View metadata, citation and similar papers at core.ac.

uk brought to you by CORE

provided by Universidade do Minho: RepositoriUM

Warehouse Design and Planning:

A Mathematical Programming Approach

Carla A.S. Geraldes1 , Maria Sameiro Carvalho2 , and Guilherme A.B. Pereira2
1
Centro ALGORITMI and Polytechnic Institute of Bragança, Portugal
carlag@ipb.pt
2
Centro ALGORITMI and University of Minho, Portugal
sameiro@dps.uminho.pt, guilherme.pereira@algoritmi.uminho.pt

Abstract. The dynamic nature of today’s competitive markets compels

organizations to an incessant reassessment in an effort to respond to con-
tinuous challenges. Therefore, warehouses as an important link in most
supply chains, must be continually re-evaluated to ensure that they are
consistent with both market’s demands and management’s strategies. A
number of warehouse decision support models have been proposed in
the literature but considerable difficulties in applying these models still
remain, due to the large amount of information to be processed and
to the large number of possible alternatives. In this paper we discuss a
mathematical programming model aiming to support some warehouse
management and inventory decisions. In particular a large mixed-integer
nonlinear programming model (MINLP) is presented to capture the trade-
offs among the different inventory and warehouse costs in order to achieve
global optimal design satisfying throughput requirements.

Keywords: Supply chain management, Warehouse models, Inventory

management, Mathematical modelling.

1 Introduction
Market competition requires continuous improvement in the design and opera-
tion of supply chains. A supply chain can be considered as a network of entities
whose efficiency and effectiveness is highly determined by the performance of
the overall network (see Fig. 1).
In a supply chain network, products need to be physically moved from one
location to another. During this process, they may be buffered or stored at
certain facilities (warehouses) for a certain period of time for strategic or tactical
reasons. Within this context, warehouses play an important role in supply chains
and are a key aspect in a very demanding, competitive and uncertain market.
On the other hand, modern supply chain principles compel companies to
reduce or eliminate inventory levels. Additionally warehouses require capital,
labour, and information technologies, which are expensive resources. So, why do
we still need warehouses?

B. Murgante et al. (Eds.): ICCSA 2012, Part III, LNCS 7335, pp. 187–201, 2012.

c Springer-Verlag Berlin Heidelberg 2012
188 C.A.S. Geraldes, M.S. Carvalho, and G.A.B. Pereira

Information Flow

Suppliers Customers
Distribution
centres
Plants

Material Flow

Fig. 1. A typical supply chain network

According to Bartholdi and Hackman [1] there are four main reasons why
warehouses are useful:
1. To consolidate products in order to reduce transportation costs and provide
customer service;
2. To take advantage of economies of scale;
3. To provide value-added processing services, and
4. To reduce response time.
Thus, warehouses will continue to be an important node at the logistic network
by the fact that if a warehouse cannot process the orders quickly, effectively,
and accurately, then all the supply chain optimization efforts will suffer (see
Tompkins [12]).
In distribution logistic where market competition requires higher performances
from the warehouses, companies are compelled to continuously improve the de-
sign and planning of warehouse operations. Furthermore, the ever increasing
variety of products, the constant changes in customer demands and the adop-
tion of agile management philosophies also bring new challenges to reach flex-
ible structures that provide quality, efficiency and effectiveness to the logistics
operations.
Some major decisions involved in the warehouse design and operation prob-
lems are illustrated in Fig. 2 (see Gu et al. [4]). Warehouse design and planning
decisions typically run from a functional description, through a technical spec-
ification, to equipment selection and determination of the layout. The overall
structure decision determines the material flow patterns within the warehouse,
the specification of the functional areas and the flows between the areas. Siz-
ing and dimensioning decisions determine the total size of the warehouse as
well as the space allocation among functional areas. Layout definition is the
detailed configuration within a functional area and equipment decisions define
 Warehouse Design and Planning 189

Performance Evaluation

Overall Receiving
Structure

Storage
Sizing and
Layout
Dimensioning

Order Picking

Equipment Operation
Selection Strategy Shipping

Warehouse Design Warehouse Operation

Fig. 2. A framework for design and operation problems (adapted from Gu et al. [4])

an automation level for the warehouse and identify equipment types. Finally
operating policies refer to storage, picking and routing decisions.
Hassan [7] presented a framework for the design of a warehouse. The proposed
framework accounts for several factors and operations of warehousing such as:
1. Specification of warehouse type and purpose;
2. Analysis and forecasting of the demand;
3. Definition of operating policies;
4. Establishment of inventory levels;
5. Class formation;
6. Definition of functional areas and general layout;
7. Storage partition;
8. Selection of equipment for handling and storage;
9. Design of aisles;
10. Determination of space requirements;
11. Location and number of Input/Output points;
12. Location and number of docks;
13. Arrangement of storage;
14. Zone formation.
Once these warehouse decisions are strongly interrelated, warehouse design is a
highly complex task where conflicting objectives impose specific trade-offs.
Despite the various decision models available in scientific literature, the ma-
jority addresses isolated or simplified problems in order to provide the best
solution. However, most of the real problems are unfortunately not well-defined
and often cannot be reduced to multiple isolated sub-problems. Therefore, ware-
house design often requires a mixture of analytical skills and creativity. Anyhow,
research aiming an integration of various decisions models and methods is badly
190 C.A.S. Geraldes, M.S. Carvalho, and G.A.B. Pereira

needed in order to develop a methodology for systematic warehouse design (see

Rouwenhorst et al. [10]).
Furthermore, as the literature review in next section shows, most research
efforts have been dedicated to warehouse operations decisions instead of design
decisions. This is not surprising since design decisions models are more difficult
to develop and treat analytically once they require the integration of several and
complex issues.
In this paper we present a warehouse and inventory mathematical model that
jointly integrates issues concerning:
– The size of the warehouse;
– The external storage additional capacity, if needed;
– The replenishment quantities and reorder points of products to be stored.
Our aim is to test an integrated approach that takes into account inventory and
some warehouse design decisions.
In the Section 2 of this paper we will present a brief literature review on ware-
house design and planning issues. The purpose of this section is not limited to the
specific studied problem but also covers other important topics in warehouse lit-
erature. The Section 3 will present the proposed warehouse sizing and inventory
model formulation and a description of the methodology used to solve the model.
Computational results will be presented and summarized in Section 4, and finally
some conclusions and future work directions are reported in Section 5.

2 Literature Review
Warehousing is concerned with all the material handling activities that take place
within a warehouse. It includes the receiving of products, storage, order-picking,
accumulation, sorting and shipping operations. Basically, one can distinguish
two types of warehouses: distribution warehouses and production warehouses.
According to Van den Berg and Zjim [14], a distribution warehouse is a ware-
house in which products from different suppliers are collected (and sometimes
assembled) for delivery to a number of customers. On the other hand, a produc-
tion warehouse is used for the storage of raw materials, semi-finished products
and finished products in a production facility.
There are many activities that occur at a warehouse. Typically, distribution
warehouses receive products - Stock Keeping Units (SKUs) - from suppliers,
unload products from the transport carrier; store products, receive orders from
customers, assemble orders, repackage SKUs and ship them to their final destina-
tion. Frequently, products arrive packaged on large scale units and are packaged
and shipped on small units. For example, SKUs may arrive in full pallets but
must be shipped in cases. Fig. 3 shows the typical functional areas and flows
within warehouses.
At the receiving area products (or items) are unloaded and inspected to ver-
ify any quantity and quality inconsistency. Afterwards, items are transferred
 Warehouse Design and Planning 191

Receiving

Putaway

Putaway
Reserve Storage
and
Pallet Picking

Replenishment

Cross-docking
Replenishment

Case Picking
Broken Case
Picking

Accumulation, Sortation and

Packing

Shipping

Fig. 3. Typical warehouse functions and flows

to a storage zone or are directly placed to the shipping area (this is called a
cross-docking operation). We can distinguish two types of storage areas: reserve
storage area and forward or picking area. The reserve area is the place where
the products stay until they are required by costumers’ orders. The picking area
is a relatively small area, typically used to store fast moving products. Most of
the flows between these areas are the result of replenishment processes. Order
picking is one of the most important functions in most warehouses. SKUs are
retrieved from their storage positions based on customers’ orders and moved to
the accumulation and sorting area or directly to the shipment area. The picked
units are then grouped by customers order, packaged and stacked on the right
unit load and transferred to the shipping area.

2.1 Warehouse Design and Planning

Warehouse design can be defined as a structured approach of decision making
at distinct decision levels in an attempt to meet a number of well-defined per-
formance criteria. At each level, multiple decisions are interrelated and therefore
it is necessary to cluster relevant problems that are to be solved simultaneously.
According to Rouwenhorst et al. [10] a warehouse design problem is a “coherent
cluster of decisions” and they define decisions to be coherent when a sequential
optimization does not guarantee a globally optimal solution.
The design of a warehouse is a highly complex problem. It includes a large num-
ber of interrelated decisions involving warehouse processes, warehouse resources
and the organization of the warehouse (see Heragu [8]). Rouwenhorst et al. [10]
192 C.A.S. Geraldes, M.S. Carvalho, and G.A.B. Pereira

classify the management decisions concerning warehousing into strategic decisions,

tactical decisions and operational decisions. Strategic decisions are long term deci-
sions and always mean high investments. The two main issues are concerned with
the design of the process flow and with the selection of the types of the warehous-
ing systems. Tactical management decisions are medium term decisions based on
the outcomes of the strategic decisions. The tactical decisions have a lower impact
than the strategic decisions, but still require some investments and should there-
fore not be reconsidered too often. At the operational level, processes have to be
carried out within the constraints set by the strategic and tactical decisions made
at the higher levels. In this level, the concern includes the operational policies such
as storage policies and picking and routing operations.
After determining warehouse location and its size, layout decisions must in-
clude areas definition and what size should be allocated to each functional area.
The forward-reserve problem (FRP) is the problem of assigning products to the
functional areas. In this problem the critical decision concerns the choice of prod-
ucts that will be stored in the forward area. Van den Berg et al. [13] proposed
a binary programming model to solve the FRP in the case of unit load replen-
ishment, and presented efficient heuristics that provide tight performances guar-
anties. These replenishments can occur during busy or idle picking periods. The
objective was to minimize the number of urgent or concurrent replenishments of
the forward area during the busy periods. Although addressing this problem is a
strategic decision problem, it is strongly associated upon some tactical problems
such as how the items will be distributed among the functional areas. However,
the approach usually adopted is to solve the problems sequentially by generating
multiples alternatives for the functional area size problem and then determine
how the products can be allocated for each of the alternatives.
Gray et al. [5] developed an integrated approach for the design and operation
of a typical order-consolidation warehouse. This approach included warehouse
layout, equipment and technology selection, item location, zoning, picker rout-
ing, pick generation list and order batching. Due to the complexity of the over-
all problem, they developed a multi-stage hierarchical decision approach. This
hierarchical approach used a sequence of coordinated mathematical models to
evaluate the major economic trade-offs and to reduce the decision space to a few
number of alternatives. They also used simulation technique for validation and
fine tuning of the resulting design and operating policies.
Heragu et al. [8] developed a mathematical model and a heuristic algorithm
that jointly determines the size of the functional areas and the allocation of the
product in a way that minimizes the total material handling and storage costs.
The proposed model uses real data readily available to a warehouse manager
and considers realistic constraints.
Geraldes et al. [2] adapted the mixed-integer programming model proposed
by Heragu et al. [8] to tackle the storage allocation and assignment problems
during the redesign process of a Portuguese company warehouse.
 Warehouse Design and Planning 193

More recently Strack and Pochet [11] presented a robust approach that inte-
grates aspects such us: (i) the size of the functional areas; (ii) the assignment
and allocation of products to storage locations in the warehouse; (iii) the re-
plenishment decision in the inventory management. This is probably the most
integrated decision model found in this area, nevertheless still assumes fixed and
known capacity for the warehouse.

2.2 Inventory Decisions

The adoption of new management philosophies compels companies to eliminate

or reduce inventory levels. In addition to warehouse management decisions, an
appropriate inventory policy may result in a reduction of the total warehousing
costs and can also improve the efficiency of the operating policies within the
warehouse. The aim of inventory management is to minimize total operating
costs satisfying customer service requirements (see Ghiani et al. [3]). To accom-
plish this, an optimal ordering policy must answer questions, for each SKU, such
us when to order and how much to order.
Two different inventory policies arise (see Hadley and Whitin [6]): continuous
review policy and the periodic review policy. The first policy implies that the
stock level will be monitored continuously. Whenever the inventory on hand
decreases to a predetermined level, referred to as the reorder point, a new order
is placed to replenish the inventory level. The placed order is a fixed quantity
that minimizes the total inventory costs and is normally called the economic
order quantity. In the second policy, the inventory level is checked at specific
fixed time periods and an order decision is made to complete the stock to a
desired upper limit. In this system the inventory level is not monitored at all
during the time interval between orders, so it has the advantage of little or no
required record keeping. The disadvantage is less direct control. Such system
also requires that a new order quantity must be determined each time a periodic
order is made. The operating costs taken into account in both inventory policies
are the acquisition cost, the holding cost and the shortage cost.
These basic policies can be adapted to take into account special situations such
us single or multi-item models with or without a constraint on the total storage
space, deterministic or stochastic demands, lost sales, etc. For more details and
examples see Ghiani et al. [3] and Nahmias [9].

3 Mathematical Programming Model

Storage may be considered one of the major warehouse functions. Some funda-
mental decisions occur during the design and planning of a warehouse: (i) how
much inventory should be kept in the warehouse for each SKU; (ii) how fre-
quently and at what time should the inventory for a SKU be replenished, and
(iii) what should be the size of the warehouse.
The first two decisions lead to the traditional inventory management problem
and the last one is probably one of the most important aspect in designing a
194 C.A.S. Geraldes, M.S. Carvalho, and G.A.B. Pereira

warehouse. Once warehouse size is determined, it will act as a constraint that

may last for a long period of time.
Generally there are two basic choices for warehousing. The first is to operate
a owned (or private) warehouse and the second is to rent storage space from a
public warehouse. Depending on which is least expensive a warehouse manager
may use only one type or adopt a mixed strategy - both owned and rented
storage space.
In this section we present a mathematical programming model that integrates
the warehouse sizing problem and the inventory decisions. The mathematical
model determines the optimal overall size for a warehouse in a way that mini-
mizes the inventory and warehouse costs. In developing the model we also assume
that external storage space is available to rent to allow the selection of the most
adequate storage alternative: storage space ownership, rented warehouse space
or a combined solution. Furthermore, an inventory control policy will be based
on continuous review policy (reorder point system).
The problem can be stated as follow: a distribution company delivers a set
of i products to its customers. Customers demand is well defined and unit ac-
quisition, processing, storage and inventory carrying costs are known. Given
warehousing costs (both own storage space and rented space) the objective is to
determine the size and type (own or rented) of storage space required as well as
the quantity of each product to order and respective reorder point, for a level of
service and that minimizes total system costs.

3.1 A MINLP Model for the Warehouse Design Problem

The following notation, adapted from Strack and Pochet [11], is used:
Parameters:
i : Product number (i = 1, ..., I)
CostCar : Inventory carrying cost
CostAcqui : Acquisition cost of product i
CostShort : Shortage cost
CostRecp : Reception cost
CostCapaS : External capacity cost
CostCapaW : Capacity cost of the private warehouse
CostCapaF W : Private warehouse capacity fixed cost
E(Ui ) : Expected value of the demand of product i
L : Supply lead time
dLi : Demand of product i during L
L
μi : Average demand of product i during L
σiL : Standard deviation of demand of product i during L
M : Large positive number
 Warehouse Design and Planning 195

Decision variables:

1 if we have a private warehouse
w=
0 otherwise
CapaW : Capacity of the private warehouse
CapaS : External additional storage capacity
Qi : Replenishment quantity of product i
ri : Reorder point of product i

The general formulation of the model can be stated as:

I
    I
Qi L
minimize Costcar × + ri − µi + CAcqui × E(Ui )
i=1
2 i=1
I  ∞
E(Ui )
+ CostShort × × (dL L L
i − ri )f (di ) ddi (1)
i=1
Q i ri
I
 E(Ui )
+ CostRecp × + CostCapaS × CapaS
i=1
Qi
+ CapaW × w × (CostCapaW + CostCapaF W ) ,

subject to:

I

(Qi + ri − µL
i ) ≤ CapaW + CapaS , (2)
i=1
CapaW ≤ M w , (3)
CapaW ≥ w , (4)
Qi , ri ≥ 0 , (5)
CapaW, CapaS ≥ 0 , (6)
w ∈ {0, 1} . (7)

The objective function (1) is the expected warehouse and inventory costs per
period. Concerning the inventory costs we have taken into account: carrying
cost, acquisition cost and shortage cost. The warehouse costs are composed by
the reception cost, the additional external storage capacity cost and the costs
of the private warehouse. The integrity of the model is ensured by the capacity
constraint (2) that guaranties that the required storage capacity is met. Con-
straints (3)-(4) serve to include the costs of the private warehouse. Finally, a set
of variables must be nonnegative (5)-(6) and another is considered binary (7).
The above model considers inventory and warehouse size decisions since it in-
tegrates both issues supporting decision makers defining the best warehousing
alternative, taking into account space requirements determined by customers’
demand.
196 C.A.S. Geraldes, M.S. Carvalho, and G.A.B. Pereira

3.2 Methodology

The above model jointly integrates inventory decisions and size decisions (both
ownership and rented storage space). It is a mixed-integer nonlinear program-
ming model with a large number of variables when real cases are considered.
To evaluate the computational performance involved in solving the proposed
integrated model, experimental tests were performed using LINGO 12.0 solver
which uses branch-and-bound algorithm for integer nonlinear problems. All tests
were performed on an Intel Core 2Duo 1.4 GHz CPU and 3GB RAM.

4 Computational Results
Instances for different scenarios were randomly generated to assess the behaviour
of the model when the number of products increases (see Table 1). Table 2 shows
parameter values used to generate the testing problems.

Table 1. Analysed scenarios

Scenario I II III IV V
SKU [units] 10 100 500 1000 5000

Table 2. Parameter values for the numerical examples

Parameter Value

CostCar 3
CostShort 50
CostRecp 5
CostCapaS 20
CostCapaW 3
CostCapaF W 10
E(Ui ) Uniform [1, 50]
dL
i N (μL L
i , σi )

The computational results for the different testing cases are shown in Table 3.
As it can be seen it was possible to analytically solve to optimality all the test
scenarios in a very satisfactory computational time. In general the algorithm is
very efficient and converge to the optimal solution in a very short time. Never-
theless, the computational time of LINGO rises as the problem size increases.
 Warehouse Design and Planning 197

Table 3. Model computational results

Scenario
I II III IV V
Total variables 13 103 503 1003 5003
Nonlinear variablesa 12 102 502 1002 5002
Iterations 203 647 1762 4020 16021
CPU time [mm : ss] 00 : 03 00 : 11 00 : 41 01 : 20 14 : 01
Global Global Global Global Global
State
Opt. Opt. Opt. Opt. Opt.
a
Variables involved in the nonlinear relationships of the model.

In order to point out some more features of the proposed model, more exper-
iments were conducted for a warehouse problem with 500 SKUs (scenario III).
Again parameter values from Table 2 were used and service levels between 95%
and 99% were considered.
Table 4 shows that a private warehouse can be less expensive than renting all
the storage capacity, even considering higher amounts of stock. When the pri-
vately operated warehouse has an upper storage limit, the total optimal storage
capacity equals the optimal capacity for the strategy of using only a public ware-
house. This happens whenever the storage upper limit of the private warehouse
is less than the optimal storage solution of the public warehouse.

Table 4. Warehouse dimension and cost

w M CapaS CapaW Total Cost

0 — 694 — 51060
unlimited 0 842 45724
300 394 300 48960
500 194 500 47560
1
600 94 600 46860
700 0 700 46149
900 0 842 45724

Additional results also show that the inventory costs have a significant im-
pact in warehouse management (see Table 5). As expected the integration of an
appropriate inventory policy may result in a reduction of the total warehousing
costs.
The gathered results also indicate that a mixed strategy (both owned and
rented) would be better than renting all the storage space. Warehouse manage-
ment should then achieve high levels of utilization for owned storage space and
198 C.A.S. Geraldes, M.S. Carvalho, and G.A.B. Pereira

Table 5. Warehouse total costs

w M Inventory Costs Warehouse Costs Total Costs

0 — 23908 27152 51060
unlimited 23878 21846 45724
300 23908 25052 48960
500 23908 23652 47560
1
600 23908 22952 46860
700 23903 22246 46149
900 23878 21846 45724

use additional rented space on a short-term basis to meet peak space require-
ments.
As it can be seen in Table 6, different cost elements have different impacts in
warehouse total cost.

Table 6. Warehouse inventory and operational costs

w M CostCar CostAqu CostShort CostRecp

0 — 1056 21436 1416 13272

unlimited 1279 21436 1163 10900
300 1056 21436 1416 13272
500 1056 21436 1416 13272
1
600 1056 21436 1416 13272
700 1065 21436 1402 13146
900 1279 21436 1163 10900

Since numerical tests were performed without having the real value of the dif-
ferent parameters involved in the model, and in order to highlight some relevant
issues in the performance of the model, it is important to investigate how much
influence some fluctuations in different cost elements may have in the optimal
solution. For this purpose some of these costs will be selected and will vary (-10%
and 25%) one at a time.
Denoting x and x∗ as the initial parameter value and the new value, and
w(x) and w(x∗ ) as the size of the storage capacity we will measure the impact
on warehouse size of some variations using the following equation:
w(x) − w(x∗ )
× 100% . (8)
w(x∗ )
Table 7 summarizes the impact of cost fluctuations in the warehouse optimal
size. A detailed look at the different cost changes reveals that the reception cost
 Warehouse Design and Planning 199

(CostRecp) has a larger impact in warehouse size. As expected, an increase in

this cost forces the model to increase the optimal order quantities and conse-
quently the warehouse storage size increases. In a similar way decreasing the
reception cost causes a decrease of the storage size area. On the other hand, the
carrying cost (CostCar) and the shortage cost (CostShort) have lower impacts
in the optimal solution with changes in warehouse size less than 1.2%.

Table 7. Impact in warehouse size (%)

Parameter Variation CapaS CapaW

−10 −0.71 −0.59

CostCar
25 1.01 1.20
−10 0.58 0.48
CostShort
25 −0.99 −1.17
−10 4.83 4.73
CostRecp
25 −9.51 −9.55
−10 −4.67 —
CostCapaS
25 10.86 —
−10 — −3.55
CostCapaF W
25 — 8.08

The results also show that despite the inventory costs have a significant im-
pact in the total warehouse costs, also variations in the warehouse costs mean
significant impact in the storage size area. For example, the model is very sensi-
tive to the external capacity cost (CostCapaS) where an increase of 25% causes
a reduction of 10.86% of the rented area. In a similar way a 25% increase in the
cost of the owned warehouse (CostCapaF W ) causes an reduction of 8.08% of
the size storage area.
The fluctuations that were carried out in some of the parameters of the model
revealed that the size of the warehouse is more sensitive to changes of some
parameters than to others. Nonetheless, the warehouse size solution is quite
robust once the proposed model integrates the different costs in a way that
balances the trade-offs among them, minimizing the effect of the change.
Finally it should be noted that in our numerical results the parameter vari-
ations were done one at a time. Future research might also be performed to
investigate joint effects of different parameters on the quality of the solution of
the model.

5 Conclusions and Future Work

Most of the times, inventory decisions and warehouse design decisions are in-
dependently considered. In fact, a single decision model that integrates several
200 C.A.S. Geraldes, M.S. Carvalho, and G.A.B. Pereira

decisions concerning warehouse design and planning is very complex due to the
tremendous amount of information to be processed, to the large number of ex-
isting alternatives, to the existence of various and often conflicting objectives
and to the uncertainty inherent in the material flow into, through and out of the
warehouse.
Throughout this work our aim was to show the value of integrating inventory
decisions and warehouse design decisions, in particular the strategic decision that
concerns the size of a warehouse. For that purpose a mathematical programming
model was proposed and discussed.
The proposed mathematical programming model jointly integrates: (i) the size
of a private storage warehouse; (ii) the external additional storage capacity, if
needed; and (iii) the replenishment quantities and reorder points of the products.
Although the model is a mixed-integer nonlinear programming model with
a large number of variables, usually difficult to solve by general optimization
packages, it was possible to solve to optimality some test scenarios in a very sat-
isfactory computational time. Nevertheless for large instances the computational
time increases considerably.
Computational results were obtained for five different scenarios randomly gen-
erated. The gathered results also suggest that, for this particular situation, a
mixed strategy, both owned and rented storage capacity, would have lower to-
tal costs than renting all the storage space. High levels of utilization should be
achieved for owned storage space and additional rented space should be used to
meet peak space requirements. It has also been shown that inventory costs have
a significant impact in warehouse system costs.
A sensitivity analysis was performed to observe the impact of cost fluctuations
in the warehouse optimal size. For this purpose some costs were selected and
varied one at a time. The results show that the size of the warehouse is more
sensitive to changes of some parameters than to others.
Even though the presented model integrates two important decisions con-
cerning the design of a warehouse, many other decisions were not included. For
example the size of the functional areas inside the warehouse; the assignment
and allocation problem of the products; the picking and routing strategies, etc.
In summary, despite some advances in integrated approaches, further research
focusing integrated models where different processes in the warehouse are jointly
considered (and its corresponding dynamic nature), is still required. Given the
prevalence of warehouses in the supply chain networks we believe that such research
achievements can have a significant impact in the supply chain performance.

References
1. Bartholdi, J., Hackman, S.: Warehouse & Distribution Science (2006). Release 0.76,
http://www.warehouse-science.com
2. Geraldes, C., Carvalho, S., Pereira, G.: A warehouse design decision model. In: Pro-
ceedings of the International Engineering Management Conference 2008, IEMC-
Europe 2008, Estoril, Portugal (2008) ISBN: 978-1-4244-2289-0, IEEE Catalog
Number: CFP08EMS
 Warehouse Design and Planning 201

3. Ghiani, G., Laporte, G., Musmanno, R.: Introduction to Logistics Systems Plan-
ning and Control. John Wiley & Sons Ltd, England (2004)
4. Gu, J., Goetschalckx, M., McGinnis, L.F.: Research on warehouse operation: A
comprehensive review. European Journal of Operational Research 177, 1–21 (2007)
5. Gray, A., Karmarkar, U., Seidman, A.: Design and operation of an order-
consolidation warehouse: models and application. European Journal of Operational
Research 58, 14–36 (1992)
6. Hadley, G., Whitin, T.: Analysis of Inventory Systems. Prentice-Hall, Englewood
Cliffs (1963)
7. Hassan, M.: A framework for the design of warehouse layout. Facilities 20(13/14),
432–440 (2002)
8. Heragu, S., Du, L., Mantel, R.J., Schuur, P.C.: Mathematical model for warehouse
design and product allocation. International Journal of Production Research 43(2),
432–440 (2005)
9. Nahmias, S.: Production and Operation Analysis, 3rd edn. McGraw-Hill Interna-
tional Editions (1997)
10. Rouwenhorst, B., Reuter, B., Stokrahm, V., van Houtum, G., Mantel, R., Zijm,
W.: Warehouse design and control: Framework and literature review. European
Journal of Operational Research 122, 515–533 (2000)
11. Strack, G., Pochet, Y.: An integrated model for warehouse design and planning.
European Journal of Operational Research 204, 35–50 (2010)
12. Tompkins, J.A., White, J.A., Bozer, Y.A., Tanchoco, J.M.A.: Facilities Planning.
John Wiley & Sons, New York (2003)
13. Van den Berg, J., Sharp, G., Gademann, A., Pochet, Y.: Forward-reserve allocation
in a warehouse with unit-load replenishments. European Journal of Operational
Research 111, 98–113 (1998)
14. Van den Berg, J., Zjim, W.: Models for warehouse management: Classification and
examples. International Journal of Production Economics 59, 519–528 (1999)
15. Van den Berg, J.: A literature survey on planning and control of warehousing
systems. IEEE Transactions 31, 751–762 (1999)