear power plants, and task analysis. It
contains a checklist of good practice of safe behavior performed during maintenance
tasks. These good practices include: thorough planning, efficient resource
management, pre-job reviewing and briefing, risk assessment, conscious checking,
situation awareness, memory reminders, usage of feedback information,
communication and coordination, housekeeping, post-job evaluation and
      To effectively transfer the human performance safety audit system to Taipower
personnel, course material associated with the audit system was also developed. It
consists of eight modules: (1) introduction to system safety, (2) the safety
management plan, (3) hazard analysis, (4) basic concepts of human performance and
human error, (5) causes for human error, (6) techniques to reduce human error, (7)
management of safe performance, and (8) the use of audit system.

Title: The Performance Evaluation of The Information Management and Assistance
        Programs for Small and Median Enterprises
Principal Investigator: Lee-Ing Tong
Sponsor: Chung-Shan Institute of Science & Technology of Ministry of National
Keywords: Small and Median Enterprises, Satisfaction
      The Small and Median Enterprises Bureau of Ministry of Economic Affairs
provided eleven assistance programs to help the small and median enterprises in
Taiwan to enhance their competitive power in various world markets in 1999. The
main objective of this study is to construct several scales-the small and median
enterprises satisfaction scale for the ministry of Economic Affairs’ assistance
programs, the expertise evaluation scale for the effectiveness of the assistance
programs, and the self-evaluation scale for the principal investigators of the eleven
assistance programs. The information obtained from these scales can be utilized to
improve the effectiveness of the future assistance programs sponsored by the Small
and Median Enterprises Bureau of Ministry of Economic Affairs

Title:A Quantitative Approach to Determine the Common Pallet Size for Distribution
Principal Investigator: Fuh-Hwa Liu
Sponsor:National Science Council
Keywords:Pallet, Loading, Distribution, Material Handling

      The purpose of this research is to introduce the approach to determine the
common pallet size for distribution systems. Transportation, material handling and
storage systems that associated to the pallet operation are the decision factors.

Title:Vehicle Routing Problems with Time Window
Principal Investigator: Fuh-Hwa Liu
Sponsor:National Science Council
Keywords:Heterogeneous Fleet, Heuristics, Time Windows, Vehicle Routing

     This research is concerning the vehicle routing problem with multiple vehicle
types and time window constraints. Several insertion-based savings method heuristics
are presented. The heuristics will test on twenty-four 100-customer problems
modified from the literature. We would do some experiment to demonstrate that
heuristics with the consideration of a sequential route construction parameter would
yield significantly better solution quality than all other heuristics tested.

Shared By: