This article was downloaded by: [Simon Fraser University]

On: 15 March 2015, At: 19:21

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

International Journal of Computer Integrated

Manufacturing
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/tcim20

Advanced sales and operations planning framework in

a company supply chain
a a a
Jun-Zhong Wang , Su-Tzu Hsieh & Ping-Yu Hsu
a
Department of Business Administration , National Central University , Jungli City , Taiwan
Published online: 02 Nov 2011.

To cite this article: Jun-Zhong Wang , Su-Tzu Hsieh & Ping-Yu Hsu (2012) Advanced sales and operations planning
framework in a company supply chain, International Journal of Computer Integrated Manufacturing, 25:3, 248-262, DOI:
10.1080/0951192X.2011.629683

To link to this article: http://dx.doi.org/10.1080/0951192X.2011.629683

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 International Journal of Computer Integrated Manufacturing
Vol. 25, No. 3, March 2012, 248–262

Advanced sales and operations planning framework in a company supply chain

Jun-Zhong Wang*, Su-Tzu Hsieh and Ping-Yu Hsu
Department of Business Administration, National Central University, Jungli City, Taiwan
(Received 29 September 2010; final version received 29 September 2011)

Traditional sales and operations planning (S&OP) focuses on balancing company supply and demand, and is utilised
utilised to align plans that support a business-strategic goal. An obvious drawback of previous decision models is
their lack in considering the supply chain network. This article proposes a new global S&OP planning framework to
integrate four supply chain stages of demand, purchasing, production and transportation with different planning
strategies. Model applicability is tested by a real life case. Several simulation scenarios are conducted to evaluate the
effectiveness and efficiency of the proposed model and heuristics. This study adopts the integrated method as a
decision support tool, thereby enhancing coordination between financial and physical activities.
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

Keywords: supply chain planning; sales and operations planning; heuristics

1. Introduction inventory level mechanism derived from the con-

To survive the current intensive, competitive environ- straints of demand and capacity. Makris and Chrysso-
ment, companies are forced to adjust their operation louris (2010) reported a Bayesian network uncertainty
plans more frequently and dramatically than ever evaluating customisation orders from the demand
before. Operation planning includes capacities of aspect. Makris et al. (2011) built a buyer uncertainty-
production sites, cost and production characteristics accessing model for high customised demands. Some
because a company in the global context must precisely researches focus on production and strategy of
calculate costs and benefits before implementation capacity modification (Olhager et al. 2001, Affonso
(Shapiro 2007). Given the demand for precise calcula- et al. 2008). Several scholars have focused on optimis-
tion of the cost and profit in supply chain manage- ing the production procedure (Feng et al. 2008), while
ment, sales and operations planning (S&OP) is gaining others have emphasised multiprocessor flow with
increasing recognition (Feng et al. 2008). From the financial constraints (Bertel et al. 2008). Various works
management perspective, sales and operations plan- have investigated order uncertainty of demand fulfil-
ning not only serves as the foundation of a production ment, such as Chen-Ritzo et al. (2010) to align supply
plan for the supply chain network, but also as the flexibility with order uncertainty, and Makris and
preliminary model for solving buyer-vendor or inven- Chryssolouris (2010) used a modern technology for
tory-distribution problems. high customisation order uncertainty. Several studies
S&OP is a pillar in supply chain planning because have stressed the order taking process in supply chain
of its ability to not only vertically integrate, but to also systems (Mourtzis et al. 2008, Makris et al. 2008).
support coordination between supply chain members Previous researches have attempted to solve S&OP
(Affonso et al. 2008). The S&OP requirement output of issues involving a single or partial perspective, while
the customer is the demand input for the downstream most studies describe how to support S&OP in a
vendor. A precise upstream customer S&OP plan can specific planning stage. Few studies have focused on
greatly help the supplier in time delivery. An accurate the S&OP problem in an integrated view with a
vendor S&OP promise can also raise customer comprehensive S&OP stage process that includes
satisfaction. This is why many companies are in demand forecast and pricing, purchasing cost of bill
demand as advanced system figures of S&OP. of material (BOM), production cost, inventory cost
Previous works have provided rich S&OP re- and transportation aspects. This article develops a
searches of single perspective demand forecast, pro- planning framework to integrate the planning aspects
duction planning, inventory level planning or cash flow of demand, purchasing, production planning and
management. Genin et al. (2005) provided a projected distribution.

*Corresponding author. Email: [email protected]

ISSN 0951-192X print/ISSN 1362-3052 online

Ó 2012 Taylor & Francis
http://dx.doi.org/10.1080/0951192X.2011.629683
http://www.tandfonline.com
 International Journal of Computer Integrated Manufacturing 249

The rest of the article is organised into four solved by a Lagrangian heuristic or a genetic
sections. Section 2 reviews and analyses the contribu- algorithm.
tions of previous research in the scope of S&OP. Goetschalckx et al. (2002) proposed two models
Section 3 proposes a solution framework according to and adopted the Benders decomposition to solve
the supply chain context and management considera- logistics network designs of warehousing, transporta-
tions for the S&OP problem, and designs a decision tion centres and production base selection. The first
model to solve the problem using the heuristic method. model focuses on setting transfer prices in the global
Section 4 evaluates the effectiveness of the heuristic supply chain with the objective of maximising the after
solution with data derived from a leading TFT-LCD tax profit of an international corporation. The second
TV company, reports the computational results and model focuses on production and distribution alloca-
provides analysis and conclusions. tion with seasonal demands in a single country system.
Bhutta et al. (2003) considered exchange rates, multiple
products and multiple costs to propose a mixed integer
2. Literature review
programming method to plan production, warehous-
2.1. Supply chain planning ing, transportation and investment issues to provide
Supply chain refers to an integrated system, which multinationals a decision model by which to allocate
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

synchronises a series of interrelated business processes their production capacities. The model maximises total
to acquire raw materials and parts, add value or profit by simultaneously considering restrictions such
transform components into finished goods and dis- as production costs and variable costs for production
tribute products to either retailers or customers. These capacity. Yeh (2005) modified the basic network model
processes can be divided into inbound logistics and with minimum cost to design a multi-stage supply
outbound logistics (Min and Zhou 2002). The former chain network. The model determines how to set up a
concerns storage of raw materials, parts and supplies, production site and a distribution centre. The network
and supports the production process from procure- takes nodes to supply chain facilities and arcs to
ment of production materials to production planning logistic flow to develop a hybrid heuristic algorithm by
and eventually to final product distribution. The latter combining a greedy algorithm and a local search.
comprises all physical distribution activities related to Chern and Hsieh (2007) considered the master plan-
providing customer services such as order receipt, ning problem for a supply chain network using
fulfilment, inventory management, warehousing and outsourcing capacity. They proposed a proper model
transportation. Erengüç et al. (1999) divided supply with mixed integer linear programming (MILP) to
chain into three stages including supplier stage, plant minimise total cost of supply chains including material,
stage and distribution stage from the operational production, processing, transportation and inventory
perspective, and designed a decision model for each holding and developed a heuristic approach based on a
stage. In methodology application, mathematical greedy algorithm to solve the model. Bertel et al.
programming is a useful method to construct a (2008) focused on multiprocessor budgeting planning
decision model in strategic and operation levels for in the supply chain and proposed a greedy algorithm to
the supply chain (Dhaenens-Flipo and Finke 2001, maximise average cash flow with financial constraints.
Goetschalckx et al. 2002, Bhutta et al. 2003). Previous Many cases reported in the literature are developed
researchers have designed heuristics to search a feasible solutions based on optimisation methods because of
solution that considers the influence of problem size on poor communication and coordination throughout the
efficiency (Goetschalckx et al. 2002, Jang et al. 2002, supply chain (Rudberg and Thulin 2009). The research
Bhutta et al. 2003, Yeh 2005, Chern and Hsieh 2007). puts forward one solution to planning problems.
Jang et al. (2002) considered decision problems
including network design, capacity, production plan-
ning and distribution planning. To operate efficient 2.2. Sales and operations planning
supply chains necessitates considering these problems As a tactical-based integrated planning process, S&OP
simultaneously. A proposed system framework com- envisages a plan that involves several functional units
prises an optimised supply network-design module, an to ensure that the aligned plans support the business
integrated planning module for production and dis- strategic goal. The S&OP plan develops production
tribution operations covering suppliers to customers, a based on a sales plan that translates into a capacity-
model management module supporting flexible math- requirement estimation. The difference between sales
ematical modelling, and a data management module. demand and supply capability results in a correspond-
Decision models among system modules can be ing inventory plan and backlog plan (Olhager et al.
divided into three sub-models, including inbound 2001). The scope of tactical planning involves procure-
network, distribution network and outbound network, ment, production, distribution and sales (Thomas and
 250 J.-Z. Wang et al.

Griffin, 1996). These issues can further combine to order taking for high-customised products. Mourtzis
buyer-vendor, inventory-distribution and production- et al. (2008) proposed a checking availability procedure
distribution (Vidal and Goetschalckx 1997). Tactical for a customised component before order taking.
planning integrates and consolidates data needed in Makris et al. (2008) used XML to build an order
cross functions of sales, production, distribution and taking system with a ship repair case. Chryssolouris
procurement in planning processes (Genin et al. 2007). et al. (2004) provided a data exchange mechanism for
A basic planning logic flow of S&OP is: various partners in the same supply chain to exchange
data among heterogeneous IT systems. Chryssolouris
. Demand forecast calculation in the sales depart- et al. (2004) experimented with a heterogeneous data
ment in term of product groups; exchanged mechanism in a ship repair industry.
. Execute calculation and capacity balancing for Previous studies have discussed S&OP and supply
each plant by the planner and manager; chain planning from various viewpoints such as
. Simulation of scenarios and feasible plans for the procurement, production, scheduling, inventory or
whole company supply chain; transportation. However, research based on an inte-
. Modification of resource requirements and avail- grated perspective is scarce. This article solves the
abilities and feasible plans validation by the planning problem at the tactical level and considers the
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

planner; environment constraint within the scope of the supply

. Compare the scenarios and select the best plan in chain including procurement, production, inventory
the monthly meeting. and transportation. Based on the proposed problem-
solving framework, this work considers the problem as
During the S&OP process, managers want feasible a 0–1 MILP problem and proposes a heuristic to
plans under the planning objective of maximising obtain solutions.
overall profit within product revenue, production cost,
distribution cost, and inventory cost with the following
3. An approach for advanced sales and operations
assumptions: Demand is defined such that at least one
planning
customer orders each product. The unit costs of
transportation, the fixed costs on plants and DCs, 3.1. Planning framework
and the site capacities in every data set are extracted For quick market response, demands fulfilment is a top
from the ERP system. The product holding cost of a priority for an entire supply chain. Unfortunately,
product over two periods is the same level as its price when capacity is tight, limited resources must be
reduction. Product refers to BOMs. The result of utilised more efficiently. In the resource allocation
S&OP should fit with managerial concerns. For process, members in a supply chain are responsible for
example, minimising total cost or maximising profit different tasks and generate their own costs via
is the common practice for planning objectives. Several procurement, production and distribution. Allocating
researches on S&OP have focused on planning resources efficiently according to demands is necessary
procedures in the product group to turn forecast to the whole supply chain to maximise overall profits.
demand into capacity requirement and inventory level This study produces operation plans to maximise profit
(Genin et al., 2005), to adjust capacity shifting strategy while increasing product revenue, and to minimise the
(Olhager et al. 2001;, Affonso et al. 2008), to optimise cost of materials, production, transportation and
production procedures (Feng et al., 2008), and to align inventory holding. All information regarding capacity
supply flexibility with order uncertainty (Chen-Ritzo limitations and deadline demand is involved in a given
et al. 2010). Makris and Chryssolouris (2010) proposed supply chain network.
an uncertainty evaluation mechanism by Bayesian This article proposes an integrated multi-period,
networks to determine whether the manufacturer multi-product, multi-echelon production-distribution
should take customised orders and the risk level. decision framework as shown in Figure 1 and decision
Makris et al. (2011) built a risk assessment mechanism model in the supply chain to allocate customer demand
by Bayesian networks to figure out purchasing in given time periods. The framework generates real
uncertainty of the customised order. time procurement, production and distribution plans
In the broad scope of the supply chain, a more for distributed manufacturing sites and distribution
systematic method is mandatory for precise and centres over the supply chain network at the opera-
effective processing of every stage. Without a systema- tional level. These plans consider a given demand and
tic method, processing massive amounts of informa- BOM relations among supply chain members, includ-
tion using a manual operation would be impossible. ing master planning for each manufacturing site under
Previous works has focused on solving similar issues. a material requirement planning (MRP) environment
Mourtzis et al. (2008) built a process control system of and a distribution centre for customer demands.
 International Journal of Computer Integrated Manufacturing 251
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

Figure 1. Solution framework of S&OP in the supply chain.

The proposed planning framework first aggregates

the total demand from global market orders, and then 3.2. Decision model formulation
obtains operation data from the ERP database as The model is formulated as a mathematical decision
input parameters in the decision model. This article model which maximises the overall profit of product
utilises two ways to solve the production-distribution revenue, production, distribution and inventory cost,
problem. One is based on commercial software for an subject to the various kinds of constraints. In the
optimal solution, such as CPLEX and LINGO. The design of model, we make the following assumptions:
commercial software solver is a black box to explore
effective suitable solutions. The black box means the . Cut the planning time horizon into time buckets.
solver yields a solution without interaction with the . Demand is deterministic.
user after sending master data to the software. Most . Backlogging is allowed.
commercial software cannot solve large scale MILP . Distribution and manufacturing lead times are
problems in a reasonable time (Fischetti and Lodi fixed.
2003). Using the heuristic algorithm is another way to . Exchange rates are fixed.
obtain the planning solution. In contrast to the . Products are stored as inventory.
software solution, the proposed heuristics in this . Assembly of product refers to BOMs.
article are an alternative solution for large size
problems. The following describes the decision model We use the following notations to define the
and heuristics. decision model.
 252 J.-Z. Wang et al.

3.2.1. Index and parameter UPDCdcp Required unit shipping capacity for
The decision model provides several objective func- product p from DC d to channel c
tions for different planning strategy. We use the UCPp Required unit production capacity for
following notations to define the data field of product p
mathematical model: UIPp Required unit storage capacity for
product p
. Index: UIMm Required unit storage capacity for
P set of product groups, p 2 P material m
T set of planning periods, t 2 T LTFPfp Cycle time of product p to plant f
M set of materials, m 2 M LTVFvf Lead time from vendor v to plant f
F set of plants, f 2 F LTFDfd Lead time from plant f to DC d
C set of channels, c 2 C LTDCdc Lead time from DC d to channel c
D set of distribution centres (DCs), d 2 D 
1; if plant f can produce product p
V set of vendors, n 2 V SPFtfp
0; otherwise
. Parameter: . Decision variable:
BLPtcp Backlog quantity of product p to
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

PRIp Sales price for product p

DEMtcp Demand quantity of product p to channel c in period t
channel c in period t RCPtcp Received quantity of product p to c
CAPf Total available production capa- channel c in period t
city of plant f PCMtvm Procurement quantity of material m to
CAPFDfd Maximum shipping capacity for vendor v in period t
product p from plant f to DC d QPtfp Quantity of product p produced at
CAPDCdc Maximum shipping capacity plant f in period t
for product p from DC d to IPtfp Quantity of product p stored at plant f
channel c in period t
WHPf Maximum storage capacity for IMtfm Quantity of material m stored at plant
product in plant f f in period t
WHMf Maximum storage capacity for TMVFtvfm Quantity of material m purchased from
material in plant f vendor v to plant f in period t
WHPDd Maximum storage capacity for TPFDtfdp Quantity of product p shipped from
product in DC d plant f to DC d in period t
BOMpm Material m consumption quantity TPDCtdcp Quantity of product p shipped from
for producing unit product p DC d to channel c in period t
SUPvm Maximum supply capacity of ven-
dor v for material m 
1; if shipping path is used for material in period t
BM A very big number SVFtvf
0; otherwise
CTBLPp Unit backlog cost for product p 
CTBLPp Unit procurement cost of material 1;if shipping path is used for product in period t
SVFtfd
m for vendor v 0;otherwise
CTMvm Unit shipping cost of product p 
for plant f 1;if shipping path is used for product in period t
SVFtdc
CTMANfp Unit shipping cost for product p 0;otherwise
from plant f to DC d
CTTRAFDfdp Unit shipping cost for product p . Objective function and constraints:
from DC d to channel c
CTINVPp Unit holding cost for product p The model includes three objective functions. Con-
CTINVMm Unit holding cost for material m straint (1) states the total revenue. Constraints (2)–(6)
Ev Exchange rate for vendor v present backlog cost, procurement cost, production
Ef Exchange rate for plant f cost, inventory cost and shipping cost, respectively.
Ed Exchange rate for DC d The objective of maximum profit is to maximise the
Ec Exchange rate for channel c net profit:
UPFDfdp Required unit shipping capa-
city for product p from plant f to Max Profit ¼ Max TREV  TBLC  TMAT
DC d TMAN  TINV  TTRA:
 International Journal of Computer Integrated Manufacturing 253
X
The objective of minimum total cost is to minimise SDCtdc  1; 8c; t: ð11Þ
the costs of procurement, production, backlog, in- d

ventory and shipping:

Constraints concerning the production: this group
Min Total Cost ¼ Min TREV  TMAT of constraints limits the capability in production and
 TMAN  TINV  TTRA: inbound logistics. Constraint (12) ensures that the
received product quantity in a DC is equal to
XXX
TREV ¼ t
RCPcp  PRIp  Ec ð1Þ aggregated shipment from plants. Constraint (13)
t c p describes the restriction of the shipping quantity.
Constraint (14) describes the restriction of shipping
XXX
TBLC ¼ t
BLPcp  CTBLPp  Ec ð2Þ capacity from plant to DC. Constraint (15) describes
t c p the availability of the shipping route. Constraint (16) is
the balance equations for the inventory of products in
XXXX
TMAT ¼ t
TMVFvfm  CTMm  Ev a plant. Constraint (17) identifies if the manufacturing
t v m f
sites possess enough production technique to produce
the products. Constraint (18) describes the production
ð3Þ
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

capacity restriction of the plant. Constraint (19)

XXX describes the material inventory capacity restriction
TMAN ¼ QPtfp  CTMANfp  Ef ð4Þ of the plant. Constraint (20) describes the product
t f p
inventory capacity restriction of the plant.
XXX X X
TINV ¼ IPtfp  CTINVPp  Ef ð5Þ TPFDtLTFD
fdp
fd
¼ TPDCtdcp ; 8d; p; t ð12Þ
t p f c
f

XXXX TPFDtfdp  SFDtfd  BM; 8f; d; p; t ð13Þ

TTRA ¼ TPFDtfdp  CTTRAFDfdp
t f d p
X
XXXX TPFDtfdp  UPFDfdp  CAPFDfd ; 8f; d; t ð14Þ
 Ed þ TPDCtdcp p
t d c p
X
 CTTRADCdcp  Ec : ð6Þ SFDtLTFD
fd
fd
 SDCtdc ; 8f; d; t ð15Þ
c

Constraints concerning the sales: this group of tLTFPfp

X
constraints limits the demand fulfilment of product IPtfp ¼ IPt1
fp þ QPfp  TPFDtfdp ; 8f; p; t ð16Þ
group. Constraint (7) ensures the backlog quantity for d

demand. Constraint (8) ensures that the received  

product quantity in a channel is equal to aggregated 1 > 1  SPFtfp QPtfp ; 8d; t ð17Þ
shipment from DCs. Constraint (9) describes the
restriction that the shipping quantity is less than X
product demand. Constraint (10) describes the restric- UIPp  IPtfp  WHPf ; 8f; t ð18Þ
p
tion of shipping capacity from DC to channel.
Constraint (11) ensures that at least one route is X
available from DC to channel. UCPp  QPtfp  CAPf ; 8f; t ð19Þ
p

RCPtcp þ BLPtcp ¼ DEMtcp ; 8c; p; t ð7Þ X

UIPp  IPtfp  WHPf ; 8f; t: ð20Þ
p
X
TPDCtLTDC
dcp
dc
¼ RCPtcp ; 8c; p; t ð8Þ
d Constraints concerning the procurement: Con-
straint (21) ensures that the received material quantity
TPDCtLTDC
dcp
dc
 SDCtLTDC
dc
dc
 DEMtcp ; 8d; c; p; t in the plant is equal to aggregated shipment from
ð9Þ vendors. Constraint (22) is the balance equations for
X the inventory of materials. Constraint (23) describes
TPDCtdcp  UPDCdcp  CAPDCdc ; 8d; c; t ð10Þ the material inventory capacity restriction of the plant.
p Constraint (24) describes the material supply
 254 J.-Z. Wang et al.
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

Figure 2. Conception of the S&OP heuristics.

restriction of the vendor. Constraint (25) ensures that

at least one vendor is available for the plant.
Constraint (26) describes the restriction of the shipping
quantity from vendor to plant.
X
TMVFtLTVF
vfm
vf
¼ PCMtvm ; 8v; m; t ð21Þ
f
X
IMtfm ¼ IMt1
fm þ TMVFtLTVF
vfm
vf

v
X ð22Þ
t
 QPfp  BOMpm ; 8f; m; t
X p
UIMm  IMtfm  WHMf ; 8f; t ð23Þ
m
X
TMVFtvfm  SUPvm ; 8v; m; t ð24Þ
f
X
SVFtvf  1; 8f; t ð25Þ
v

TMVFtvfm  SVFtvf  BM; 8v; f; m; t: ð26Þ

3.3. S&OP heuristic procedure

The integrated planning problem is complicated to
solve due to the number of integer and binary variables
representing network nodes, product items, materials
and the number of planning periods. Some difficulties
arise when trying to solve this model. The number of Figure 3. S&OP heuristics flow chart.
variables and constraints increases exponentially as the
number of demands or planning period increases.
Considering the problem complexity and the time approximate optimal solution, including Lagrangian
requirement for computation, an alternative solution is relaxation, the branch and bound method and the
necessary. Two types of developed heuristics can solve Bender decomposition. The other is based on intuition
problems. One is based on the optimisation theory to or decision rules applied to problems. The difficulties in
 International Journal of Computer Integrated Manufacturing 255

solving the S&OP problem are to find adequate flow capacity constraint of manufacturing sites.
within a given supply chain network, or flow routes in Assign the demand of product p to manufac-
the supply chain network that are not unique. When turing site f in priority order. Adjust capacity
allocating demands, it is hard to decide which time of the manufacturing sites, remove the order
interval to plan for product and inventory holding, and and continue the next order until all demands
which route to use. Thus, only a few computationally have been assigned to production, or all
efficient heuristics are available for solving the capacities have been utilised. Finally, calculate
problem. the shortage of product demand in each
This study developed a heuristic solution of period.
resources allocation to solve the problem. The (3) Calculate the material requirements using the
expectation is confirmed by the computational result. BOMs and estimate whether the supply
The following describes each heuristic procedure. The capability of vendors is affordable in each
developed heuristic in this study (Figure 2) selects a period. If the supply capability is overloaded
candidate from sets of material vendors, manufactur- with material requirements, the production
ing sites and DCs and then decides production plan needs to be modified to fit the material
quantities and inventory subject to capacity restric- constraint.
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

tions of each stage in the supply chain. The heuristic (4) Consider the shortage of product demand and
also considers transportation quantities. remaining capacity and materials in each
The S&OP heuristic of this research uses a five step- period, and balance the shortage demand to
procedure as shown in Figure 3 and Appendix 1: remaining capacity from the perspective of
forward planning. The available forward per-
(1) Aggregate the total demand from orders. iod in each product is restricted by inventory
From a multi-tasking perspective, demands cost of product. The lowest inventory cost of
of the product in a period can be assigned to product is first selected to reallocate.
several planning periods according to the
comparative forward planning cost of one Consider the distribution centre-channel transpor-
product as compared with others. Sort the tation cost and channel demands to find the lowest link
production priority for products that consid- and decide the amounts of products to be shipped from
ers variable profit determined from the pro- distribution centres to channels. Next, consider the
duct price, transportation cost and inventory manufacturing site-distribution centre transportation
cost across planning periods. cost and distribution centre storage capacity to find the
(2) Execute the capacity allocation according to lowest link and decide the amounts of products to be
the production priority, and allocate the shipped from manufacturing sites to distribution
product production quantity under the centres.

Figure 4. The supply chain network for the case company.

 256 J.-Z. Wang et al.

Table 1. The demands forecast volume for each product group.

Time periods
Channel Product T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12
C1 K206-10U1 253 334 313 205 241 398 245 310
C2 ST43-15U1 222 327 337 202 216 259 256 249
C3 LT12-23U1 383 490 260 333 304 210 374 323
C4 K206-10U1 256 291 218 248 291 350 351 293
C5 ST43-15U1 293 303 391 355 325 289 310 204
C6 LT12-23U1 340 345 380 352 320 306 353 276

Table 2. The material purchasing BOM cost for each

4. Scenario simulation with solution framework products group.
4.1. Case description
Bill of material
The proposed generic model can fit into the multi-site Unit
purchasing K206- ST43- LT12-
supply chain network. This research implemented the Material cost 10U1 15U1 23U1
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

model with permission from the Taiwan TFT-LCD TV

company that owns multi-sites in Asia, with sales Panel 30 3 4 10
worldwide. Moore’s law states that the price of PCBA 80 1 2 4
Mec. 45 1 2 6
electronic products drops 100% every 12 months. To
cope with the market racket test, it is important for top
managers to rapidly and precisely point out product
combination rates between cash cow products and Table 3. The products sales prices, production cost and
tomorrow star products. Top managers are eager to inventory cost for each product group.
learn the cost benefit of various types of product Production
combinations, which this research seeks to provide. Product Product price cost Inventory cost
The following illustrates the supply chain network of
K206-10U1 400 28 7
the TFT-LCD TV manufacturer. ST43-15U1 700 31 12
The company has manufacturing sites located in LT12-Z3U1 1300 44 22
Taiwan and China. Three product types were chosen
from the company’s main product line, characterised
by LCD size. The three main sizes of LCD TV include distances between factories, DCs and customer desti-
K206-10U1-000 (9.6 in), ST43-15U1-000 (15 in) and nations in Table 4. The product carrying cost of a
LT12-23U1-000 (23 in). The LCD TV products are product over two periods is the same level as its price
assembled with parts such as panels, PCBAs and reduction. Demand forecast, material purchasing
mechanical components (Figure 4). This research BOM cost, product sales prices, production and
proposes a planning architecture to simulate four inventory cost, and transportation cost are shown in
S&OP stage factors of capacity, procurement cost, Tables 1–4. These datasets are characterised by
inventory cost and logistic cost with loose or tight suppliers, producers, distributors, channels and the
strategies. Sixteen S&OP simulation scenarios were planning periods of time horizon. The proposed
derived from these four planning factors with two method in this research derived the material plan in
strategies. Sixteen production and inventory simula- Table 5, the production plan in Table 6 and the
tion plans were derived from planning variances of distribution plan in Tables 7 and 8.
three product groups, three material groups, six For a practical research model that can fit into
suppliers, two manufacturing sites, four warehousing industry practices, this research designed a model with
and distribution centres, with six customers over a four planning factors of capacity, procurement cost,
planning horizon of 12 periods. The historical opera- inventory cost and logistic cost with two loose or tight
tional data of the company were extracted from the strategies. Sixteen simulation scenarios were designed
ERP database and transformed as planning data into based on the above four planning factors with two
the research model for a planning run. planning strategies. The experimental data were con-
Customer demand from the channel was defined ducted to test the validity of the research model. The
for each product as an order from at least one planning performances between the heuristic and the
customer. Transportation costs, fixed costs of plants, optimal solution were compared by the error gap each
DCs and site capacities were randomly generated. The solution generated. Capacity, procurement cost, pro-
differentiations of distribution costs were set to the duct inventory cost and product logistics cost were
 International Journal of Computer Integrated Manufacturing 257

Table 4. The transportation cost of each product type.

Site DC Product Transportation cost DC Channel Product Transportation cost

F1 D1 K206-10U1 6 D1 C1 K206-10U1 12
ST43-15U1 12 ST43-15U1 24
LT12-23U1 18 LT12-23U1 36
D2 K206-10U1 15 C2 K206-10U1 18
ST43-151 29 ST43-15U1 36
LT12-23U1 47 LT12-23U1 54
F2 D1 K206-10U1 13 D2 C1 K206-10U1 24
ST43-15U1 26 ST43-15U1 32
LT12-23U1 45 LT12-23U1 48
D2 K206-10U1 4 C2 K206-10U1 8
ST43-15U1 8 ST43-15U1 16
LT12-23U1 12 LT12-23U1 24

Table 5. The material plan of each material type.

Downloaded by [Simon Fraser University] at 19:21 15 March 2015

Time periods
Vendor Site Material T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12
V1 F1 Panel 903 1882 157 218 338 0 874 156 0 0 0 0
F2 Panel 5073 5118 5451 5219 4965 4560 5181 4455 0 0 0 0
V2 F1 PCBA 0 1935 1448 0 0 896 0 0 0 0 0 0
F2 PCBA 1484 0 0 1307 1211 0 1547 905 0 0 0 0
V3 F1 Mec. 1930 2605 1728 1677 1459 0 2001 1103 0 0 0 0
F2 Mec. 0 0 0 0 0 928 0 0 0 0 0 0
V4 F1 Panel 5000 4875 5000 5000 5000 4729 5000 5000 0 0 0 0
F2 Panel 0 0 0 0 0 271 0 0 0 0 0 0
V5 F1 PCBA 2371 862 669 2096 2126 993 2378 2100 0 0 0 0
F2 PCBA 629 2138 2331 904 874 2007 622 900 0 0 0 0
V6 F1 Mec. 1207 1172 909 1085 1275 2309 1125 1643 0 0 0 0
F2 Mec. 2793 2828 3091 2915 2725 1691 2875 2357 0 0 0 0

Table 6. The production plan of each product type.

Time periods
Site Product T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12
F1 K206-10U1 0 395 183 403 360 478 531 370 310 0 0 0
ST43-15U1 0 222 327 337 202 216 259 256 249 0 0 0
LT12-23U1 0 383 490 260 333 304 210 374 323 0 0 0
F2 K206-10U1 0 167 152 29 93 155 205 137 293 0 0 0
ST43-15U1 0 293 303 391 355 325 289 310 204 0 0 0
LT12-23U1 0 340 345 380 352 320 306 353 276 0 0 0

selected as key factors representing environment demand control, loose material purchasing BOM
uncertainty. Each key factor can be taken under two cost, loose production and inventory cost, and loose
planning strategies of loose or tight scenarios. Table 9 transportation cost. These four planning strategies
shows a list of 16 scenarios based on these four represent four S&OP stages of demand, purchasing,
planning factors with two planning strategies. production and transportation. The output per-
formance of heuristic frameworks for these four
stages is the same as the software solution at
4.2. Scenario explanation
3,592,046. That is, total cost derived from heuristic
4.2.1. Scenario one frameworks is the same as the software solution that
The S&OP plans have the same planning strategies proves the heuristic frameworks of this research are
of loose strategies. The S&OP plan has loose robust.
 258 J.-Z. Wang et al.

Table 7. The distribution plan of each product type from site to DC.

Time periods
Site DC Product T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12
F1 D1 K206-10U1 0 0 342 236 403 360 478 531 370 310 0 0
ST43-15U1 0 0 222 327 337 202 216 259 256 249 0 0
LT12-23U1 0 0 383 490 260 333 304 210 374 323 0 0
D2 K206-10U1 0 0 0 0 0 0 0 0 0 0 0 0
ST43-15U1 0 0 0 0 0 0 0 0 0 0 0 0
LT12-23U1 0 0 0 0 0 0 0 0 0 0 0 0
F2 D1 K206-10U1 0 0 0 0 0 0 0 0 0 0 0 0
ST43-15U1 0 0 0 0 0 0 0 0 0 0 0 0
LT12-23U1 0 0 0 0 0 0 0 0 0 0 0 0
D2 K206-10U1 0 0 167 152 29 93 54 217 226 293 0 0
ST43-15U1 0 0 293 303 391 355 325 289 310 204 0 0
LT12-23U1 0 0 340 345 380 352 320 306 353 276 0 0
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

Table 8. The distribution plan of each product type from DC to channel.

Time periods
DC Channel Product T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12
D1 C1 K206-10U1 0 0 0 253 236 313 205 241 398 245 310 0
C2 ST43-15U1 0 0 0 222 327 337 202 216 259 256 249 0
C3 LT12-23U1 0 0 0 383 490 260 333 304 210 374 323 0
C4 K206-10U1 0 0 0 89 0 90 155 237 133 125 0 0
C5 ST43-15U1 0 0 0 0 0 0 0 0 0 0 0 0
C6 LT12-23U1 0 0 0 0 0 0 0 0 0 0 0 0
D2 C1 K206-10U1 0 0 0 0 0 0 0 0 0 0 0 0
C2 ST43-15U1 0 0 0 0 0 0 0 0 0 0 0 0
C3 LT12-23U1 0 0 0 0 0 0 0 0 0 0 0 0
C4 K206-10U1 0 0 0 167 152 29 93 54 217 226 293 0
C5 ST43-15U1 0 0 0 293 303 391 355 325 289 310 204 0
C6 LT12-23U1 0 0 0 340 345 380 352 320 306 353 276 0

Table 9. Factor index of experiment test.

Factor Description
LC Loose capacity, the total demand at per planning period are set to 75% *95% of total production capacities
TC Tight capacity, the total demand at per planning period are set to 95% *115% of total production capacities
PCL Lower procurement cost for components, the costs are set to the original component price
PCH Higher procurement cost for components, the added costs are correspond to variation in component price
ICL Lower product inventory cost, the costs are set to the original cost
ICH Higher product inventory cost, the added costs are correspond to reduction in product price
TCL Lower product logistics cost from sites and channels, the costs are set to the original cost
TCH Higher product logistics cost from sites and channels, the added costs are correspond to fluctuation in exchange rate

4.2.2. Scenario two production and inventory cost, and loose in

Changed demand control in scenario two is tight, transportation cost. Output performance of the
and the other three planning factors remain at the heuristic research model is slightly lower than the
same loose strategy. That is, the strategy planning software solution. This test result showed that the
factors are tight in demand forecast control, performance of this study is only slightly lower
loose in material purchasing BOM cost, loose in than the optimiser.
 International Journal of Computer Integrated Manufacturing 259

Table 10. Planning output in different scenarios.

Planning output comparison

Scenario Factor combination LINGO Heuristic GAP (%)
1 LC-PCL-ICL-TCL 3,592,046 3,592,046 0.00
2 TC-PCL-ICL-TCL 3,533,108 3,331,144 5.72
3 LC-PCH-ICL-TCL 2,106,441 2,106,441 0.00
4 TC-PCH-ICL-TCL 2,061,951 1,947,508 5.55
5 LC-PCL-ICH-TCL 3,591,976 3,591,948 0.00
6 LC-PCH-ICH-TCL 2,106,371 2,106,343 0.00
7 TC-PCL-ICH-TCL 3,531,407 3,326,355 5.81
8 TC-PCH-ICH-TCL 2,060,250 1,943,495 5.67
9 LC-PCL-ICL-TCH 3,148,610 3,148,610 0.00
10 TC-PCL-ICL-TCH 3,084,789 2,9123,50 5.59
11 LC-PCH-ICL-TCH 1,663,005 1,658,409 0.28
12 TC-PCH-ICL-TCH 1,613,632 1,528,714 5.26
13 LC-PCL-ICH-TCH 3,148,512 3,148,312 0.00
14 LC-PCH-ICH-TCH 1,662,907 1,662,907 0.00
15 TC-PCL-ICH-TCH 3,082,643 2,907,361 5.68
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

16 TC-PCH-ICH-TCH 1,611,486 1,524,701 5.39

Table 11. Time performance in different scenarios.

Time (S) Planning result (max profit)

Problem size (f-d-c-p) LINGO S&OP heuristics Saving time LINGO S&OP heuristics Gap (%)
5-5-5-5 7 5 2 2,605,070 2,451,879 5.88
5-5-10-10 45 9 36 8,546,040 8,012,000 6.25
5-5-15-15 181 18 163 19,347,900 18,134,940 6.27
5-10-15-20 891 110 781 28,169,360 26,617,832 5.51
5-10-15-25 1399 424 975 34,520,300 32,573,231 5.64
10-10-20-20 1942 437 1505 38,535,900 36,061,434 6.42
10-10-20-25 3187 1206 1981 48,949,800 45,908,227 6.21
10-10-25-30 6247 1578 4669 74,042,400 70,565,354 4.70
10-15-25-25 11,099 2087 9012 62,073,500 58,888,592 5.13
10-15-30-30 16,299 2286 14,013 91,135,100 85,949,232 5.69
15-15-30-35 29,007 9755 19,252 107,313,400 103,246,648 3.79
15-15-35-40 – 10,265 – – 184,665,054 –

research model and software solution are listed in

4.2.3. Scenario three Table 10.
The restored planning strategy of demand forecast The proposed heuristic was coded in heuristic by
returns to loose and the changed second planning Borland Cþþ. To prove the execution performances of
factor of material purchasing BOM cost returns to the research model, the error gap of the research model
tight strategy, while the third and fourth strategies for the Borland Cþ þ was compared to the contrasted
remain at loose strategy. That is, demand forecast solution of LINGO 9.0. The error gap between the
control is loose, material purchasing BOM cost is tight, heuristic solution and the optimal solution was
production and inventory cost is loose, transportation calculated as (LINGO – Heuristic)/LINGO 6 100%.
cost is loose, and output performance of the heuristic Table 10 shows that the result of the heuristic solution
research model is the same as the software solution at is only slightly worse, but the optimal solution is equal.
2,106,441. The gap difference is 56%.
The switching of planning strategies was repeated Finally, the execution performances of time con-
from loose to tight for every planning factor for sumption for the proposed heuristic and LINGO were
the remaining simulation scenarios until all 16 compared. Computational experiments were con-
combinations of planning factors were run. The ducted to test the proposed decision model and
comparisons of output performances for the heuristic solution approach. This work extended the scale of
 260 J.-Z. Wang et al.

numerical examples. Thirty problems were randomly Bhutta, K.S., et al., 2003. An integrated location, production,
generated for each problem size shown in Table 11. and investment model for a multinational corporation.
International Journal of Production Economics, 86 (3),
For large-scale problems, finding the optimal solution 201–216.
is time consuming. It takes about 16,299 s and Chern, C.-C. and Hsieh, J.-S., 2007. A heuristic algorithm
29,007 s, respectively, to solve 10-15-30-30 and 15– for master planning that satisfies multiple objectives.
15–30–35 numerical examples of the problem size. Computers and Operational Research, 34 (11), 3491–
Along with increasing problem size, the incremental 3513.
Chen-Ritzo, C.-H., et al., 2010. Sales and operations
saving time of the proposed heuristic becomes larger planning in systems with order configuration uncertainty.
than LINGO. The incremental time saving proves that European Journal of Operational Research, 205 (3), 604–
the proposed heuristic is robust. The proposed plan- 614.
ning framework is generic and not specific to the above Chryssolouris, G., et al., 2004. Towards the internet based
case study. supply chain management for the shiprepair industry.
International Journal of Computer Integrated Manufac-
turing, 17 (1), 45–57.
Dhaenens-Flipo, C.D. and Finke, G., 2001. An integrated
model for an industrial production-distribution problem.
5. Conclusions IIE Transactions, 33 (9), 705–715.
Erengüç, S.S., Simpson, N.C., and Vakharia, A.J., 1999.
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

This work builds an integrated problem-solving frame-

Integrated production/distribution planning in supply
work to solve S&OP problems in a consensus. chains: an invited review. European Journal of Opera-
Operation-related data can be gathered from the tional Research, 115, 219–236.
ERP system and simulated for tactical planning of Feng, Y., D’Amours, S., and Beauregard, R., 2008.
firms. This integrated planning framework provides The value of sales and operations planning in
top managers a complete consensus view of S&OP oriented strand board industry with make-to-order
manufacturing system: cross functional integration
stages of demand, purchasing, production and trans- under deterministic demand and spot market recourse.
portation at one time. This study adopted an International Journal of Production Economics, 115 (1),
integrated perspective and systematic methodology to 189–209.
allocate limited resources more efficiently in the supply Fischetti, M. and Lodi, A., 2003. Local branching. Mathe-
chain. The methodology defines decision parameters matical Programming, 98 (1–3), 23–47.
Genin, P., Lamouri, S., and Thomas, A., 2005. Sales and
and variables with industrial characteristics and operations planning optimization. Applied Optimization,
proposes a decision model for a multi-period tactical 94, 191–204.
planning problem. A resources allocation heuristic Genin, P., Thomas, A., and Lamouri, S., 2007. How to
based on resource constraints was developed through manage robust tactical planning with an APS (Advanced
intuition procedures to solve this problem. Finally, Planning Systems). Journal of Intelligent Manufacturing,
18 (2), 209–221.
effectiveness of the proposed heuristic was tested using Goetschalckx, M., Vidal, C.J., and Dogan, K., 2002.
an extensive computational experiment test. We Modeling and design of global logistics systems: a review
conclude that the heuristic provides feasible solutions of integrated strategic and tactical models and design
for problems and is flexible enough to design scenarios algorithms. European Journal of Operational Research,
to execute what-if analysis. The planning results can 143, 1–18.
Jang, Y.-J., et al., 2002. A combined model of network design
also pass to members for more detailed production and production/distribution planning for a supply net-
scheduling or transportation planning. We believe that work. Computers and Industrial Engineering, 43 (1–2),
companies can benefit from the proposed approach 263–281.
and solution described in this article. Makris, S. and Chryssolouris, G., 2010. Customer’s behavior
Future research can extend the proposed model modeling for manufacturing planning. International
Journal of Computer Integrated Manufacturing, 23 (7),
based not only on a profit-oriented planning strategy, 619–629.
but also as a multi-criteria decision analysis. This Makris, S., Zoupas, P., and Chryssolouris, P.G., 2011.
work focuses on an interactive decision support Supply chain control logic for enabling adaptability
system (DSS) in uncertain environments. under uncertainty. International Journal of Production
Research, 49 (1), 121–137.
Makris, S., et al., 2008. On the information modeling for the
electronic operation of supply chains: a maritime case
References study. Robotics and Computer Integrated Manufacturing,
Affonso, R., Marcotte, F., and Grabot, B., 2008. Sales and 24 (1), 140–149.
operations planning: the supply chain pillar. Production Min, H. and Zhou, G., 2002. Supply chain modeling: past,
Planning & Control, 19 (2), 132–141. present and future. Computers and Industrial Engineering,
Bertel, S., Fenies, P., and Roux, O., 2008. Optimal cash flow 43, 231–249.
and operational planning in a company supply chain. Mourtzis, D., et al., 2008. Supply chain modeling and
International Journal of Computer Integrated Manufac- control for producing highly customized products. CIRP
turing, 21 (4), 440–454. Annals – Manufacturing Technology, 57 (1), 451–454.
 International Journal of Computer Integrated Manufacturing 261

Olhager, J., Rudberg, M., and Wikner, J., 2001. Long-term 30 Dpt ¼ 0
capacity management: linking the perspectives from 31 Else
manufacturing strategy and sales and operations plan- 32 //---If product demands are more than capacity, then
ning. International Journal of Production Economics, 69 update the demands gap DEMpt---
(2), 215–225. 33 DEMpt ¼ Dpt*Wpf – CAPft
Rudberg, M. and Thulin, J., 2009. Centralised supply chain 34 CAPft ¼ 0
master planning employing advanced planning systems. 35 EndIf
Production Planning & Control, 20 (2), 158–167. 36 Else
Shapiro, J.F., 2007. Modelling the supply chain. 2nd ed. 37 Break
Australia: Duxbury press, Thomson Brooks. 38 EndIf
Yeh, W.C., 2005. A hybrid heuristic algorithm for the 39 EndFor
multistage supply chain network problem. International 40 //---Step 3---
Journal of Advanced Manufacturing Technology, 26, 675– 41 //---Calculate procurement quantity Qmvt based on
685. production quantity Qpft and supply constraint---
Vidal, C.J. and Goetschalckx, M., 1997. Strategic produc- 42 //Calculate the material requirements Dmt according to
tion-distribution models: a critical review with emphasis BOMs
on global supply chain models. European Journal of 43 For p 1 to max p, pþþ
Operational Research, 98, 1–18. 44 For f 1 to max f, fþþ
45 Dmt ¼ Dmt þ Qpft*BOMmp
46 Endfor
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

Appendix 1: HHeuristic heuristic pseudo code 47 EndFor

48 //Allocate the procurement quantity Qmvt to
1 //The variables are all mean set suppliers---
2 p ! product, m ! material or component, t ! period, 49 For v 1 to max PRIPv, vþþ
v ! vendor 50 For t 1 to max t, tþþ
3 f ! fab, d ! distribution centre, c ! channel 51 //Find available supply capacity REMmvt---
o ! order 52 If MAX(REMmvt, m 1 to max m) 40
4 Dpct, ! Demand Initialisation 53 For m 1 to max m, tþþ
5 CFp, CMp, INVp, TRAfdp, TRAdcp ! Cost 54 //---Consider material demands Dmt, available supply
Initialisation capacity REMmvt simultaneously to decide
6 Qpft, Ipft, Qmvt, Imvt, Qpfdt, Qpdct ! Planning prosurement quantity Qmvt---
Initialisation 55 Qmvt ¼ MIN (Dmt, REMmvt)
7 Pi, Dmt, Dpt, Wif, BOMmi, CAPft, , CAPdt, DEMpit, 56 //---If material demands are small than capacity, then
DEMmt, SUPmt, Pmv, REMmvt update the available REMmvt---
8 ! Constraint Initialisation 57 If Dmt 5 REMmvt
9 PRIPpft, PRIMv, PRIfdo, PRIdco ! Priority 58 REMmvt ¼ REMmvt – Dmt
Initialisation 59 Dmt ¼ 0
10 //---Step 1--- 60 Else
11 //––––––––––Sort production priority of 61 //---If material demands are small than capacity, then
product––––––––––– update the demands gap DEMmt---
12 //Decide production priority based on Price Pp, 62 DEMmt ¼ Dmt –REMmvt
manufacture cost CFp, material cost CMp, inventory 63 REMmvt ¼ 0
cost INVp, transportation cost TRAp 64 EndIf
13 PRIPo(p, f, t) ¼ Sort(Pp, CFp, CMp, INVp) 65 EndFor
14 //Decide procurement priority based on procurement 66 Else
cost from low to high 67 Break
15 PRIMo ¼ Sort(Cmv) 68 EndIf
16 //Decide transportation priority based on transportation 69 EndFor
cost from low to high 70 EndFor
17 PRIMfdo ¼ Sort(TRAfdp) 71 //---Step 4---
18 //Decide transportation priority based on transportation 72 //---Consider the shortage of product demand and
cost from low to high remaining capacity and materials, and modify the
19 PRIMdco ¼ Sort(TRAfdcp) production plan---
20 //---Step 2--- 73 For t 1 to max t, tþþ
21 //---Calculate production quantity based on production 74 // Find if the planning period needs to reallocate
quantity Qpft and demands--- 75 If DEMit or DEMmt 4 0
22 For o 1 to max PRIPo, oþþ 76 //Search available capacity and material from current
23 //Find available capacity CAPft--- period t to forward period t-x, x is forward periods---
24 If MAX(CAPft, f 1 to max f) 4 0 77 If MIN(CAPft, REMmvt, t t -1 to t-x, f 1 to max
25 //---Consider product demands Dpt, available capacity f, m 1 to max m, v 1 to max v) ¼ 0
CAPft simultaneously to decide production quantity--- 78 Break
26 Qpft ¼ MIN (Dpt, CAPft/Wpf) 79 Else
27 //--If product demands are small than capacity, then 80 //Calculate the production quantity TempI and modified
update the available CAPft-- inventory for forward planning---
28 If Dpt 5 CAPft/Wpf 81 TempI ¼ MIN(DEMpt, CAPf(t-x)/Wi, REMmv(t-x)/
29 CAPft ¼ CAPft – Dpt*Wpf BOMmp)
 262 J.-Z. Wang et al.

82 Qpf(t-x) ¼ Qpf(t-x) þ TempI 106 Dpct ¼ Dpct – CAPdt

83 Ipft ¼ Ipft þ TempI 107 CAPdt ¼ 0
84 //Calculate the procurement quantity TempM and 108 EndIf
modified inventory for forward planning--- 109 Else
85 TempM ¼ MIN(DEMp(DEMmt þ TempI* BOMmi, 110 Break
REMmv(t-x)) 111 EndIf
86 Qmv(t-x) ¼ Qmv(t-x) þ TempM 112 EndFor
87 Imvt ¼ Imvt þ TempM 113 //---Calculate transportation quantity from site to DC
88 EndIf based on transportation priority PRIfdo and channel
89 Else demands Dpt---
90 Break 114 For o 1 to max PRIfdo, oþþ
91 EndIf 115 //Find DC inventory demands CAPdt to fulfill---
92 EndFor 116 If MAX(CAPdt, d 1 to max d) 4 0
93 //---Step 5--- 117 //---Consider available product production Qpft, DC
94 //---Calculate transportation quantity from DC to inventory demand CAPdt simultaneously to decide
channel based on transportation priority PRIfdo and transportation quantity---
channel demands Dpt--- 118 Qpft ¼ MIN (Qpft, CAPdt)
95 For o 1 to max PRIfdo, oþþ 119 //--If product demands are small than inventory, then
96 //Find demands Dpct to fulfill-- update the available CAPdt--
97 If MAX(Dpct, c 1 to max c) 4 0 120 If CAPdt 5 Qpft
Downloaded by [Simon Fraser University] at 19:21 15 March 2015

98 //---Consider product demands Dpct, available DC 121 Qpft ¼ Qpft – CAPdt

inventory CAPdt simultaneously to decide 122 CAPdt ¼ 0
transportation quantity--- 123 Else
99 Qpdct ¼ MIN (Dpct, CAPdt) 124 //---If product demands are more than inventory, then
100 //---If product demands are small than inventory, then update the demands CAPdt---
update the available CAPdt-- 125 CAPdt ¼ CAPdt – Qpft
101 If Dpct 5 CAPdt 126 Qpft ¼ 0
102 CAPdt ¼ CAPdt – Dpct 127 EndIf
103 Dpct ¼ 0 128 Else
104 Else 129 Break
105 //---If product demands are more than inventory, then 130 EndIf
update the demands Dpct--- 131 EndFor