Design of a Global Decision Support System for a manufacturing SME:

Towards participating in Collaborative Manufacturing

Hao W. Lin
a,n
, Sev V. Nagalingam
b
, Swee S. Kuik
b
, Tomohiro Murata
c
a
Harbin Institute of Technology, Shenzhen Graduate School, Rm. 424, Bld. D, HIT Campus, Xili University Town, Shenzhen, Guangdong 518055, P.R. China
b
School of Advanced Manufacturing and Mechanical Engineering, University of South Australia, Australia
c
Graduate School of Information, Production and Systems, Waseda University, Japan
a r t i c l e i n f o
Article history:
Received 15 February 2007
Accepted 5 July 2011
Available online 22 August 2011
Keywords:
Collaborative Manufacturing
Collaborative networks
Global Decision Support System
System interoperability
Multi-objective optimisation
a b s t r a c t
This paper discusses the conceptual design of a Global Decision Support System for a manufacturing
Small or Medium Enterprise (SM/E), which actively participates in Collaborative Manufacturing. In
order to implement the proposed concept, a Web Services based system architecture is proposed to
offer maximum interoperability between all the distributed participants of a Collaborative Manufactur-
ing Network (CMN) and their management information systems. Furthermore, this conceptual design
utilises a Collaborative decision-support model that effectively interacts with the decision-makers and
the management information systems/tools exist in the network, and provides appropriate support to
all necessary decision-making steps towards the attainment of the networks strategic goals, while
making full benets of the network resources.
& 2011 Elsevier B.V. All rights reserved.
1. Introduction
In recent years, many manufacturing enterprises that are
operating worldwide show an interest for Collaborative Manu-
facturing (CM). This new business strategy offers manufacturers
the critically needed competitive advantages (Camarinha-Matos
and Macedo, 2010; Chung et al., 2004; Johansen et al., 2005).
CM is a concept that involves the establishment of Collaborative
Manufacturing Networks (CMNs) in order to fully exploit the
core competencies of every manufacturer within a network.
The strategy is aiming to achieve best possible fullment of
customer demands and improvement of their overall net prot,
agility, and competitiveness towards the global market (Danilovic
and Winroth, 2005; Kuik et al., 2010). However, CM heavily relies
on improved data, information, and knowledge transparency
typically a commonly recognised decision-making approach to
achieve balanced prots, costs, and risks among the participants
(DAmours et al., 1999; Lagerstrom and Andersson, 2003; Li and
Lai, 2005; Zhang et al., 2004). This reliance suggests that an
integrated manufacturing decision-support infrastructure is
essential for a CMN to successfully deliver the positive outcomes.
Enhancing the existing capabilities on supporting the manage-
ment and production activities are traditionally restricted to
in-house operations and department-oriented operations. Advan-
cing to the CM era, a corresponding new generation of manufac-
turing systems must also expand their features to administrate
collaborative activities between the local enterprise and its
business partners within the CMN (Chiu and Lin, 2004; Cil et al.,
2005; Perrin et al., 2003).
Since, collaborative activities are highly complex and dynamic
(Cil et al., 2005; Perrin et al., 2003; Xu et al., 2009), adequate
interoperability between manufacturing systems that are distrib-
uted across the CMN is essential for the success of this network.
To a certain extent, this interoperability issue is not properly
addressed by most of the conventional integrated manufacturing
systems (Chiu et al., 2006; Lin et al., 2009). Especially these
systems are established by closely coupling computer systems
with inexible interfaces that are hard-coded to accommodate
the purpose of a Business-to-Business (B2B) relationship. Under a
customised interface, these systems full the objectives of infor-
mation sharing, and they proved adequate in sustaining the
automation of most pre-dened business operations. However,
hand-coded interfaces are not readily adaptive to the frequent
changes as experienced within a CMN. As a result, participants
within the network must invest invaluable resources in perform-
ing substantial updates just to maintain the operation of their
existing systems. In order to conform to these integration archi-
tectures, the system must be commonly endorsed by all business
partners to ensure smooth transaction of collaborative manage-
ment activities. Furthermore, the schema of shared information
and knowledge must be updated accordingly whenever the CMN
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/ijpe
Int. J. Production Economics
0925-5273/$ - see front matter & 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2011.07.001
n
Corresponding author. Tel.: 86 755 2603 3148.
E-mail addresses: h.lin@y.kurenai.waseda.jp (H.W. Lin),
sev.nagalingam@unisa.edu.au (S.V. Nagalingam),
swee.kuik@unisa.edu.au (S.S. Kuik), t-murata@waseda.jp (T. Murata).
Int. J. Production Economics 136 (2012) 112
changes its formation. The maintenance cost of these systems is
therefore a critical drawback.
The current manufacturing integration architectures for CM
business activities are facing heavy technical and nancial bur-
dens (Brunnermeier and Martin, 2002; Chiu et al., 2006). These
issues are signicantly apparent for Small and Medium Enter-
prises (SMEs) due to scarce nancial resources and limited
technical abilities. Instead of rapidly expanding the manufactur-
ing capabilities and capacities to cope with the highly dynamic
global market, CM usually constitutes lower risks and is more
nancially justied for manufacturing SMEs to achieve the same
effects of internal expansion (Danilovic and Winroth, 2005;
Johansen et al., 2005; Lin et al., 2005; Loeser, 1999; Nadvi,
1995; Wang et al., 2004; Wheelen and Hungar, 2000). Further-
more, a survey of Australian Manufacturing Industry in 2005
demonstrated the willingness of the organisation to collaborate
with other organisations (Intelligent Manufacturing Systems,
2005). In our work, the development of a Global Decision Support
System (GDSS) enables optimised decision-making via facilitating
interactions amongst the stand-alone manufacturing systems,
and the adoption of a generic collaborative decision-making
model. Subsequently, the GDSS is critical in building up the
willingness for networked collaboration.
This paper is organised as follows: in Section 2, the back-
ground on manufacturing SMEs, CMN, and business process (BP)
modelling are presented. In Section 3, the system architecture for
the GDSS is discussed. In Section 4, a process-based collaborative
decision-support model (CDSM) is proposed. The CDSM depicts
the execution of global decision-making processes within a Small
or Medium Manufacturing Enterprise (S/MME) and its business
partners. Section 5, a real case example is given to illustrate the
conceptual design of the GDSS. Finally, the concluding remarks
and future work are discussed in Section 6.
2. Background
The authors of this paper have performed extensive literature
reviews and on-site business analysis of an enterprises business
activities in association with CM. The outcomes of these studies
demonstrated the need of an improved GDSS for supporting the
current CM business phenomena by more readily attaining
successful decision-making outcomes.
2.1. Critical success factors for manufacturing SMEs
The denition of manufacturing SMEs varies among countries,
but generally the classifying parameters being the number of
employees (under 200 employees in Australia) (Australian Bureau
of Statistics, 2002) and the annual turnover (maximum of
40 million euros in Europe) (Small Business Service, 1996). These
two parameters have dened manufacturing SMEs characteristics
that are considerably different to large enterprises. Huin (2003)
surveyed the characteristics of manufacturing SMEs among
30 companies by conducting 95 interviews with executives, and
identied the key strategic and operational characteristics of an
S/ME. Based on Huins ndings (2003) and our previous work
(Nagalingam and Lin, 2000), it is concluded that there are three
key business objectives that are critical to the success of manu-
facturing SMEs. Firstly, management activities throughout every
organisation unit of the S/ME must be integrated to the max-
imum, so that the workow across all functional and manage-
ment boundaries are better aligned with the strategic goals of the
enterprise. Furthermore, since external factors are highly inuen-
tial to an S/MME, business partners must also participate in
relevant decision-making activities whenever appropriate. Sec-
ondly, SMEs should adopt knowledge management as an essential
activity so that work transition can be accomplished smoothly in
events of task transferring and unexpected staff turnover. Thirdly,
the S/MME must be proactive with decision-making, so that the
organisation is capable of confronting forecast distortions, unex-
pected events, and demanding customers in an effective manner.
2.2. Participating in a Collaborative Manufacturing network
From the perspectives of a manufacturing oriented S/ME, a
CMN is formed when the S/ME establishes highly transparent
collaborative relationships with its business partners, who
include customers, suppliers, and contractors such as illustrated
in Fig. 1. The key objective of the collaborative strategy is to
coordinate all the resources of the S/MME, suppliers, and con-
tractors to best full the demands of the existing customers and
Manufacturing S/ME
(Internal Business Processes)
Supplier 1
Supplier 3
Supplier 2 Supplier y
Contractor 2
Contractor 1
Contractor z
Contractor 3
Analyse, create and
fulfill customer
demands
Procure materials and other
manufacturing supplies from
suppliers that demonstrate
maximum synergy with the
manufacturing S/ME
Outsource production processes to
contractors that demonstrate
maximum synergy with the
manufacturing S/ME
Customer 1
Customer 3
Customer 2
Customer x
Supplier Network
Contractor Network
Customer Network
Collaborative Manufacturing Network
Fig. 1. CMN topology.
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 2
continuously expand the customer network (Johansen et al.,
2005; Kuik et al., 2010).
Due to the globalised market trend, most SMEs no longer have
partnerscustomers conned within the local market, but are dis-
tributed worldwide (Binder and Clegg, 2007). By incorporating the
partners as part of the CMN, the manufacturing SMEs can engage in
more proactive management strategies so that partner and consumer
demands can be better understood, and trading issues can be made
more apparent (Binder and Clegg, 2007; Chung et al., 2004). With
more attentions given to customers (and consumers), manufacturers
can substantially increase business opportunities by attracting new
customers and also retaining the existing ones. The S/ME-to-con-
sumer relationship is classied as downstream or forward vertical
collaboration in the CMN topology (Lin et al., 2005). As for the S/ME-
to-supplier relationship, the key role of suppliers is to supply raw
materials or other Original Equipment Manufacturer (OEM) goods to
the S/MME. All services must satisfy the requirements set out by the
S/ME such as quality assurance, delivery time reliability, and costs in
order to attract business opportunities. This type of the relationship is
classied as upstream or backward vertical collaboration in CMN
topology (Lin et al., 2005). Considering the roles of contractors, they
are generally to provide extended manufacturing and other suppor-
tive services to the S/ME. Examples of organisations that provide the
services include product packaging, Information Technology (IT),
logistic, legal consulting, and others. This type of S/ME-to-contractor
relationship is classied as a horizontal collaboration in the CMN
topology (Lin et al., 2005). All outsourced processes must not be any
of the SMEs core competencies, as they are critical in characterising
the competitiveness of the SME. The actual driving force behind
outsourcing is to maintain a lean enterprise in order to avoid the risks
involved in expanding capabilities that are new to the S/ME.
2.3. Business process management modelling
In recent years, the concept of business process management
(BPM) modelling is becoming an increasingly popular manufac-
turing modelling technique. This technique provides managers
with the visibility, exibility, and agility needed to manage their
businesses (Butler et al., 2002). In BPM modelling, the formation
of BPs is executed at the enterprise level, so that every stake-
holder is aware of the relationships between the processes, and is
clear about their roles (Perrin et al., 2003). The BP model has a
typical hierarchical structure, where each entity is empowered to
execute decisions made by the local experts or managers. As a
rule, the decision outputs of upper processes would always dene
the constraints or the desired targets for their lower processes,
and the decision outputs of the lower processes would express their
commitments towards the upper processes (Aguilar-Save n, 2004). In
our research, BPM refers to the administration of complex relation-
ships between all the BPs in a CMN. BPM guides every available
activity in doing their parts to satisfy customer demands at minimal
resource utilisation. With this management model, the performance
of each process is transparent to the top management level, which
implies business performance can be evaluated more extensively and
that manufacturing bottlenecks can be easily identied. A key concept
of BPM is to empower, to operate and optimise their own processes,
while satisfying the strategic goals set by the senior managers and
constraints presented by all stakeholders of the CMN (Butler et al.,
2002). This decentralised management strategy veries that the BPM
concept can be adapted to establish the suitable management
technique for a CMN.
Extending from the customer oriented business model sug-
gested by McCormack and Rauseo (2005), the CMN oriented
business model examined in our study is shown in Fig. 2. The
model depicts the hierarchical relationships between the business
collaboration lifecycle of the CMN, the core-BPs of each member
of the CMN, and the key supporting BPs whose roles are to ensure
the fullment of core business activities.
3. System architecture
The system architecture of GDSS is made up of the hardware
infrastructure and software packages. Firstly, an information
framework is suggested based on a rationale that constitutes a
suitable information interoperability foundation for the GDSS.
Then the architectural design of the GDSS is introduced in detail.
3.1. Information framework for GDSS
The GDSS must adopt a widely accepted interoperable infor-
mation framework in the sense that those distributed systems
belonging to different members of the CMN can be integrated
seamlessly. A platform is provided to enable collaborative
Business
collaboration
life cycle
Core
business
process
Supporting
business
process
Search for suitable
collaborative business
partners
Evaluate, define, and
establish collaborative
processes
Execute collaborative
processes in accordance to
business requirements
Analyse, maintain, and
update collaborative
processes
Product
research and
development
Marketing
Business
relationship
management
Sales
management
Financial
management
Production
planning
Quality
assurance and
business
partner support
Local
shop-floor
scheduling
Capacity
planning
Inventory
planning
Production
monitoring
Logistics Order picking
Purchasing Outsourcing
Business
opportunity
creation
Business
opportunity
conversion
Order
fulfillment
Support &
Service
Fig. 2. CMN oriented business model for manufacturing SMEs.
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 3
decision-making without confronting the issues related to legacy
systems. The rationale suggested by Chiu et al. (2006) is used to
select an interoperability framework for implementing the GDSS
that facilitates CM operations:
Business connectivity: the framework should ensure afford-
able connections with CMN participants systems for rapid
implementation and ease-of-use capabilities.
Business benets: the framework should facilitate coherence of
knowledge management across all members of a CMN and seek
to achieve convergence through voluntary compliance on mini-
mising the business impact while maximising the benets.
Financial considerations: the framework should provide a
technology roadmap for choosing and implementing reliable,
scalable, and secure systems that are able to meet the rate of
technological changes.
E-government standards: existing government standards
should be adopted wherever they are available and
appropriate.
Forward outlook: the framework should grow in parallel with
the technological developments and establish a process to
grow the framework.
Usage limitations: the framework should provide a high level
or minimum basis for interoperability within the CMN without
any imposed standards, but rather collects and reects stan-
dards in use.
Product support: the framework should provide exibility in
the selection of vendors, preferably being vendor and product
neutral.
Technical support or staff experience: ideal selection of an
appropriate framework depends largely on the technical sup-
port or staff experience available however, these factors
impact heavily on the SMEs in reality.
Three interoperability framework architectures, namely: Elec-
tronic Data Interchange (EDI) (Hartley, 1993), Common Object
Request Broker Architecture (CORBA) (Vinoski, 2004), and Web
Services Architecture (WSA) (Cabrera et al., 2004) have been
reviewed by Chiu et al. (2006). The performance of the architec-
tures are analysed based on the above rationales. As an outcome
of this analysis, WSA is considered as the most appropriate
interoperability framework for the development of the GDSS.
Web Services (WS) through ease of integration, exibility, and
support of Extensible Markup Language (XML) and HyperText
Transfer Protocol (HTTP) provide a suitable interoperability frame-
work to realise the GDSS for a CMN. In terms of business connectivity,
a CMN can reach vast number of business partners through the
Internet. The XML transformation and other associated technologies
facilitate system level integration of services and business partners.
The technical standards of WS are widely endorsed by the software
vendors, such as Sun Microsystems Java 2 PlatformEnterprise Edition
(J2EE) and Microsoft .Net framework, and governments (Australian
Government Information Management Ofce, 2003) worldwide.
Standards-based systems are the choice of manufacturing SMEs that
desire to move to the future rather than reinvent the past. Every
application and computerhuman interaction can be modelled as a
WS to facilitate collaborative knowledge management across mem-
bers of a CMN. This high degree of system exibility helps minimise
the cost of switching among alternative requirements or partners, and
enables the indispensable features of CMN that include indepen-
dence, equality of the partners, and uid boundaries.
3.2. GDSS architectural design
This section introduces the architectural design, which is
depicted in Fig. 3, towards the implementation of the GDSS. As
shown in Fig. 3, the system architecture design consists of
4 modules: (i) the database and knowledgebase, (ii) the web-
based client application, (iii) the collection of decision-support
WS that enables the functionalities of the GDSS, and (iv) the
softwares-to-WS interfaces that enable standalone information
systems to interoperate with the GDSS.
3.2.1. Database and knowledgebase module
The database and knowledgebase module focuses on the
designing of a relational database that is capable of storing loosely
structured decision-making data, information, and knowledge in a
highly systematic table-based information system environment.
In our research project, Microsoft SQL 2000 database server
(Rankins et al., 2003) has been selected to implement the
database. It is highly scalable, reliable, and exible to maintain
due to its functional rich user interface for the target audience,
the SMEs. Unlike transactional systems, an ideal decision support
system requires data and knowledge to be organised in a
simplied structure that maximise the efciency of analytical
queries. Furthermore, the timely update of operational data and
information has a similar importance in maximising the accuracy
of quality of a decision. It is essential to ensure that at the point of
decision analysis, all relevant decision parameters have been
strategically maintained (either manually or pre-programmed)
in order to establish a reliable representation of the actual
problem.
For our study, knowledge can be summarised as a depository
of know-hows on the interpretation and reasoning of decision-
attributes towards achieving the objectives of a decision. Knowl-
edge can be gained through educations, learning, and experience,
and it can be classied as tacit (knowledge exist in the state of
decision-makers minds) or explicit (knowledge that have been
formally documented). One of the critical challenges in our study
is to assist a CMN in the conversion of tacit knowledge to explicit
knowledge so that it can be formally stored in the knowledgebase.
For the GDSS, it can be generally summarised that four types of
knowledge are stored in the knowledgebase such as (i) business
process denition, (ii) business rules and policies, (iii) operation
know-hows, and (iv) historical experience (Lin et al., 2005, 2009).
3.2.2. Web-based client application
The web-based client application is implemented using Micro-
softs ASP .Net programming language in our research project.
This application allows decision-makers to interact with the GDSS
with minimum equipment requirements and highest exibility.
Most web browsers of different vendors are capable of displaying
the interface and supporting all functionalities. In this project,
however, Microsoft Internet Explorer version 6 and later can best
support the dynamic features and maintain the correct displaying
format of the interface.
In GDSS, the key roles of the client application is to display
concise information in relation to the decision and to guide
decision-makers through a pre-dened decision-making work-
ows as described by the CDSM introduced in Section 4. Using the
Web Services paradigm, all related information and knowledge
are made readily accessible to the decision-maker throughout
every step of the decision-making sequence. A template of the
web-based client application is depicted in Fig. 4.
3.2.3. Decision-support WS
The collection of WS can be further categorised into ve
groups, with each serving a unique feature of the GDSS. These
groups are: (i) BPM Services, (ii) Knowledge Management
Services, (iii) Decision Analysis Model Management Services,
(iv) Information Management Services, and (v) Computer-based
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 4
Fig. 3. Architectural design for GDSS.
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 5
Delphi Process Services (Lin et al., 2005, 2009). These services
merely constitute the basic requirements of the GDSS. Due to the
scalability of the WSA, new services can be created indepen-
dently, and seamlessly interoperate with the existing services,
thus expanding the functionality of the GDSS. Each decision-
support WS is designed to address a specic feature or function
of the GDSS. Being implemented as a WS, the web-based client
application could dynamically and rapidly bind with these
features in any sequence, to create a highly intelligent decision-
support package that best address the requirements for decision-
making in a CMN. Furthermore, any of the features may be
selected to interoperate with the business partners WS-enabled
system using the existing network infrastructure that is enabling
the Internet, and consequently expanding the capabilities for the
local system. Overall, a highly complex system can be achieved in
the CMN by using these functionally focused WS features.
3.2.4. Softwares-to-WS interfaces
Often, decision-support WS are required to extract information
and learn new knowledge from standalone or dedicated information
systems such ERP, SCM, Management Information System (MIS),
and Customer Relationship Management (CRM). Other times, they
are required to interact with decision analysis tools and models such
as mathematical optimisation solver software applications. In these
situations, an intermediate interface must be created to enable
interoperability between the GDSS and the standalone information
systems (Chiu et al., 2006; Lin et al., 2005). For example, Visual Basic
for Application (VBA) programme can be written to inter-connect an
Excel-based decision model with the GDSS via a particular Decision
Analysis Model Management Service.
4. Collaborative decision-support model
This section focuses on orchestrating all elements of the entire
collaborative decision-making progresses throughout a CMN.
Our research project has lead to the development of a CDSM,
which provides a systematic way of supporting rational decision-
making in SMEs using quantitative methods.
4.1. Decision-making concepts
In CM, the essence of decision-making is to identify the unique
objectives, capabilities, constraints, and commitments of all
functional units of the CMN toward a manufacturing process,
and decide on how to optimally dispose the overall available
resources to full the business objectives. Under such business
phenomena, decision-making is highly complex due to large
number of distributed but interrelated manufacturing variables,
as well as the large number of alternatives for achieving the
objectives. A systematic decision-making process is required to
Fig. 4. Web-based client application interface.
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 6
address this overwhelming complexity. Simon (Marakas, 1999)
proposed a four phase decision-making model (Turban et al.,
2005) that is the most concise and yet completes characterisation
of a rational decision-making approach. The four decision-making
phases as suggested by Simon (Turban et al., 2005) are namely
intelligence, design, choice, and implementation. The intelligence
phase is used to simplify and make knowledgeable assumpt-
ions about the real world problem, so that decision-makers could
comprehend the situations and correctly dene the potential
problems and or opportunities. The design phase involves the
selection of an appropriate model to analyse the decision and
thus nding the potential decision alternatives for the most likely
Scan for environmental
parameters and performance
indicators that deviate from
their desired values
Identify problems and/or
opportunities and construct
formal decision statements
Select appropriate decision
alternatives and scenario, and
forward the information to the
associated functional units
Analyse decision using local
models, and formulate goal
function(s) based on the
decision alternatives
Analyse decision using local
models, and formulate goal
function(s) based on the
decision alternatives
Analyse decision using local
models, and formulate goal
function(s) based on the
decision alternatives
Collect goal functions from the
functional units and ensure the
functions are valid
Solve the meta-goal
programming model
Outcome satisfied?
Implementation of
the decision outcome
No
Yes
Define meta-goals for all goal
functions under consideration
Activate computer-based
Delphi process and refine
previous phases as required
1 2
3
4.1 4.2 4.n
5 6
7
8
9 10
Decision-maker(s) approve(s)
meta-goals and establish
priority for each meta-goal
Group-based
interactive process?
No
Yes
Collect performance
feedbacks on the
implementation of decisions
11
Fig. 6. Collaborative decision-support model.
Intelligence Design
Choice Outcome
* Generally the starting point of a decision-making process
Implementation
Failure
Success
Solution testing
Model validation
Problem
statement
Alternatives
Solution
Simplification
and
assumptions
* Identify objectives,
capabilities, constraints,
and commitments of all
functional units of the
CMN
Fig. 5. Decision-making model for Collaborative Manufacturing (adapted from Small Business Service, 1996).
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 7
scenario. The choice phase focuses on using algorithms to solve
the model and nd the solutions from the decision alternatives.
The implementation phase veries the performance of the solu-
tions obtained. This model is conceptually illustrated in Fig. 5.
4.2. Collaborative decision-support modelling
CDSM is usually in a continuous loop. However, the scanning
for business performance parameter deviations is usually con-
sidered as the rst step of any decision-making process. During
any stage of the decision-making process, the decision-makers if
necessary can retrace to any of the previous stages to rene their
results. By projecting CDSM to Simons model (Turban et al.,
2005) of decision-making, Steps from 1 to 2 are categorised as the
intelligence phase, Steps from 3 to 4 belong to the design phase,
Steps from 5 to 9 are in the choice phase, and nally Step 10 is the
implementation phase. The CDSM is depicted in Fig. 6.
Conforming to the BPM concept, the CDSM assumes local
empowerment to each entity of the CMN. A critical challenge
however, is to best orchestrate local decisions toward the attain-
ment of strategic goals for the entire CMN. Stepping up to the
challenge, the CDSM employs Interactive Meta-Goal Program-
ming (IMGP) technique and Computerised Delphi Process to
respectively, enlighten issues associated with human-to-model
interaction and conicting of interests between entities. IMGP is a
mathematical optimisation technique proposed by Caballero et al.
(2006). Building on the fundamental ideas of Goal Programming,
IMGP offers a multi-level modelling structure that is highly
suitable for representing the hierarchical structured organisa-
tional decision problems in a CMN environment. The CDSM
suggests that each entity is empowered to build Goal Program-
ming models to express its local decision problem. The local
models are then collated and that meta-goals are designed to
dictate the level of attainment for each original goal function with
respect to the high level strategic objectives. Meta-goals are goals
of the original goal functions, and they provide a concise and yet
consistent manner to evaluate the quality of sub-decisions with
respect to the performance of the entire CMN. By adjusting the
meta-goal objectives, decision-makers can effectively interact
with the model and conduct what-if analysis to discover
optimal overall decision. Also, meta-goals are commensurable
performance parameters between all entities of the CMN, which
conveniently form the common reference point for multiple
decision-makers in resolving conicting interests during Delphi
Processes. Similar to what is depicted by Turoff and Hiltz (1995),
the CDSM employs a computerised Delphi Process to establish
consensus decisions amongst all entities of the CMN. It is
expected that each entity is interested in optimise its local
performance, and that its respective contribution towards the
overall CMN performance is represented by the commensurable
meta-goals. In group decision-making, various congurations are
proposed by each entity, and that any conicts must be resolved
in a manner that guarantees the aspiration of the CMN as a whole.
Under the Delphi Process, conicts are resolved over numerous
rounds, with each round trying to anonymously establish a
consensus for disputing factor that contribute the most variation
towards the CMN performance. Consequently, the IMGP and
Delphi Process enabled CDSM can effectively facilitate group
decision-making in a CMN, and this concept is demonstrated
using a case example in Section 5.
5. Research methodology and discussion
This section illustrates an example on a decision-making process
that can be constructed and applied in a practical situation within
the GDSS platform. The results of the example is analysed, and the
performance of the conceptual design for the GDSS are discussed.
5.1. Background of a small manufacturing enterprise
In this section, the decision-making process for a production-
planning task of a small manufacturing enterprise that manufac-
tures medical accessories is illustrated. Due to business conden-
tiality, most of the data used in this example are hypothetical. In
this example, we considered that three different functional units
participated in a decision-making process: the Sales and Fore-
casting unit that set goals to customer demands and business
prots; the operational planning unit that set goals to the utilities
of the available production capacity; the scheduling unit that set
goals to the utilities of the current machine group formation.
5.2. Consolidate information and data
The problem considered is in this example is to decide the
production quantity for 10 different types of suture for the following
Table 1
Technology coefcients obtained from different functional units.
Suture
type 1
Suture
type 2
Suture
type 3
Suture
type 4
Suture
type 5
Suture
type 6
Suture
type 7
Suture
type 8
Suture
type 9
Suture
type 10
Production quota (box of dozen) 200 150 180 130 230 50 90 85 130 255
Expected prot ($/box) 20 22 25 30 15 28 18 23 22 26
Manufacturing cost ($/box) 25 26 28 30 20 25 22 20 24 28
Production time on swaging machine
group 1 (h)
0.1 0.2 0.1 0.2
Production time on swaging machine
group 2 (h)
0.1 0.1 0.15
Production time on swaging machine
group 3 (h)
0.15 0.2 0.15
Packaging time (h) 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04
Quality cheque time (h) 0.025 0.025 0.025 0.025 0.025 0.025 0.025 0.025 0.025 0.025
Needle type 1 consumption (unit) 1 1 1 1
Needle type 2 consumption (unit) 1 1 1
Needle type 3 consumption (unit) 1 1 1
Thread type 1 consumption (cm) 20 20 20
Thread type 2 consumption (cm) 20 20
Thread type 3 consumption (cm) 20 20 20
Thread type 4 consumption (cm) 20 20
: empty elds imply technology coefcients not applicable for the corresponding suture type.
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 8
week. The suture production process consists of three major steps,
which are swaging (joining the thread to the needle), packaging, and
quality checking. The production lead-time for different types of
suture varies, as ner needles are more difcult to handle than the
coarse ones. Since there are three different types of needles available
(with different diameters), the 10 available swaging machines are
divided into three groups, where machines of the same group are set
up to produce a particular needle size. Although, any machine can set
up to produce other needle sizes, it is not desirable by the production
workers, as setting up is a time consuming process, and the slightest
miss-set-up would destroy the costly die in the swaging mechanism.
The technology coefcients for the functional units are summarised in
Table 1. The constraints to this problem are the available stock of the
needles and the threads, which are supplied by the inventory
management unit. These constraints are summarised in Table 2.
The multi-objective problem is then modelled by the following Goal
Programming equations, and the model is analysed using the algo-
rithm proposed by Caballero et al. (2006). The model is solved by
WhatsBest!, a Microsoft Excel plug-in optimisation tool developed by
Lindo Systems Inc. (Lindo Systems Inc., 2006).
Parameters
i Types of suture considered in production planning
Decision variables/alternatives
x x
1
,x
2
,:::x
i
(Production quantity for suture type i, i 1,:::,10)
Goals proposed by the sales and forecasting unit
x
1
d

1
d

1
200 (Production quantity for suture type 1)
x
2
d

2
d

2
150 (Production quantity for suture type 2)
x
3
d

3
d

3
180 (Production quantity for suture type 3)
x
4
d

4
d

4
130 (Production quantity for suture type 4)
x
5
d

5
d

5
230 (Production quantity for suture type 5)
x
6
d

6
d

6
50 (Production quantity for suture type 6)
x
7
d

7
d

7
90 (Production quantity for suture type 7)
x
8
d

8
d

8
85 (Production quantity for suture type 8)
x
9
d

9
d

9
130 (Production quantity for suture type 9)
x
10
d

10
d

10
255 (Production quantity for suture type 10)
20x
1
22x
2
25x
3
30x
4
15x
5
28x
6
18x
7
23x
8
22x
9
26x
10
d

11
d

11
30,000
(Prot target)
Goals proposed by the operational planning unit
25x
1
26x
2
28x
3
30x
4
20x
5
25x
6
22x
7
20x
8
24x
9
28x
10
d

12
d

12
50,000
(Budget utility)
0:1x
1
0:1x
2
0:15x
3
0:2x
4
0:1x
5
0:2x
6
0:1x
7
0:15x
8
0:15x
9
0:2x
10
d

13
d

13
350
(Swaging utility)
0:04x
1
x
2
x
3
x
4
x
5
x
6
x
7
x
8
x
9
x
10
d

14
d

14
70 (Packaging utility)
0:025x
1
x
2
x
3
x
4
x
5
x
6
x
7
x
8
x
9
x
10
d

15
d

15
35 (Quality cheque utility)
Goals proposed by the scheduling unit
0:1x
1
0:2x
4
0:1x
7
0:2x
10
d

16
d

16
175 (Swaging machine group 1 capacity)
0:1x
2
0:1x
5
0:15x
8
d

17
d

17
105 (Swaging machine group 2 capacity)
0:15x
3
0:2x
6
0:15x
9
d

18
d

18
70 (Swaging machine group 3 capacity)
Constraints
20x
1
22x
2
25x
3
30x
4
15x
5
28x
6
18x
7
23x
8
22x
9
26x
10
Z10,000
(Prot-break-even point)
x
1
x
4
x
7
x
10
r650 (Needle type 1 constraint)
x
2
x
5
x
8
r500 (Needle type 2 constraint)
x
3
x
6
x
9
r350 (Needle type 3 constraint)
20x
1
x
2
x
3
r10,000 (Thread type 1 constraint)
20x
4
x
6
r3500 (Thread type 2 constraint)
20x
7
x
8
x
9
r6500 (Thread type 3 constraint)
20x
5
x
10
r10,000 (Thread type 4 constraint)
In this model, the undesired deviations for the functional goals are
d

1
,d

2
,d

3
,d

4
,d

5
,d

6
,d

7
,d

8
,d

9
,d

10
,d

11
,d

12
,d

13
,d

14
,d

15
,d

16
,d

17
,d

18
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 9
According to the IMGP algorithm, rstly the payoff matrix
must be calculated to determine the boundaries of the solution
when the meta-goals are optimised one at a time. The payoff
matrix result is summarised in Table 3. The decision-makers use
this solution as a guide to prioritise the existing goal functions
and evaluate the tradeoffs between the outputs obtained from
different priority schemes. Realistically, the decision variables
denote quantity, and thus should be calculated in terms of
integers. To simply the evaluation process however, non-integer
values are used, and the nal results are rounded off to the
nearest integers.
Based on the payoff matrix solution, the decision-makers can
analyse and compare different decision outcomes by suggesting
their preference on the degree of meta-goal achievements.
As an example, if a decision-maker has assigned all of the goal
functions to the same priority level, and suggested the meta-goal
achievements as listed in points below, the solution of this meta-
goal model is summarised in Table 4
i. Aggregate undesired deviations of all the goal functions
cannot be more than 0.3.
ii. Maximum undesired deviations of all the goal functions
cannot be more than 0.07.
iii. Number of unattained goals cannot be more than 5.
Table 2
Goal targets and hard constraints.
Equation type Description Target/constraint value Undesired deviations
Goals Production target for suture type 1 200 (box of dozen) Slack
Production target for suture type 2 150 (box of dozen) Slack
Production target for suture type 3 180 (box of dozen) Slack
Production target for suture type 4 130 (box of dozen) Slack
Production target for suture type 5 230 (box of dozen) Slack
Production target for suture type 6 50 (box of dozen) Slack
Production target for suture type 7 90 (box of dozen) Slack
Production target for suture type 8 85 (box of dozen) Slack
Production target for suture type 9 130 (box of dozen) Slack
Production target for suture type 10 255 (box of dozen) Slack
Prot achievable from the current production plan 30,000 ($) Slack
Budget utilisation for the current production plan 50,000 ($) Surplus
Total swaging capacity for the current production plan 350 (h) Surplus
Packaging capacity for the current production plan 70 (h) Surplus
Quality cheque capacity for the current production plan 35 (h) Surplus
Group 1 swaging machine capacity 175 (h) Surplus
Group 2 swaging machine capacity 105 (h) Surplus
Group 3 swaging machine capacity 70 (h) Surplus
Constraints Prot-break-even point that the manufacturer must achieve 10,000 ($)
Stock availability for needle type 1 650 (unit)
Stock availability for needle type 2 500 (unit)
Stock availability for needle type 3 350 (unit)
Stock availability for thread type 1 10,000 (cm)
Stock availability for thread type 2 3500 (cm)
Stock availability for thread type 3 6500 (cm)
Stock availability for thread type 4 10,000 (cm)
: empty elds imply choice of slack or surplus deviation is not applicable, as hard constraint equations must be strictly satised.
Table 3
Meta-goal payoff matrix solution.
Objective Achieved score Decision-
makers
preference
Aggregate
undesired
deviations
Maximum
undesired
deviation
Number of
unattained
goals
Aggregate
undesired
deviations
0.24 0.1 4 0.3
Maximum
undesired
deviation
0.471 0.057 9 0.07
Number of
unattained
goals
0.645 0.323 2 5
Table 4
Solution for single priority level model.
Goal functions Value
Meta-goals
Aggregate undesired deviations 0.3
Maximum deviation 0.076
Number of unattained goals 5
Goals
Production quantity for suture type 1 185 (box of dozen)
Production quantity for suture type 2 142 (box of dozen)
Production quantity for suture type 3 170 (box of dozen)
Production quantity for suture type 4 120 (box of dozen)
Production quantity for suture type 5 230 (Box of Dozen)
Production quantity for suture type 6 50 (box of dozen)
Production quantity for suture type 7 90 (box of dozen)
Production quantity for suture type 8 85 (box of dozen)
Production quantity for suture type 9 130 (box of dozen)
Production quantity for suture type 10 255 (box of dozen)
Prot achievable 32,594.65 ($)
Budget utility 36,472.64 ($)
Swaging utility 207.48 (h)
Packaging utility 58.29 (h)
Quality cheque utility 36.43 (h)
Machine group 1 capacity 102.52 (h)
Machine group 2 capacity 49.97 (h)
Machine group 3 capacity 55 (h)
Underlined goals imply targets not met.
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 10
So far in this example, the analysis focused only on the goal
priority preference set by a single decision-maker. To incorporate
multiple preferences, the CDSM suggests that Delphi Process should
be activated to deduce a consensus priority preference scheme that
can be used towards the IMGP analysis. Furthermore, through Delphi
Process, the goal attainment preferences proposed by different
functional units can be updated accordingly in order to achieve better
decision outcomes. For example, the analysis outcome of the example
presented here indicates that the quality checking capacity is the
bottleneck of the manufacturing process, which negatively affects the
fullment of production quota. To improve this situation, the quality-
checking unit is to be consulted on the ability to increase its output
capacity. Once an updated quality-checking capacity is obtained, the
corresponding payoff matrix of the new model can be calculated and
that meta-goal targets can be adjusted accordingly. In this approach,
the Meta-Goal Programming process and Delphi Process are iterated
until a solution that satises all associated decision-makers is found.
Below, we describe a possible rened meta-goal analysis model
that follows the Delphi Process. The goal functions are allocated into
four priority levels, and that goal functions per priority level are
analysed by a unique meta-goal such as described in the following
points. This IMGP model can be easily evaluated using WhatsBest!
Plug-in, but the solution is not included here.
Priority level 1: Finance unit demands, the prot and
budget goals are to be in this priority.
These goals are also closely monitored and
supported by the enterprises high-level
management.
Priority level 2: The target dened by the sales and
marketing unit is considered as the second
priority to ensure customer satisfaction.
Priority level 3: Production, packaging, and quality goals
are in this priority level.
Priority level 4: The production-scheduling unit and
machine group formation that minimises
machine re-setup are placed in the lowest
priority while satisfying the meta-goals of
higher priority levels.
5.3. Results, outcomes, and performance
The feasibility of the conceptual design for the GDSS was
analysed based on the results of our study. Our analysis have
identied that our CDSM has the following strengths:
i. Data, information, and knowledge are shared via the WSA
based interoperability framework, which is highly efcient in
establishing system integrations on the application-to-
application level.
ii. Data, information, and knowledge can be acquired from
stand-alone systems within the CMN at timely manner, thus
decision-makers are well supported towards making informed
decisions.
iii. Decision-making tools can be developed individually and
incorporated to the GDSS to further enhance the capabilities
of the CDSM.
iv. Knowledge coherence is achieved, since the business objec-
tives of the CMN are transparent to all participants of the
network.
v. The CDSM ensures that all decision-making activities are
appropriately orchestrated and synchronised toward the
achievement of optimal decisions.
vi. The GDSS is highly adaptable as new data, information, and
knowledge can be incorporated into the system at any time
to support new management requirements.
vii. The local desired goals are collected and analysed by the
IMGP model, and it ensures that the actual nal decisions can
be practically fullled across all functional units under the
current manufacturing environment, and that the decision is
close to global optimal as much as possible.
viii. The meta-goal approach enables decision-makers to ef-
ciently analyse and compare alternate solutions, and identify
which manufacturing resources may be adjusted in order to
improve the overall outcome.
ix. The Delphi Method provides a mediating environment for
decision-makers to quickly identify issues that prevent
decision consensus, and hence management resources could
be focused on eliminating those issues.
6. Conclusion
This paper analysed current business environment within a
CMN setting in considering for the perspectives of a participating
S/MME, and argued that a GDSS is necessary to enable the
manufacturing oriented S/ME in near optimal decision-making
within the CMN. A conceptual design of the GDSS is provided in
this paper as a key contribution. In our approach, rstly, Microsoft
.Net WSA is selected to design the system architecture for the
GDSS. Fundamentally established on the existing Internet proto-
cols and other cross-platform standards, the WSA delivers max-
imum information system interoperability on all levels ranging
from a standalone simply application to a suite of software.
Furthermore, the WSA-based GDSS design ensures that maximum
system scalability and re-useability are readily achieved.
Secondly, a CDSM is proposed to guide the process of optimised
decision-making within the CMN. The model enables different
functional units to analyse the decision under consideration
separately and propose goals that are in favour of their respective
functional units performance. These goals are collected and then
analysed using the IMGP approach, which allows decision-makers
to concisely interact during the process of reaching a solution and
quickly converge to a solution acceptable to all the decision-
makers. The conceptual design of the GDSS is justied by a
simulated case study on an Australian manufacturer. In our study,
we have found that the GDSS is capable of integrating the existing
distributed information systems within the local manufacturer
and its CMN. This enables the managers of the local manufacturer
to efciently conduct collaborative decision-making activities in
relation to other participants of the CMN. In addition, the GDSS
enables an S/MME to realise the benets of a CMN, without
extensive alteration to their existing computer network architec-
ture and software systems. The next step of our research project is
to model other decision-making processes in compliance with the
CDSM for the core BPs. Example of these BPs for sustainability in
manufacturing include collaborative product design where design
objectives of individual functional units are considered at the
CMN level, and production order allocation problem where
integer-based Meta-Goal Programming process is used to opti-
mally distribute production orders amongst a group of manufac-
turers. Thus, the GDSS establishes a fundamental platform for
future applications of decision-support approaches that can be
continuously added in supporting the dynamic management
processes for achieving sustainability in manufacturing in a CMN.
Acknowledgement
Authors acknowledge the funding provided by Australian
Research Council for this research project, the support provided
by industry partner, Dynek Pty Ltd., South Australia, Australia,
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 11
and the review comments for this article by Mr. B. Crook,
Managing director of Dynek Pty Ltd.
References
Aguilar-Save n, R.S., 2004. Business process modelling: review and framework.
International Journal of Production Economics 90 (2), 129149.
Australian Bureau of Statistics, 2002. Small Business in Australia. /http://www.
abs.gov.au/ausstats/abs@.nsf/lookupMF/97452F3932F44031CA256C5B00027F
19S (viewed 12.02.05.).
Australian Government Information Management Ofce, 2003. Interoperability
Technical Framework for the Commonwealth Government. /http://www.
agimo.gov.au/publications/2003/08/frameworkS (viewed 12.15.05.).
Binder, M., Clegg, B., 2007. Enterprise management: a new frontier for organisa-
tions. International Journal of Production Economics 106 (2), 409430.
Brunnermeier, S.B., Martin, S.A., 2002. Interoperability costs in the US automotive
supply chain. International Journal of Supply Chain Management 7 (2), 7182.
Butler, M., Clarke, S., Holden, J., Holt, M., Jones, G., Jones, T., Kellett, A., Smith, E.,
2002. Business Process Management: Improving Business EfciencyButler
Direct Limited, UK.
Caballero, R., Ruiz, F., Uria, M.V.R., Romero, C., 2006. Interactive meta-goal
programming. European Journal of Operational Research 175 (1), 135154.
Cabrera, L.F., Kurt, C., Box, D., 2004. An Introduction to the Web Services
Architecture and Its Specications. Microsoft MSDN. /http://msdn.microsoft.
com/library/default.asp?url=/library/en-us/dnwebsrv/html/introwsa.aspS
(12.12.05.).
Camarinha-Matos, L., Macedo, P., 2010. A conceptual model of value systems in
collaborative networks. Journal of Intelligent Manufacturing 21 (3), 287299.
Chiu, M., Lin, G., 2004. Collaborative supply chain planning using the articial
neural network approach. Journal of Manufacturing Technology Management
15 (8), 789796.
Chiu, M., Lin, H.W., Nagalingam, S.V., Lin, G.C.I., 2006. Inter-operability framework
towards virtual integration of SMEs in the manufacturing industry. Interna-
tional Journal of Manufacturing Technology and Management 9 (3/4),
328349.
Chung, W.W.C., Yam, A.Y.K., Chan, M.F.S., 2004. Networked enterprise: a new
business model for global sourcing. International Journal of Production
Economics 87 (3), 267280.
Cil, I., Alpturk, O., Yazgan, H.R., 2005. A new collaborative system framework based
on a multiple perspective approach: InteliTeam. Decision Support Systems 39
(4), 619641.
DAmours, S., Montreuil, B., Lefranc-ois, P., Soumis, F., 1999. Networked manufac-
turing: the impact of information sharing. International Journal of Production
Economics 58 (1), 6379.
Danilovic, M., Winroth, M., 2005. A tentative framework for analyzing integration
in collaborative manufacturing network settings: a case study. Journal of
Engineering and Technology Management 22 (12), 141158.
Hartley, J., 1993. Electronic Data Interchange: Gateway to World-Class Supply
Chain ManagementEconomist Intelligence Unit, London.
Huin, S.F., 2003. Intelligent Resources Planning for Small-and-Medium-Sized
Manufacturers. Ph.D. Thesis. University of South Australia, Adelaide.
Intelligent Manufacturing Systems, 2005. Survey of Australian Manufacturing
Industry. /http://www.au.ims.org/download/IMSAURnDNeedsSurveyResults.
pdfS (viewed 02.6.06.).
Johansen, K., Comstock, M., Winroth, M., 2005. Coordination in collaborative
manufacturing mega-networks: a case study. Journal of Engineering and
Technology Management 22 (3), 226244.
Kuik, S.S., Nagalingam, S.V., Amer, Y., 2010. Challenges in implementing sustain-
able supply chain within a collaborative manufacturing network. In: Proceed-
ings of Eighth International Conference on Supply Chain Management and
Information Systems. Hong Kong.
Lagerstrom, K., Andersson, M., 2003. Creating and sharing knowledge within a
transnational teamthe development of a global business system. Journal of
World Business 38 (2), 8495.
Li, E.Y., Lai, H., 2005. Collaborative work and knowledge management in electronic
business. Decision Support Systems 39 (4), 545547.
Lin, H.W., Nagalingam, S.V., Chiu, M., 2005. Development of a collaborative
decision-making model for a network of manufacturing SMEs. In: Pasquino,
N. (Ed.), Proceedings of 18th International Conference on Production Research.
University of SalernoCampus of Fisciano (SA) Italy, Fischiano (SA) Italy.
CD-Rom Publication.
Lin, H.W., Nagalingam, S.V., Lin, G.C.I., 2009. An interactive meta-goal program-
ming based decision analysis methodology to support collaborative manufac-
turing. Robotics and Computer-Integrated Manufacturing 25, 135154.
Lindo Systems Inc., 2006. WhatsBest! /http://www.lindo.com/S (viewed 04.
28.06.).
Loeser, B.O., 1999. How to set up a cooperation network in the production
industry: example of the HuberSuhner AG. Industrial Marketing Manage-
ment 28 (5), 453465.
Marakas, G.M., 1999. Decision Support Systems in the Twenty-First CenturyPren-
tice-Hall, Inc., USA.
McCormack, K., Rauseo, N., 2005. Building an enterprise process view using
cognitive mapping. Business Process Management Journal 11 (1), 6374.
Nadvi, K., 1995. Industrial Clusters and Networks: Case Studies of SME Growth and
InnovationUnited Nations Industrial Development Organisation.
Nagalingam, S.V., Lin, G.C.I., 2000. A distributed group decision support systems
for virtual computer intergrated manufacturing. In: Proceedings of the Sixth
International Conference on Automation Technology. Taiwan, pp. 377382
(hardcopy).
Perrin, O., Wynen, F., Bitcheva, J., Godart, C., 2003. A model to support collabora-
tive work in virtual enterprises. In: van derAalst, W.M.P. (Ed.), Proceedings of
International Conference of Business Process Management. Springer-Verlag,
Eindhoven, The Netherlands.
Rankins, R., Bertucci, P., Jensen, P., 2003. Microsoft SQL Server 2000 Unleashed,
second ed., Indianapolis.
Small Business Service, 1996. Small and Medium Enterprise Denitions. European
Commission. /http://www.sbs.gov.uk/sbsgov/action/layer?r.s=sl&topicId=700
0000237S (viewed 12.2.05.).
Turban, E., Aronson, J.E., Liang, T.-P., 2005. Decision Support Systems and
Intelligent Systemsseventh ed. Pearson Education, Inc., New Jersey.
Turoff, M., Hiltz, S.R., 1995. Computer based Delphi processes. In: Adler, M., Ziglio,
E. (Eds.), Gazing into the Oracle: The Delphi Method and its Application to
Social Policy and Public HealthJessica Kingsley Publishers, London.
Vinoski, S., 2004. CORBA in a Loosely Coupled World. Object Management Group.
/http://www.looselycoupled.com/opinion/2004/vinos-corba-infr0908.htmlS
(viewed 12.12.05.).
Wang, D., Nagalingam, S.V., Lin, G.C.I., 2004. Development of a parallel processing
multi-agent architecture for a virtual CIM system. International Journal of
Production Research 42 (17), 37653785.
Wheelen, T.L., Hungar, D.J., 2000. Strategic Management and Business Policy:
Entering 21st Century Global Societyseventh ed. Prentic Hall, London.
Xu, H., Koh, L., Parker, D., 2009. Business processes inter-operation for supply
network co-ordination. International Journal of Production Economics 122 (1),
188199.
Zhang, S., Shen, W., Ghenniwa, H., 2004. A review of Internet-based product
information sharing and visualization. Computers in Industry 54 (1), 115.
H.W. Lin et al. / Int. J. Production Economics 136 (2012) 112 12