Académique Documents
Professionnel Documents
Culture Documents
To cite this article: Reha Uzsoy, John W. Fowler & Lars Mönch (2018): A survey of semiconductor
supply chain models Part II: demand planning, inventory management, and capacity planning,
International Journal of Production Research, DOI: 10.1080/00207543.2018.1424363
Article views: 20
A survey of semiconductor supply chain models Part II: demand planning, inventory
management, and capacity planning
Reha Uzsoya, John W. Fowlerb and Lars Mönchc*
a
Edward P. Fitts Department of Industrial and Systems Engineering, North Carolina State University, Raleigh, NC, USA;
b
Department of Supply Chain Management, Arizona State University, Tempe, AZ, USA; cDepartment of Mathematics and Computer
Science, University of Hagen, Hagen, Germany
(Received 23 December 2016; accepted 26 December 2017)
Part I of this three-part series described semiconductor supply chains from the decision-making and functional perspec-
tives, using this as a framework to review the industrial engineering and operations research literature on the problems
arising in these supply chains. Part I then reviewed the literature on Strategic Network Design, supply chain coordina-
tion, sustainability and simulation-based decision support. This paper, Part II, reviews the areas of Demand Planning,
Inventory Management and Capacity Planning in semiconductor supply chains. Part III concludes the series by dis-
cussing models to support Master Planning, Production Planning and Demand Fulfilment in this industry.
Keywords: supply chain management; semiconductor industry; capacity planning; inventory management; demand
uncertainty
1. Introduction
The ever-growing demand for integrated circuits driven by the development of new applications such as the Internet of
Things and mobile devices, together with the increasing complexity of semiconductor production technology, has ren-
dered effective supply chain management crucial to success in the semiconductor industry. This series of three papers
uses the Supply Chain Planning Matrix of Meyr, Wagner, and Rohde (2015) as a framework to review the extensive lit-
erature in this area. Part I (Mönch, Uzsoy, and Fowler, forthcoming-a) reviewed the literature on the Strategic Network
Planning, supply chain coordination, sustainability and simulation-based decision support. The present paper continues
by reviewing the literature on Demand Planning, Inventory Management and Capacity Planning. The rectangles framed
red in Figure 1 indicate the planning functions discussed in the remainder of this paper.
Accurate demand estimates are clearly important for all aspects of supply chain management. In the semiconductor
supply chain, the Capacity Planning activity, which determines how to adjust the capacity of existing facilities in the
face of changing demand, takes place in the medium term due to the long lead times required for acquiring new produc-
tion equipment. As mentioned in Part I, random co-production and alternative bills of material (BOMs) allows many
combinations of alternative manufacturing processes and intermediate items to produce a given end item. Hence, the
problems of Inventory Management are closely related to those of Master Planning and Production Planning discussed
in Part III (Mönch, Uzsoy, and Fowler, forthcoming-b) of this series.
After discussing Demand Planning in the following section, we review the literature on Inventory Management in
Section 3. Section 4 discusses Capacity Planning approaches for semiconductor supply chains. A summary and conclu-
sions are provided in Section 5, together with some directions for future research.
2. Demand planning
2.1 Problem setting
The Demand Planning activity shown in Figure 1 produces the demand forecasts at different levels of aggregation (pro-
duct family, geographical region, individual SKU and time) that are crucial to the planning processes of any supply
chain. Quantifying the uncertainty in demand forecasts is also important for planning safety stocks across the supply
chain and for formulating contractual arrangements with customers and suppliers. Capacity Planning decisions, which
Capacity Planning
Demand
Planning
mid-
term
Master Planning
Inventory
Manage-
Material ment
Require- Production Planning
ments
short- Planning
Demand
term
Fulfillment
Transport
Scheduling
Planning
are critical to success due to the high capital costs of equipment and facilities, require accurate demand forecasts. Synte-
tos et al. (2016) review the literature on forecasting in the supply chain context. However, the demand planning activity
must go beyond simple statistical forecasting in order to be effective; expert judgement from different sources must be
taken into account, as well as inputs from other supply chain partners (Kilger and Wagner 2015). In many semiconduc-
tor companies the term ‘demand planning’ is used to describe the allocation of customer demand among different facili-
ties or among different product lines. The cross-functional activities referred to as Sales and Operations Planning taking
place between the manufacturing and sales organisations (Vollmann et al. 2005) are also relevant to this function. We
discuss the former allocation under the heading of Master Planning in Part III (Mönch, Uzsoy, and Fowler, forthcom-
ing-b), and the latter in Part I (Mönch, Uzsoy, and Fowler, forthcoming-a) under the heading of supply chain coordina-
tion. A number of commercial software systems are available to support different aspects of this activity (cf. Roundy
2001).
Bannister, Bickford, and Swanke (2014) give an interesting overview of how this function is performed in one par-
ticular company. They also suggest a demand smoothing technique that begins by classifying demand forecasts into two
groups, prioritised and non-prioritised. The former represents demand that needs to be met with high probability while
the latter represent less reliable or urgent demand. Their procedure follows the logic of the load-levelling (heijunka) pro-
cedure used by Toyota (Liker 2004) to maintain an even manufacturing output by distributing priority demand in excess
of the cumulative average over time periods preceding the peak, while filling deficits with non-priority orders. They
evaluate the effectiveness of this procedure using safety stock calculations on a sample of items and find that it reduces
safety stock requirements.
Several factors combine to render demand forecasting in the semiconductor industry especially challenging:
(1) Short product life cycles limit the amount of data that can be collected and to which statistical forecasting
approaches can be applied. The drivers of demand for products at different stages of the life cycle can be quite
different, as can the consequences of failing to meet demand. Early in the life cycle, demand is driven by cus-
tomers’ willingness to design the new device into their products (design wins), so failing to meet demand can
lead to significant losses of revenue if customers turn to a competing product. Later in the life cycle, a firm may
International Journal of Production Research 3
encourage its customers to discontinue use of an older product and adopt a newer, more profitable one. In yet
other cases, a device may become extremely valuable as replacement parts for long-lived legacy systems that are
difficult or expensive to replace, altering its demand characteristics and its role in the firm’s product portfolio.
(2) Some products in a firm’s portfolio may be substitutes, while others may be complementary, such as a particular
CPU and its associated chipset.
(3) The long lead times in semiconductor supply chains due to the long fab cycle times (Mönch, Fowler, and Mason
2013) require demand forecasts to be made well in advance of demand realisation, increasing forecast uncer-
tainty. The same difficulty arises for demand forecasts supporting capacity acquisition and expansion decisions
due to the long procurement lead times for equipment and tooling.
(4) The same device may be sold in different markets with quite different characteristics, and hence very different
demand drivers.
(5) The prices of most semiconductor devices decrease significantly over their life cycle (Aizcorbe, Oliner, and
Sichel 2008; Flamm 1993). A new product will command a premium price early in its life cycle, but prices
decrease rapidly as competing products enter the market (Leachman and Ding 2007), affecting the demand for
the product.
denotes the realised demand in period t. When the time advances to the next period s + 1, additional information
becomes available allowing new demand forecasts to be generated. In the additive MMFE, the size of the updates is
assumed to be unrelated to the size of the forecast, so the forecast update for period t computed in period s is repre-
sented by a random variable est ¼ Dst Ds1;t , which is assumed to be normally distributed with mean 0. This implies
that the forecasts Dst represent the conditional expectation of demand in period t given all available information at time
s, i.e. Dst ¼ E½Dtt jFs where Fs is a σ-field describing the information available at the end of period s such that
Fs ⊆ Fs+1. Therefore, the successive demand forecasts for period tfDst ; s t g form a martingale such that
Dst ¼ E ½D~st jFs for all s ~s t. The martingale property implies that the forecast updates in any period are unbiased
and uncorrelated with updates that occur in other periods. In the multiplicative MMFE version, the forecast updates are
proportional to the size of the demand forecasts. More detailed discussion can be found in Norouzi and Uzsoy (2014).
In contrast to the growth curve approaches, the approach of Cakanyildirim and Roundy (2002c) is not appropriate
for products with short life cycles, but rather for products such as memory, microprocessors and hard disc controllers
for whose functionality there is a stable demand that may be met by successive generations of products. The main con-
cern is with forecasts of capacity requirements for the purposes of equipment acquisition, which requires aggregate fore-
casts over a longer term. Thus, the goal is forecasting the market size for a particular device and the fraction of this
market that the products of a given firm will be able to capture. The primary result of this work is the statistical charac-
terisation of the forecast errors affecting these demand forecasts.
Cakanyildirim and Roundy (2002c) aggregate devices into product families based on both functionality (e.g. mem-
ory, microprocessor) and technology level (line width or minimum feature size). Within each product family, a given
technology will follow an S-shaped growth curve as it is introduced and adopted in the market, reaches maturity and is
eventually displaced by a newer technology. At a given point in time, a firm will be interested in forecasting the
demands for a product family involving the current technology as well as all future technologies by which that family
may be produced. Figure 2 illustrates the demand for a hypothetical family of memory products produced using three
different technologies (CMOS 8, CMOS 10 and CMOS 12) over time. The demand for the product family at a particu-
lar time is given by the sum of demands over all technologies by which the family can be produced. The estimated mar-
ket size for the product family is given by the upper envelope of the individual growth curves, while the vertical
distance between individual technology curves gives the demands for the different technologies.
The authors develop a framework to evaluate the quality of semiconductor demand forecasts assuming the additive
MMFE model of Heath and Jackson (1994), which is appropriate for time-stationary time series. They use historical
observations of realised demand and forecasts, in particular the sequence of forecasts made for a given period t until its
actual demand was observed, to estimate the variance-covariance matrices characterising the forecast updates. The pur-
pose of this exercise is to estimate the variance of the forecast errors related to future periods, allowing better quantifica-
tion of the risks associated with equipment acquisition decisions. Extensive computational experiments and sensitivity
analyses using both randomly generated and industrial data indicate that the proposed approach performs well.
Figure 2. Technology migration within a product family (Cakanyildirim and Roundy 2002c).
International Journal of Production Research 5
Other authors have taken an alternative approach to demand forecasting by building stochastic models of the beha-
viour of individual customers. Higle and Kempf (2011) model the evolution of an individual customer order by assum-
ing that it is initially placed F periods in advance of its original requested delivery date, resulting in a forecast horizon
of F periods. Using the notation of Heath and Jackson (1994), let dst denote the demand forecast for period t made in
period s. In each period t, this forecast can be updated in both quantity and timing; a fraction dnst will become due in
period t–n, and the amount due in period t will be altered by a fraction γst. Thus, the demand forecast dst evolves fol-
lowing the expression
X
nmax
dst ¼ dtnþ1;s1 dntnþ;s1 1 þ ctnþ1;s1 ; (1)
n¼nmin
where nmin and nmax define the time interval over which the demand represented in an initial order may be reallocated.
The resemblance to the MMFE model is evident.
Adriaansen et al. (2013) use agent-based simulation to model the sales of microprocessors in the high-end gaming
market, focusing on consumers’ decision to purchase a more powerful computer. Each consumer is represented by an
agent who makes purchasing decisions based on its internal logic. Agreement between the simulation models and
observed sales can be obtained by calibrating two parameters for each agent: the number of games lost since the pur-
chase of the last system, and the amount of money available for their gaming hobby. Once the simulation model is cali-
brated and validated, it can be used to estimate several quantities of interest to marketing, such as the optimal timing of
new device releases and the distribution of the time between purchases for each consumer, that are not readily accessible
from conventional sales data.
3. Inventory management
3.1 Problem setting
The inventory management problems faced by the semiconductor industry resemble those in many other industries: how
to determine the levels and locations of safety stocks throughout the supply chain to maintain an appropriate trade-off
between customer service and costs. Problems in the semiconductor industry are complicated by several factors: long
lead times and uncertain yield in wafer fabrication, random co-production at final test, alternative BOMs and the global
nature of the supply chain. These factors also affect how safety stocks are accounted for in Production Planning, which
will be discussed in Part III of this survey.
Stochastic inventory models (Axsäter 2015; Zipkin 2000) show that the amount of safety stock required to provide a
specified level of customer service depends on the distribution of the lead time demand – the amount of demand arising
during the time interval between a replenishment order being placed and its being received. In most semiconductor sup-
ply chains, the orders that replenish inventories, notably the die bank inventory of probed wafers between wafer probe
and assembly/test and the finished goods inventories between final testing and distribution, are placed with production
systems subject to considerable uncertainty. For die bank inventory, the replenishment lead time consists mainly of the
cycle time of the wafer fab, which is long and subject to uncertain yield. Finished goods inventory is replenished by the
assembly and final testing operations that are subject to uncertain yield and random co-production due to binning.
Inventory management problems also arise in semiconductor supply chains for spare parts for machines. Spare parts
are required for both preventive and breakdown maintenance. Missing spare parts can lead to long down times, but
spare parts inventory is expensive and can become obsolete when production equipment models change.
Inventory control models that have been developed to address these problems can be grouped under three principal
headings:
(1) the management of inventories in the presence of co-production,
(2) the location and sizing of safety stocks in the supply chain and
(3) the management of spare parts inventories to support semiconductor manufacturing.
International Journal of Production Research 7
There is also an extensive literature on inventory problems in the face of random yield, where the amount received
after an order is placed is uncertain; for reasons of brevity we shall not discuss this literature, referring the reader to the
review by Yano and Lee (1995) as well as the extensive literature on the topic (Nasr, Maddah, and Salameh 2013;
Wright and Mehrez 1998).
formulations of this problem. The first is that of stochastic service, where the amount of material the customer receives
when their order is placed is a random variable due to the possibility of stockouts at stages inside the supply chain.
These models are based on the classical formulation by Clark and Scarf (1960), and have been addressed extensively in
the literature (Axsäter 2015; Diks, de Kok, and Lagodimos 1996; Inderfurth 1994; Zipkin 2000). The development of
optimal policies for this type of system involves stochastic dynamic programming methods that do not scale well to
large networks. In addition, the development of solution methods for general network topologies has proven difficult;
although solutions exist for serial and assembly networks (Federgruen and Zipkin 1984; Rosling 1989), divergent net-
work structures have proven challenging.
The second approach, the Guaranteed Service Approach (GSA), originates in the work of Simpson (1958) and its
extension by Graves and Willems (2000), and has attracted considerable interest (Eruguz et al. 2016). The GSA assumes
that each stage of the supply chain operates under a periodic-review constant-base-stock policy. Under this approach,
delivery of the entire order within a specified service time is guaranteed as long as the total demand observed by the
inventory point does not exceed a maximum reasonable demand. Any shortages arising when demand exceeds the maxi-
mum reasonable demand will be managed as an exception to routine inventory management procedures whose costs are
not considered in the model. The GSA model results in a deterministic equivalent optimisation problem that can be
solved using dynamic programming (e.g. Graves and Willems 2000) or mathematical programming (e.g. Magnanti et al.
2006; You and Grossmann 2011).
Wieland et al. (2012) use a guaranteed service model to study the location and quantity of safety stocks for Intel’s
Channel Supply Demand Operations. This supply chain consists of three echelons: finished CPU inventories at assem-
bly/test plants, boxing sites that configure the CPUs with added components such as heat sinks and package them, and
finished goods warehouses from which they are shipped to meet customer demand. The basic approach is the model of
Graves and Willems (2000), extended to consider general acyclic supply chain networks (Humair and Willems 2011)
and periodic as opposed to continuous review (Bossert and Willems 2007). The authors discuss the determination of
appropriate service levels, the impact of capacity constraints in determining which safety stock targets cannot be met
and the selection of items to include in the model due to decreasing returns on modelling low-volume items. Of particu-
lar interest is the discussion of how to incorporate the safety stock targets computed by the GSA model into the Produc-
tion Planning models used to determine production quantities at the boxing plants. Tian, Willems, and Kempf (2011)
propose an iterative scheme combining a dynamic programming model for safety stock targets and a linear program-
ming model for Production Planning. Manary and Willems (2008) discuss the modifications to the GSA model required
to address biased demand forecasts. Hung and Chang (1999) address the incorporation of safety stocks into Production
Planning models by deriving a simple formula to estimate the standard deviation of the total cumulative production of
an item in the face of random lead times and yields. The performance of the approach is examined using simulation and
found to be satisfactory.
Taking a stochastic service approach, Katircioglu and Gallego (2011) address the problem of determining base stock
levels at each location of a multi-echelon supply chain in the presence of random yields, where a specified service level
(stockout probability) must be guaranteed to customers. They develop a heuristic based on transforming the problem
from one of minimising cost subject to service constraints to one of minimising cost where backlogging is allowed at a
modified cost. Numerical experiments show that the proposed heuristics have promising performance.
Huang and Song (2010) consider a two-stage production system whose first stage uses a single raw material to pro-
duce multiple grades of semi-finished parts whose distribution is uncertain. The second stage processes the semi-finished
products to a specific grade, either the original grade of the semi-finished product or a lower grade to which the semi-
finished product is downgraded. They propose a modified base stock policy and extend it to allow modification of
downgrading decisions at the second stage to take advantage of risk pooling over the shared semi-finished products.
The downgrading decisions at the second production stage are formulated as a linear program (LP) whose objective is
to equalise service levels across grades.
Brown, Lee, and Petrakian (2000) examine the use of product postponement or delayed differentiation as an inven-
tory reduction strategy. The authors consider a semiconductor manufacturer producing devices that allow a generic,
semi-finished product to be configured late in the production process, in some cases by the final customer themselves.
Subsequent work reported in Cheng et al. (2005) develops an optimisation model to minimise stockout costs subject to
budget constraints on the inventory holding cost at each inventory location. Numerical experiments suggest that most of
the inventory budget should be allocated to end items when holding costs are low, while for high holding costs pooling
of inventory at die bank is optimal. Significant reduction in inventory holding cost is obtained by allowing different
safety stock levels for different items, as opposed to setting all safety stocks to the same number of daily demands.
International Journal of Production Research 9
4. Capacity planning
4.1 Problem setting
Capacity Planning is concerned with estimating how much equipment is needed to support a given demand or, equiva-
lently, how much a given set of equipment can produce while maintaining acceptable performance (e.g. competitive
cycle times). In semiconductor manufacturing, this is generally a medium term decision over a horizon of one to three
years due to the long lead time to obtain new equipment (Cakanyildirim and Roundy 1999). In practice, Capacity Plan-
ning has traditionally been done at the level of the factory rather than the supply chain. The most common approach is
a deterministic calculation based on mean production rates at tool groups, with allowances for routine detractors such as
preventive maintenance, set-ups and yield losses. Geng and Jiang (2009) review strategic Capacity Planning in the semi-
conductor industry, while Wu, Erkoc, and Karabuk (2005) review the capacity the literature in high-tech industries
including semiconductor manufacturing. Geng and Jiang (2009) classify Capacity Planning approaches into three main
types:
(1) Static Capacity Models
(2) Neighbourhood Search Methods – we will call these Performance Evaluation methods to avoid confusion with
neighbourhood search methods in metaheuristics
(3) Mathematical Programming Methods.
availability and yield. The number of equipment categories tracked varies between factories, but is generally between 5
and 15. Few papers on static capacity models have appeared in archival journals but numerous conference presentations
and papers (e.g. Occhino 2000; Witte 1996) attest to their widespread use.
variables in CAPS are the product starts into the manufacturing line and the allocation of the production times to the
tool groups, while the constraints determine the required processing times for all tool groups and cascades. The
multi-period LP model seeks to maximise profit with penalties for exceeding the available capacity of a tool group or
cascade.
Çatay, Erengüç, and Vakharia (2003) consider alternative tool groups in their multi-period mixed integer program-
ming (MIP) formulation that minimises the sum of tool operating costs, inventory holding costs and procurement cost
of primary tools. Since the model is computationally intractable, they develop a Lagrangian relaxation heuristic by mov-
ing the capacity constraints into the objective function. They compare the performance of the heuristic to a lower bound
developed in the paper and obtain good solutions.
Romauch and Klemmt (2015) propose a capacity planning approach for wafer fabs based on linear programming
that considers alternative tool groups and seeks to optimise the product mix. The resource pooling concept from heavy
traffic queuing theory (Harrison and López 1999) is applied. Two resources are called connected if at least one product
exists that can be processed on both resources. Connected components are called closed machine sets. Resource pools
are closed machine sets that lead to the minimal maximum load. The load on a resource pool cannot be decreased with-
out exceeding the load on resources outside the pool. Decomposition approaches for the resulting LPs based on con-
nected components and resource pools are proposed to tackle large instances.
Motivated by semiconductor manufacturing, Kim and Uzsoy (2009) develop a discrete-time capacity expansion (and
contraction) model that considers the congestion in a factory. Congestion in a period is modelled using clearing func-
tions (CF) that relate the expected output of a production resource in a planning period to its expected workload over
the period. CFs are discussed in more detail in Part III of this survey. The decision variables are the number of machi-
nes to introduce and to salvage in each workcenter in each period. The objective function to be minimised is the sum of
machine acquisition, salvage and maintenance costs in each period. The authors develop an exact method for a single
workcenter and three heuristics for the multiple workcenter case and compare the performance of the multiple workcen-
ter heuristics. Kim and Uzsoy (2008) develop a column generation procedure to approximately solve the problem, and
use it in a branch and price algorithm that finds the optimal solution for larger problem instances faster than their previ-
ous methods. The authors indicate that a prime area for future research is to develop multiple product CFs of the type
addressed by Albey, Bilge, and Uzsoy (2014, 2017).
Chien and Zheng (2012) propose a capacity expansion model for wafer fabs that minimises the maximum regret
while allowing capacity to be added to the fab, where regret is defined as either oversupply (underutilised capacity) or
shortage (lack of capacity). Oversupply and shortage can be given different weights. Chen, Chen, and Liou (2013a) dis-
cuss linear programming models for capacity planning in foundries where expensive auxiliary resources such as masks
must also be considered in capacity expansion decisions.
Several papers have focused on capacity planning for back-end operations. Similar to the paper by Chen, Chen, and
Liou (2013a) for foundries, auxiliary resources must also be considered in capacity expansion decisions. Wang and Lin
(2002) formulate an IP model to determine the type and number of testers required to meet upcoming demand under a
budget constraint and how to allocate tester capacity among the orders to maximise profit. The authors develop a genetic
algorithm and demonstrate that it performs well. Zhang et al. (2007) describe a two-level hierarchical planning method-
ology that uses MIP for mid-range capacity planning followed by short-range (weekly) Capacity Planning (Production
Planning) for back-end operations. The details of the mid-range Capacity Planning portion are described in Zhang et al.
(2007) and the Production Planning portion in Zhang et al. (2006). The Capacity Planning portion determines whether
the demand over the next few months can be met with existing capacity and what additional resources should be pur-
chased if existing capacity is insufficient. It also determines how customer demands should be allocated among different
factories and subcontractors. The model considers multiple stages (both assembly and test) and multiple facilities per
stage. The objective is to minimise the weighted sum of several objectives including unmet demand, subcontract alloca-
tion, inventory, number of assembly machines, number of testers, number of kit combinations required and cross ship-
ping between assembly, test and subcontractor sites. The model was used at Intel for a number of years to support their
back-end operations.
Habla and Mönch (2008) develop and test a model that determines production volumes and capacity allocations for
multiple product types for an entire network of front-end and back-end facilities over the medium term. Since capacity
expansion is not included in the model, it can be solved using linear programming instead of mixed integer program-
ming. Notable aspects of the model include: (1) consideration of both front-end and back-end; (2) commodities and cus-
tom products and some products with intermediate characteristics are included, each with a different level of demand
volatility; (3) both committed and forecasted orders are explicitly considered; and (4) products can be stored inside the
fab (wafer bank), before assembly (die bank), or after being completed (finished goods inventory). Backorders are
allowed and four bottleneck resources are included in the front-end and four in the back-end. An interesting design of
12 R. Uzsoy et al.
experiments is used. They have distributions for most of the parameters in the model and generate 25 unique problem
instances and report results comparing the full model with scenarios that: (1) only model the most critical bottleneck
resource in the front-end and back-end; (2) there is a high level of forecast inaccuracy; and (3) a combination of 1 and
2. They determine that there is little deviation in the solutions under the first scenario, but large deviations under the
second and third scenarios. An area for future research is to use the model in a rolling horizon fashion and to study the
interaction of their approach with more detailed Master Planning approaches.
Wang and Su (2015) develop a deterministic MIP model for strategic capacity planning for light emitting diode
(LED) makers in Taiwan considering multiple wafer fabs, assembly facilities and demand centres. The model only
allows new equipment to be added in assembly plants, and can consider individual orders. They provide a numerical
example that includes wafer fabs, assembly facilities and demand centres in Taiwan and mainland China.
Another group of authors from Taiwan develop a series of models for thin film transistor-liquid crystal display
(TFT-LCD) manufacturing, which is similar to semiconductor manufacturing and consists of three stages: Array, Cell
and Module. Chen, Chen, and Lu (2013b) provide a deterministic capacity allocation and expansion model that max-
imises long-term economic performance, considering only the bottleneck stage (Array) without subcontracting. The
basic model is deterministic and only allows for expanding the capacity of product groups by adding auxiliary tools.
They extend this model to consider the purchase of additional bottleneck equipment, inventory decisions and multiple
demand scenarios (but no recourse). Chen, Lin, and Wu (2014) allows capacity expansion at all manufacturing stages,
not just in the Array stage. They also consider the product life cycle and develop a two-phase shadow-price heuristic to
solve this problem.
and impose bounds on other variables based on the solution from the previous step; and (4) apply branch-and-bound
heuristics to find integer solutions. They compare the solutions from four different branch-and-bound heuristics and
determine that a heuristic that solves the problem one period at time provides relatively quick solutions, but a heuristic
that considers two periods at time with integer first period variables and continuous second period variables provides
better solutions in longer solution times. They also briefly discuss the business processes necessary to use SCAPS
effectively.
Geng, Jiang, and Chen (2009) present a two-stage stochastic programming model similar to SCAPS that considers
demand uncertainty. They establish a goal of 90% utilisation for each tool group and then use scenarios with different
values of Overall Equipment Effectiveness (OEE) (Nakajima 1988) to reflect capacity uncertainty, considering both the
procurement cost and the operating costs for tools in each period. The model considering both uncertainties is more
robust to uncertain capacity than that considering only uncertain demand.
Cakanyildirim and Roundy (2000, 2002a, 2002b) develop a single product model inspired by the Theory of Con-
straints (Goldratt and Fox 1986) that determines when fab capacity should be added in order to minimise the sum of
tool costs, lost sales costs and floor-shell expansion costs under stochastically increasing uncertain demand. The model
is called FIx Four EXpansions (FIFEX) and it treats the expansion times as continuous variables instead of considering
expansions at discrete times. They use a Bottleneck Purchasing Policy (BPP) that orders the tools to be added based on
minimum wafer production capacity. They show that using discrete time variables does not have a major negative
impact on performance if the time buckets are kept relatively small (e.g. six months or less). They also show that fre-
quent (e.g. quarterly) forecast and planning updates can provide significant benefit. Huh and Roundy (2005) build upon
Cakanyildirim and Roundy (2002b) by considering multiple products and develop a divide and conquer algorithm to
solve the problem. Huh, Roundy, and Çakanyildirim (2006) refine the solution algorithm, add tool retirements, and use
variance reduction techniques when sampling demand scenarios to estimate lost sales.
Zhang et al. (2004) take a different approach when considering the uncertainty in demand of multiple products over
time. They use a fairly large number of demand rays (each representing a different product mix) and minimise the sum
of investment costs and expected lost sales costs. They assume that all tool purchases are made in period 1, that there is
minimal finished goods inventory, and that there are no backorders between periods. These assumptions allow the prob-
lem to be decomposed into time periods once the tool purchase decisions have been made. They also assume that pro-
duction will always be proportional to demand rather than maximising the profit from production when capacity is less
than demand. They go on to show that under these assumptions the problem is equivalent to a maximum flow/minimum
cut problem and realistic problem instances can be solved in reasonable time.
Chen and Lu (2012) extend the work of Chen, Chen, and Lu (2013b) and Chen, Lin, and Wu (2014) for TFT-LCD
manufacturing discussed at the end of Section 4.4.1 by using a two-phase shadow price heuristic to solve a two-stage
stochastic MIP model similar to the deterministic model of Chen, Chen, and Lu (2013b) considering only the Array pro-
cess. The second stage allocates the capacity based on the realised demand. Lin, Chen, and Chu (2014) considers the
same problem as Chen and Lu (2012) but solves it as a stochastic dynamic program with an embedded LP model to
generate a capacity planning policy as the demand is realised and updated.
there is need for research evaluating the performance of these alternative models under realistic conditions where some
of their assumptions may be violated. It is perfectly possible that the models and solution techniques are robust to some
assumptions; it may also be possible to identify some of their limitations and address them through carefully tested
approximations.
Acknowledgements
The authors would like to thank the attendees of the Dagstuhl Seminar 16062 ‘Modeling and Analysis of Semiconductor Supply
Chains’ held in 7–12 February 2016. This paper profits by the insights that are a result of the active contribution of every attendee.
The seminar would not have been possible without the generous support of the Leibniz Center for Informatics, which the authors also
gratefully acknowledge. We would like to thank Kenneth Fordyce and John Milne who read and commented on an earlier draft of this
paper.
Disclosure statement
No potential conflict of interest was reported by the authors.
References
Adriaansen, T., D. Armbruster, K. Kempf, and H. Li. 2013. “An Agent Model for the High-End Gamers Market.” Advances in Com-
plex Systems. A Multidisciplinary Journal 16 (7): 1350028.
Aizcorbe, A., S. D. Oliner, and D. Sichel. 2008. “Shifting Trends in Semiconductor Prices and the pace of Technological Change.”
Business Economics 43 (3): 23–39.
Akçali, E., M. Davis, R. D. Hamlin, T. McCullough, T. Teyner, and R. Uzsoy. 2001. “A Decision Support System for Spare Parts
Management in a Wafer Fabrication Facility.” IEEE Transactions on Semiconductor Manufacturing 14 (1): 76–78.
Albey, E., Ü. Bilge, and R. Uzsoy. 2014. “An Exploratory Study of Disaggregated Clearing Functions for Production Systems with
Multiple Products.” International Journal of Production Research 52 (18): 5301–5322.
Albey, E., A. Norouzi, K. Kempf, and R. Uzsoy. 2015. “Demand Modeling with Forecast Evolution: An Application to Production
Planning.” IEEE Transactions on Semiconductor Manufacturing 28 (3): 374–384.
Albey, E., Ü. Bilge, and R. Uzsoy. 2017. “Multi-Dimensional Clearing Functions for Aggregate Capacity Modelling in Multi-Stage
Production Systems.” International Journal of Production Research 55 (14): 4164–4179.
Axsäter, S. 2015. Inventory Control. 3rd ed. New York: Springer.
Aytac, B., and S. Wu. 2011. “Modelling High-Tech Product Life Cycles with Short-Term Demand Information: A Case Study.” Jour-
nal of the Operational Research Society 62: 425–432.
Aytac, B., and S. Wu. 2013. “Characterization of Demand for Short Life-Cycle Technology Products.” Annals of Operations Research
203: 255–277.
Bannister, A. R., J. P. Bickford, and K. V. Swanke. 2014. “Demand Smoothing.” IEEE Transactions on Semiconductor Manufacturing
27 (3): 335–340.
Barahona, F., S. Bermon, O. Günlük, and S. Hood. 2005. “Robust Capacity Planning in Semiconductor Manufacturing.” Naval
Research Logistics 52 (5): 459–468.
Bard, J., K. Srinivasan, and D. Tirupati. 1999. “An Optimization Approach to Capacity Expansion in Semiconductor Manufacturing
Facilities.” International Journal of Production Research 37 (15): 3359–3382.
Bermon, S., and S. Hood. 1999. “Capacity Optimization Planning System (CAPS).” Interfaces 29 (5): 31–50.
Birge, J., and F. Louveaux. 2011. Introduction to Stochastic Programming. 2nd ed. New York: Springer.
Bitran, G., and S. Dasu. 1992. “Ordering Policies in an Environment of Stochastic Yields and Substitutable Demands.” Operations
Research 40: 999–1017.
Bitran, G., and T. Leong. 1992. “Deterministic Approximations to Co-Production Problems with Service Constraints and Random
Yields.” Management Science 38: 724–742.
International Journal of Production Research 15
Blackstone Jr, J. H., and J. Cox. 2002. “Designing Unbalanced Lines - Understanding Protective Capacity and Protective Inventory.”
Production Planning & Control 13 (4): 416–423.
Bossert, J., and S. Willems. 2007. “A Periodic Review Modeling Approach for Guaranteed Service Supply Chains.” Interfaces 37:
420–436.
Box, G., G. Jenkins, G. Reinsel, and G. Ljung. 2015. Time Series Analysis: Forecasting and Control. 5th ed. Hoboken, NJ: Wiley.
Brockwell, P., and R. Davis. 1991. Time Series: Theory and Methods. 2nd ed. New York: Springer.
Brown, A., H. Lee, and R. Petrakian. 2000. “Xilinx Improves Its Semiconductor Supply Chain Using Product and Process Postpone-
ment.” Interfaces 30 (4): 65–80.
Brown, S., T. Hanschke, I. Meents, B. Wheeler, and H. Zisgen. 2010. “Queueing Model Improves IBM’s Semiconductor Capacity
and Lead-Time Management.” Interfaces 40 (5): 397–407.
Cakanyildirim, M., and R. Roundy. 1999. Demand Forecasting and Demand Planning in the Semiconductor Industry. Technical
Report 1229, SORIE, Cornell University. Accessed February 8, 2017. https://www.utdallas.edu/metin/Research/forecastcapacity
survey.pdf
Cakanyildirim, M., and R. Roundy. 2000. Optimal Capacity Expansion Strategies under Uncertainty. Technical Report 1250, SORIE,
Cornell University.
Cakanyildirim, M., and R. Roundy. 2002a. “Evaluation of Capacity Planning Practices for the Semiconductor Industry.” IEEE Trans-
actions on Semiconductor Manufacturing 15 (3): 331–340.
Cakanyildirim, M., and R. Roundy. 2002b. Optimal Capacity Expansion and Contraction under Demand Uncertainty. Ithaca, NY:
School of Operations Research and Industrial Engineering.
Cakanyildirim, M., and R. Roundy. 2002c. “SeDFAM: Semiconductor Demand Forecast Accuracy Model.” IIE Transactions 34 (5):
449–465.
Carmon, T., and S. Nahmias. 1994. “A Preliminary Model for Lot Sizing in Semiconductor Manufacturing.” International Journal of
Production Economics 35: 259–264.
Çatay, B., S. Erengüç, and A. Vakharia. 2003. “Tool Capacity Planning in Semiconductor Manufacturing.” Computers & Operations
Research 30 (9): 1349–1366.
Chen, L., and H. Lee. 2009. “Information Sharing and Order Variability Control under a Generalized Demand Model.” Management
Science 55 (5): 781–797.
Chen, T., and H. Lu. 2012. “Stochastic Multi-Site Capacity Planning of TFT-LCD Manufacturing Using Expected Shadow-Price
Based Decomposition.” Applied Mathematical Modelling 36 (12): 5901–5919.
Chen, A., K. Yang, and Z. Hsia. 2008. “Weighted Least-Square Estimation of Demand Product Mix and Its Applications to Semicon-
ductor Demand.” International Journal of Production Research 46 (16): 4445–4462.
Chen, Y. Y., T. L. Chen, and C. D. Liou. 2013a. “Medium-Term Multi-Plant Capacity Planning Problems Considering Auxiliary Tools
for the Semiconductor Foundry.” International Journal of Advanced Manufacturing Technology 64: 1213–1230.
Chen, T., Y. Chen, and H. Lu. 2013b. “A Capacity Allocation and Expansion Model for TFT-LCD Multi-Site Manufacturing.” Jour-
nal of Intelligent Manufacturing 24 (4): 847–872.
Chen, T., J. Lin, and C. Wu. 2014. “Coordinated Capacity Planning in Two-Stage Thin-Film-Transistor Liquid-Crystal-Display (TFT-
LCD) Production Networks.” Omega 42 (1): 141–156.
Cheng, F., M. Ettl, G. Lin, and D. Yao. 2005. “Inventory Management in High Technology Value Chains.” Chap. 3 in Supply Chain
Management on Demand: Strategies and Technologies, Applications,, edited by C. An, and H. Fromm, 37–63. Berlin Heidel-
berg: Springer.
Chien, C., and J. Zheng. 2012. “Mini–Max Regret Strategy for Robust Capacity Expansion Decisions in Semiconductor Manufactur-
ing.” Journal of Intelligent Manufacturing 23: 2151–2159.
Chien, C., W. Wang, and J. Cheng. 2007. “Data Mining for Yield Enhancement in Semiconductor Manufacturing and an Empirical
Study.” Expert Systems with Applications 33: 192–198.
Chittari, P., and N. Raghavan. 2006. “Support Vector Based Demand Forecasting for Semiconductor Manufacturing.” Proceedings
IEEE International Symposium on Semiconductor Manufacturing, Tokyo, 261–264.
Choi, T., C. Hui, and Y. Yu, eds. 2014. Intelligent Fashion Forecasting Systems: Models and Applications. Heidelberg: Springer.
Chou, Y.-C., and C.-S. Wu. 2002. “Economic Analysis and Optimization of Tool Portfolio in Semiconductor Manufacturing.” IEEE
Transactions on Semiconductor Manufacturing 15 (4): 447–453.
Chou, Y., and R. You. 2001. “Resource Portfolio Planning Methodology for Semiconductor Wafer Manufacturing.” The International
Journal of Advanced Manufacturing Technology 18 (1): 12–19.
Chou, Y.-C., C.-T. Cheng, F.-C. Yang, and Y.-Y. Liang. 2007. “Evaluating Alternative Capacity Strategies in Semiconductor Manufac-
turing under Uncertain Demand and Price Scenarios.” International Journal of Production Economics 105: 591–606.
Christie, R., and S. Wu. 2002. “Semiconductor Capacity Planning: Stochastic Modeling and Computational Studies.” IIE Transactions
34 (2): 131–144.
Clark, A., and H. Scarf. 1960. “Optimal Policies for a Multi-Echelon Inventory Problem.” Management Science 6: 475–490.
Connors, D., G. Feigin, and D. Yao. 1996. “A Queueing Network Model for Semiconductor Manufacturing.” IEEE Transactions on
Semiconductor Manufacturing 9 (3): 412–427.
16 R. Uzsoy et al.
Dayhoff, J., and R. Atherton. 1986. “Signature Analysis of Dispatch Schemes in Wafer Fabrication.” IEEE Transactions on Compo-
nents, Hybrids and Manufacturing Technology 9 (4): 518–525.
Diks, E., A. de Kok, and A. Lagodimos. 1996. “Multi-Echelon Systems: A Service Measure Perspective.” European Journal of Oper-
ational Research 95 (2): 241–263.
Duenyas, I., and C. Tsai. 2000. “Control of a Manufacturing System with Random Product Yield and Downward Substitutability.”
IIE Transactions 32: 785–795.
Elias, R., D. Montgomery, S. Low, and M. Kulahci. 2008. “Demand Signal Modelling: A Short-Range Panel Forecasting Algorithm
for Semiconductor Firm Device-Level Demand.” European Journal of Industrial Engineering 2: 253–278.
Eruguz, A., E. Sahin, Z. Jemai, and Y. Dallery. 2016. “A Comprehensive Survey of Guaranteed-Service Models for Multi-Echelon
Inventory Optimization.” International Journal of Production Economics 172: 110–125.
Federgruen, A., and P. Zipkin. 1984. “Computational Issues in an Infinite-Horizon Multiechelon Inventory Model.” Operations
Research 32 (4): 818–836.
Flamm, K. 1993. “Measurement of DRAM Prices: Technology and Market Structure.” Chap. 5 in Price Measurements and their
Uses, edited by M. Foss, M. Manser and A. Young, 157–206. Chicago, IL: University of Chicago Press.
Fowler, J., S. Brown, H. Gold, and A. Schoemig. 1997. “Measurable Improvements in Cycle-Time Constrained Capacity.” Proceed-
ings of the 6th International Symposium on Semiconductor Manufacturing, San Francisco, A21–A24.
Fowler, J., S. Park, G. MacKulak, and D. Shunk. 2001. “Efficient Cycle Time-Throughput Curve Generation Using a Fixed Sample
Size Procedure.” International Journal of Production Research 39 (12): 2595–2613.
Fowler, J., L. Mönch, and T. Ponsignon. 2015. “Discrete-Event Simulation for Semiconductor Wafer Fabrication Facilities: A Tuto-
rial.” International Journal of Industrial Engineering: Theory, Applications and Practice 22 (5): 661–682.
Gallego, G., K. Katircioglu, and B. Ramachandran. 2006. “Semiconductor Inventory Management with Multiple Grade Parts and
Downgrading.” Production Planning & Control 17 (7): 689–700.
Geng, N., and Z. Jiang. 2009. “A Review on Strategic Capacity Planning for the Semiconductor Manufacturing Industry.” Interna-
tional Journal of Production Research 47 (13): 3639–3655.
Geng, N., Z. Jiang, and F. Chen. 2009. “Stochastic Programming Based Capacity Planning for Semiconductor Wafer Fab with Uncer-
tain Demand and Capacity.” European Journal of Operational Research 198 (3): 899–908.
Gerchak, Y., A. Tripathy, and K. Wang. 1996. “Co-Production Models with Random Functionality Yields.” IIE Transactions 28: 391–
403.
Goldratt, E., and R. E. Fox. 1986. The Race. Croton-on-Hudson, NY: North River Press.
Graves, S., and S. Willems. 2000. “Optimizing Strategic Safety Stock Placement in Supply Chains.” Manufacturing & Service Opera-
tions Management 2: 68–83.
Graves, S., H. Meal, S. Dasu, and Y. Qui. 1986. “Two-Stage Production Planning in a Dynamic Environment.” Chap. 2 in Multi-
Stage Inventory Planning and Control, edited by M. Beckmann, W. Krelle and Sven Axsäter, 9–43. Berlin: Springer.
Graves, S., D. Kletter, and W. Hetzel. 1998. “A Dynamic Model for Requirements Planning with Application to Supply Chain Opti-
mization.” Operations Research 46: S35–S49.
Gross, C., and J. Sohl. 1990. “Disaggregation Methods to Expedite Product Line Forecasting.” Journal of Forecasting 9: 233–254.
Grove, A. 1983. High-Output Management. New York: Random House.
Güllü, R. 1996. “On the Value of Information in Dynamic Production/Inventory Problems under Forecast Evolution.” Naval Research
Logistics 43: 289–303.
Habla, C., and L. Mönch. 2008. “Solving Volume and Capacity Planning Problems in Semiconductor Manufacturing: A Computa-
tional Study.” Proceedings of the 2008 Winter Simulation Conference, Miami, 2260–2266.
Habla, C., R. Drießel, L. Mönch, T. Ponsignon, and H. Ehm. 2007. “A Short-Term Forecast Method for Demand Quantities in Semi-
conductor Manufacturing.” Proceedings of the IEEE Conference on Automation Science and Engineering, Scottsdale, 94–99.
Harrison, J. M., and M. J. López. 1999. “Heavy Traffic Resource Pooling in Parallel-Server Systems.” Queueing Systems 33: 339–
368.
Hax, A., and D. Candea. 1984. Production and Inventory Management. Englewood Cliffs, NJ: Prentice-Hall.
Heath, D., and P. Jackson. 1994. “Modeling the Evolution of Demand Forecasts with Application to Safety Stock Analysis in Produc-
tion/Distribution Systems.” IIE Transactions 26: 17–30.
Higle, J., and K. Kempf. 2011. “Production Planning under Supply and Demand Uncertainty: A Stochastic Programming Approach.”
Chap. 14 in Stochastic Programming: The State of the Art, edited by G. Infanger, 297–315. Berlin: Springer.
Hood, S., S. Bermon, and F. Barahona. 2003. “Capacity Planning under Demand Uncertainty for Semiconductor Manufacturing.”
IEEE Transactions on Semiconductor Manufacturing 16 (2): 273–280.
Hopp, W., M. Spearman, S. Chayet, K. Donohue, and E. Gel. 2002. “Using an Optimized Queueing Network Model to Support
Wafer Fab Design.” IIE Transactions 34 (2): 119–130.
Hsu, A., and Y. Bassok. 1999. “Random Yield and Random Demand in a Production System with Downward Substitution.” Opera-
tions Research 47: 277–290.
Huang, H., and H. Song. 2010. “Modified Base-Stock Policies for Semiconductor Production System with Dependent Yield Rates.”
European Journal of Operational Research 207: 206–217.
Huh, W., and R. Roundy. 2005. “A Continuous-Time Strategic Capacity Planning Model.” Naval Research Logistics 52 (4): 329–343.
International Journal of Production Research 17
Huh, W., R. Roundy, and M. Çakanyildirim. 2006. “A General Strategic Capacity Planning Model under Demand Uncertainty.” Naval
Research Logistics 53 (2): 137–150.
Humair, S., and S. Willems. 2011. “Technical Note – Optimizing Strategic Safety Stock Placement in General Acyclic Networks.”
Operations Research 59: 781–787.
Hung, Y., and C. Chang. 1999. “Determining Safety Stocks for Production Planning in Uncertain Manufacturing.” International Jour-
nal of Production Economics 58: 199–208.
Hung, Y., and Q. Wang. 1997. “A New Formulation Technique for Alternative Material Planning – An Approach for Semiconductor
Bin Allocation Planning.” Computers & Industrial Engineering 32 (2): 281–297.
Iida, T., and P. Zipkin. 2006. “Approximate Solutions of a Dynamic Forecast-Inventory Model.” Manufacturing & Service Operations
Management 8: 407–425.
Inderfurth, K. 1994. “Safety Stocks in Multistage Divergent Inventory Systems: A Survey.” International Journal of Production Eco-
nomics 35: 321–329.
Katircioglu, K., and G. Gallego. 2011. “A Practical Multi-Echelon Inventory Model with Semiconductor Manufacturing Application.”
Chap. 6 in Planning Production and Inventories in the Extended Enterprise: A State of the Art Handbook, Vol. 2, edited by K.
G. Kempf, P. Keskinocak and R. Uzsoy, 133–151. New York: Springer.
Kilger, C., and M. Wagner. 2015. “Demand Planning.” Chap. 7 in Supply Chain Management and Advanced Planning – Concepts,
Models and Software. 5th ed., edited by H. Stadtler, C. Kilger and H. Meyr, 125–154. Berlin Heidelberg: Springer.
Kim, S., and R. Uzsoy. 2008. “Exact and Approximate Algorithms for Capacity Expansion Problems with Congestion.” IIE Transac-
tions on Scheduling and Logistics 40 (12): 1185–1197.
Kim, S., and R. Uzsoy. 2009. “Heuristics for Capacity Planning Problems with Congestion.” Computers & Operations Research 36:
1924–1934.
Koenig, L., and A. Law. 1985. “A Procedure for Selecting a Subset of Size M Containing the L Best of K Independent Normal Popu-
lations, with Applications to Simulation.” Communications in Statistics: Simulation and Computation 14 (3): 719–734.
Köksal, G., İ. Batmaz, and M. Testik. 2011. “A Review of Data Mining Applications for Quality Improvement in Manufacturing
Industry.” Expert Systems with Applications 38: 13448–13467.
Leachman, R. 2001. “Semiconductor Production Planning.” Chap. 18.6. in Handbook of Applied Optimization, edited by P. Pardalos
and M. Resende, 746–762. New York: Oxford University Press.
Leachman, R., and T. Carmon. 1992. “On Capacity Modeling for Production Planning with Alternative Machine Types.” IIE Transac-
tions 24 (4): 62–72.
Leachman, R. C., and S. Ding. 2007. “Integration of Speed Economics into Decision-Making for Manufacturing Management.” Inter-
national Journal of Production Economics 107: 39–55.
Leachman, R., R. Benson, C. Liu, and D. Raar. 1996. “IMPReSS: An Automated Production-Planning and Delivery-Quotation Sys-
tem at Harris Corporation – Semiconductor Sector.” Interfaces 26 (1): 6–37.
Lee, J. 2015. “Smart Factory Systems.” Informatik Spektrum 38 (3): 230–235.
Levi, R., M. Pál, R. Roundy, and D. Shmoys. 2007. “Approximation Algorithms for Stochastic Inventory Control Models.” Mathe-
matics of Operations Research 32: 284–302.
Levi, R., R. Roundy, D. Shmoys, and V. Truong. 2008. “Approximation Algorithms for Capacitated Stochastic Inventory Control
Models.” Operations Research 56: 1184–1199.
Liker, J. 2004. The Toyota Way: 14 Management Principles from the World’s Greatest Manufacturer. New York: McGraw-Hill.
Lin, J., T. Chen, and H. Chu. 2014. “A Stochastic Dynamic Programming Approach for Multi-Site Capacity Planning in TFT-LCD
Manufacturing under Demand Uncertainty.” International Journal of Production Economics 148: 21–36.
Liu, J., F. Yang, H. Wan, and J. Fowler. 2011. “Capacity Planning through Queueing Analysis and Simulation-Based Statistical Meth-
ods: A Case Study for Semiconductor Wafer Fabs.” International Journal of Production Research 49 (15): 4573–4591.
Liu, N., S. Ren, T. Choi, C. Hui, and S. Ng. 2013. “Sales Forecasting for Fashion Retailing Service Industry: A Review.” Mathemati-
cal Problems in Engineering, Article ID 738675.
Magnanti, T., Z.-J. Max Shen, J. Shu, D. Simchi-Levi, and C.-P. Teo. 2006. “Inventory Placement in Acyclic Supply Chain Net-
works.” Operations Research Letters 34: 228–238.
Manary, M., and S. Willems. 2008. “Setting Safety Stock Targets at Intel in the Presence of Forecast Bias.” Interfaces 38: 112–122.
Meade, N., and T. Islam. 2006. “Modelling and Forecasting the Diffusion of Innovation – A 25-Year Review.” International Journal
of Forecasting 22: 519–545.
Meixell, M., and S. Wu. 2001. “Scenario Analysis of Demand in a Technology Market Using Leading Indicators.” IEEE Transactions
on Semiconductor Manufacturing 14 (1): 65–75.
Meyr, H., M. Wagner, and J. Rohde. 2015. “Structure of Advanced Planning Systems.” Chap. 5 in Supply Chain Management and
Advanced Planning – Concepts, Models and Software. 5th ed, edited by H. Stadtler, C. Kilger and H. Meyr, 99–106. Berlin
Heidelberg: Springer.
Mönch, L., J. Fowler, and S. Mason. 2013. Production Planning and Control for Semiconductor Wafer Fabrication Facilities: Model-
ing, Analysis, and Systems. New York: Springer.
Mönch, L., R. Uzsoy, and J. W. Fowler. Forthcoming-a. “A Survey of Semiconductor Supply Chain Models Part I: Semiconductor
Supply Chains, Strategic Network Design, and Supply Chain Simulation.” International Journal of Production Research
18 R. Uzsoy et al.
Mönch, L., R. Uzsoy, and J. W. Fowler. Forthcoming-b. “A Survey of Semiconductor Supply Chain Models Part III: Master Planning,
Production Planning, and Demand Fulfillment.” International Journal of Production Research.
Nakajima, S. 1988. Introduction to TPM: Total Productive Maintenance. Portland, OR: Productivity Press.
Nasr, W. W., B. Maddah, and M. K. Salameh. 2013. “EOQ with a Correlated Binomial Supply.” International Journal of Production
Economics 144: 248–255.
Norouzi, A., and R. Uzsoy. 2014. “Modeling the Evolution of Dependency between Demands, with Application to Inventory Plan-
ning.” IIE Transactions 46: 55–66.
Occhino, T. 2000. “Capacity Planning Model: The Important Inputs, Formulas and Benefits.” Proceedings of the 2000 IEEE/SEMI
Advanced Semiconductor Manufacturing Conference, Boston, 455–458.
Pai, P., S. Yang, and P. Chang. 2009. “Forecasting Output of Integrated Circuit Industry by Support Vector Regression Models with
Marriage Honey-Bees Optimization Algorithms.” Expert Systems with Applications 36: 10746–10751.
Peres, R., E. Muller, and V. Mahajan. 2010. “Innovation Diffusion and New Product Growth Models: A Critical Review and Research
Directions.” International Journal of Research in Marketing 27: 91–106.
Pitts, K. 1988. “Discrete-Event Simulation of Wafer Fabrication Facility.” Proceedings of the 1988 Winter Simulation Conference,
San Diego, 712–718.
Romauch, M., and A. Klemmt. 2015. “Product Mix Optimization for a Semiconductor Fab: Modeling Approaches and Decomposition
Techniques.” Computers & Operations Research 53: 338–352.
Rosling, K. 1989. “Optimal Inventory Policies for Assembly Systems under Random Demands.” Operations Research 37 (4): 565–
579.
Roundy, R. O. 2001. Report on Practices Related to Demand Forecasting for Semiconductor Products. Survey of Semiconductor
Research Council (SRC) Member Companies. Accessed February 8, 2017. https://ecommons.cornell.edu/bitstream/handle/1813/
9174/TR001294.pdf?sequence=1
Sethi, S., H. Yan, and H. Zhang. 2005. Inventory and Supply Chain Management with Forecast Updates. New York: Springer.
Shanthikumar, J., S. Ding, and M. Zhang. 2007. “Queueing Theory for Semiconductor Manufacturing Systems: A Survey and Open
Problems.” IEEE Transactions on Automation Science and Engineering 4 (4): 513–522.
Simpson, K. 1958. “In-Process Inventories.” Operations Research 6: 863–873.
Snyder, L., and Z. Shen. 2011. Fundamentals of Supply Chain Theory. New York: Wiley.
Swaminathan, J. 2000. “Tool Capacity Planning for Semiconductor Fabrication Facilities under Demand Uncertainty.” European Jour-
nal of Operational Research 120 (3): 545–558.
Swaminathan, J. 2002. “Tool Procurement Planning for Wafer Fabrication Facilities: A Scenario-Based Approach.” IIE Transactions
34 (2): 145–155.
Syntetos, A., Z. Babai, J. Boylan, S. Kolassa, and K. Nikolopoulos. 2016. “Supply Chain Forecasting: Theory, Practice, Their Gap
and the Future.” European Journal of Operational Research 252: 1–26.
Thomassey, S. 2010. “Sales Forecasts in Clothing Industry: The Key Success Factor of the Supply Chain Management.” International
Journal of Production Economics 128: 470–483.
Tian, F., S. Willems, and K. Kempf. 2011. “An Iterative Approach to Item-Level Tactical Production and Inventory Planning.” Inter-
national Journal of Production Economics 133 (1): 439–450.
Toktay, L., and L. Wein. 2001. “Analysis of a Forecasting-Production-Inventory System with Stationary Demand.” Management
Science 47 (9): 1268–1281.
Tu, Y. M., Y. H. Chao, S. H. Chang, and H. C. You. 2005. “Model to Determine the Backup Capacity of a Wafer Foundry.” Interna-
tional Journal of Production Research 43 (2): 339–359.
Vollmann, T. E., W. L. Berry, D. C. Whybark, and F. R. Jacobs. 2005. Manufacturing Planning and Control for Supply Chain Man-
agement. New York: McGraw-Hill.
Wang, K., and S. Lin. 2002. “Capacity Expansion and Allocation for a Semiconductor Testing Facility under Constrained Budget.”
Production Planning & Control 13 (5): 429–437.
Wang, C., and S. Su. 2015. “Strategic Capacity Planning for Light Emitting Diode (LED) Supply Chains across Taiwan and China.”
Journal of the Operational Research Society 66 (12): 1989–2003.
Wieland, B., P. Mastrantonio, S. Willems, and K. Kempf. 2012. “Optimizing Inventory Levels within Intel’s Channel Supply Demand
Operations.” Interfaces 42 (6): 517–527.
Witte, J. 1996. “ Using Static Capacity Modeling Techniques.” Proceedings of the 1996 IEEE/SEMI Advanced Semiconductor Manu-
facturing Conference Workshop, Cambridge, MA, 31–35.
Wright, C. M., and A. Mehrez. 1998. “An Overview of Representative Research of the Relationships between Quality and Inventory.”
Omega 26 (1): 29–47.
Wu, S., M. Erkoc, and S. Karabuk. 2005. “Managing Capacity in the High-Tech Industry: A Review of Literature.” The Engineering
Economist 50 (2): 125–158.
Wu, S., B. Aytac, R. Berger, and C. Armbruster. 2006. “Managing Short Life-Cycle Technology Products for Agere Systems.” Inter-
faces 36 (3): 234–247.
Wu, S., K. Kempf, M. Atan, B. Aytac, S. Shirodkar, and A. Mishra. 2010. “Improving New-Product Forecasting at Intel Corpora-
tion.” Interfaces 40 (5): 385–396.
International Journal of Production Research 19
Yano, C., and H. Lee. 1995. “Lot Sizing with Random Yields: A Review.” Operations Research 43: 311–334.
You, F., and I. Grossmann. 2011. “Stochastic Inventory Management for Tactical Process Planning under Uncertainties: MINLP Mod-
els and Algorithms.” AIChE Journal 57: 1250–1277.
Zhang, F. 2007. “An Application of Vector GARCH Model in Semiconductor Demand Planning.” European Journal of Operational
Research 181: 288–297.
Zhang, F., R. Roundy, M. Cakanyildirim, and W. Huh. 2004. “Optimal Capacity Expansion for Multi-Product, Multi-Machine Manu-
facturing Systems with Stochastic Demand.” IIE Transactions 36 (1): 23–36.
Zhang, Z., M. Zhang, S. Niu, and L. Zheng. 2006. “Capacity Planning with Reconfigurable Kits in Semiconductor Test Manufactur-
ing.” International Journal of Production Research 44 (13): 2625–2644.
Zhang, M., S. Niu, S. Deng, Z. Zhang, Q. Li, and L. Zheng. 2007. “Hierarchical Capacity Planning with Reconfigurable Kits in Glo-
bal Semiconductor Assembly and Test Manufacturing.” IEEE Transactions on Automation Science and Engineering 4 (4):
543–552.
Zhang, M. T., J. Fu, and E. Zhu. 2008. “Dynamic Capacity Modeling in Semiconductor Assembly Manufacturing.” International
Journal of Production Research 46 (3): 739–752.
Zheng, M., and K. Wu. 2017. “Smart Spare Parts Management Systems in Semiconductor Manufacturing.” Industrial Management &
Data Systems 117 (4): 754–763.
Zhou, S., P. Jackson, R. Roundy, and R. Zhang. 2007. “The Evolution of Family Level Sales Forecasts into Product Level Forecasts:
Modeling and Estimation.” IIE Transactions 39: 831–843.
Zipkin, P. H. 2000. Foundations of Inventory Management. Singapore: McGraw-Hill.