RESEARCH ON INNOVATION OF SHIPBUILDING SUPPLY
CHAIN MANAGEMENT BASED ON COMPLEXITY
ADAPTIVE SYSTEM
Gen Li, Chunlei Lv and Yifei Cheng
School of Economics and Mnanagement, Harbin Engineering University, Nantong Street, Harbin, China
Keywords: Shipbuilding supply chain management, Complexity adaptive system, Operational mode, Innovation.
Abstract: Because of the change of the whole operational environment, traditional methods could not reasonably be
used to settle the complicated problems which occur in shipbuilding supply chain management. Based on
complexity adaptive system theory, the paper first interprets the whole operational mode of shipbuilding
supply chain, then it establishes simulation model of shipbuilding supply chain according to Multi-Agent,
finally it puts forward some innovation directions in order to resolve complicated problems which emerge
from current shipbuilding supply chain management.
1 INTRODUCTION
Shipbuilding supply chain is a series of value-
increasing activities including original material
purchasing, conveyance, process, manufacturing and
productions arriving in ship owners, in fact it is a
function net chain structure which is filled with
logistics flow, fund flow, service flow, knowledge
flow and information flow thoroughly. Shipbuilding
supply chain management is a whole management
style which integrates supplier, manufacturers,
medium-sellers and buyers through logistics and
information flow (Ren and Zhao, 2008).
Under the premise that it quickly satisfies the
need of final users, shipbuilding supply chain must
have the ability to respond to the change of market
quickly and run perfectly from the whole supply
chain (Li, 2006). In the market environment which is
full of dynamic competition and cooperation
nowadays, the enterprises could belong to different
supply chains in the meantime. The competition
among enterprises actually has already converted
into the competition among different supply chains.
Under the competition situation above, the operation
environment of shipbuilding enterprises becomes
more complicated (Zhang, 2003). Enterprises should
re-examine their own operational modes from a new
aspect. They also should allocate resources and make
strategic decisions from the greater thinking space
and the area space. The paper explains the
complicated phenomenon and resolves complicated
problems of shipbuilding supply chain management
basically with the help of complexity adaptive
system theory.
2 CHARACTERISTICS AND
RELATED MODELS OF CAS
Complex adaptive system was put forward by John
Holland (USA) when the institute of Santa Fe was
10 years old in 1994. The fundamental thought of
CAS is that: because the agents will interact with
themselves and environment, they will change
themselves and also change the environment.
2.1 Fundamental Characteristics of
CAS
The core thought of CAS is “adaption creating
complexity”, it can say that it is a leap in
understanding system motion and evolving
regulation. Its characteristics will be discussed as
follows (Chen, 2001).
First, the agent is an active and live agent. This is
the key differentiation of CAS from other systems.
This characteristic makes it able to effectively be
applicable to economic, social, ecological
complicated system etc. But other methods are more
difficult to deal with these areas.
389
Li G., Lv C. and Cheng Y..
RESEARCH ON INNOVATION OF SHIPBUILDING SUPPLY CHAIN MANAGEMENT BASED ON COMPLEXITY ADAPTIVE SYSTEM.
DOI: 10.5220/0003568803890395
In Proceedings of the 13th International Conference on Enterprise Information Systems (NMI-2011), pages 389-395
ISBN: 978-989-8425-53-9
Copyright
c
 2011 SCITEPRESS (Science and Technology Publications, Lda.)

Second, the mutual influence and interaction
between the agent and environment (including
among the agents) is the main motive for the
evolution of the system. The former systems usually
put the internal attribute of agent into the initial
position, and they don’t emphasize on the interaction
between agent and environment (including among
agents). This makes CAS able to be used into
different areas where the characteristics of agents are
absolutely different, but there are many commons in
the correlation.
Third, it introduces the random factor, so it has a
stronger ability to describe and express.
Fourth, it contacts the macro view and the micro
view organically. It makes the change of agents
become the foundation of the change of the whole
system through the interaction between the agent and
environment (Tan and Dong, 2001).
Because of the above characteristics, CAS has
different functions and values compared with other
system methods
2.2 Main Models and Simulated Tool of
CAS
There are many methods to study CAS, but it has
more guiding significance from models and
simulated tool (Dong, 2006).
2.2.1 Stimulating-reaction Model
From the micro view level, it expresses the most
basic behaviour pattern of agent in each system with
the way of unity. The performance system of each
agent is composed of three parts such as a detector,
an effecter, a set of IF/THEN rule.
The detector is been used to accept external
incitement, the effecter is been used to make a
reaction, IF|THEN rule stipulates how to respond
towards different incitements, but it differs from
common sense which is 1 to 1 correspondence. In
order to simulate real world, this rule sets up many
kinds of choices and it contains contradiction,
conflict and disagreement. Based on genetic
algorithm, the agent can successfully compare and
choose rules, and it even yields new rules through
the procession of evolution. Figure 1 is stimulating-
reaction model as follows.
2.2.2 Echo Model
Based on micro stimulating-reaction model, it can
establish echo model which is macro mode of the
whole system. In this model, the agent has three
elementary parts: offensive label, defense label and
resource pool.
Adaptive agent
message message
detector effecter
a set of rule
environment
message message
Figure 1: Stimulating-reaction model.
The basic function of echo mode includes the
performance that an agent actively contacts with
other agents. In the meantime, it gives answers to
other agents towards their contacts, if they match
successfully, then they will exchange resources.
They will save and process resources in their inner
bodies; if the resources are enough and then they
will breed new agents. Under the control of echo
model, the system will be like figure 2 as follows.
Position1
Position2
position 3
Position
Agent2
Agent1
Agent3
Reso
-urce
p
ool
Offensive label
Defense label
Reso
-urce
Figure 2: Echo model.
2.2.3 Simulation Tool
Swarm is a simulated terrace. It is an experiment tool
to study CAS (Qiu, 2003). This simulated terrace is
used to carry out discrete event simulation. It has the
same characteristics as the general affairs simulation.
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
390

It adopts realizing method which is facing to object,
so in this terrace, it provides a whole set of database
which is the basic work what the researcher
completes in some simulated experiments. The
customer can realize many functions on the
foundation of these databases.
The majority of simulation procedures of Swarm
contain four types of objects: Model Swarm,
Observer Swarm, simulation agent and environment.
Model Swarm and Observer Swarm are subcategory
of Swarm category. Swarm category is the basic
structure piece of Swarm simulation. A Swarm is the
combination between a series of objects and
behaviour timetable that are caused by the objects.
Simulation agent usually inherits methods from
Swarm Object to provide support for memory and
management of speculation control. The different
simulation systems have different environment,
figure 3 is Swarm structure.
Model
Swar
m
Observer
Swar
m
Schedule Schedule
Space
Graph
Probe
Agent
Figure 3: The structure of Swarm.
3 ANALYSIS OF COMPLEXITY
AND ADAPTABILITY ON
SHIPBUILDING SUPPLY
CHAIN
3.1 The Source of Complexity of
Shipbuilding Supply Chain System
3.1.1 The INDEPENDENCE of Entities
Shipbuilding supply chain is network system that is
constituted by many comparatively independent
entities. Overall supply chain model and
optimization target don not exist. The management
of overall supply chain will be realized through
coordination among these comparatively
independent entities. Each enterprise in supply chain
has its own benefit and management mode and there
is no core place among these members. The change
of one enterprise will be affected by other enterprises
and this change will affect the change of other
enterprises. The whole shipbuilding supply chain can
constantly study, reorganize and perfect its layer
structure and functional structure (Choi and Dooley,
2003).
3.1.2 Turbulence and Indetermination of
Environment
Shipbuilding supply chain faces inside and external
environment that is full of turbulence and
indetermination. Ally and rival will change all the
time, and itself of shipbuilding supply chain also
adjusts in order to adapt to environment. Under this
circumstance, it is very important for the members in
shipbuilding supply chain to choose suitable
suppliers and dealers, join in the suitable supply
chain and valuate the efficiency and effect of supply
chain. But among the current researches and
applications towards shipbuilding supply chain, the
authors just see the supply chain and its environment
as a static state or very slowly changing object. They
usually adopt serial plan or control mode and over-
emphasize the stability of the colleague relationship,
so it lacks of enough adaptability and is very
difficult to adapt to the operation environment which
is changing rapidly.
3.1.3 Difference of Internal Structure
The pursuit of shipbuilding supply chain is
unobstruction and agility of logistics, information
flow, knowledge flow and capital flow. Because of
the different internal structure of shipbuilding supply
chain enterprise, the corresponding information
systems are not compatible. Shipbuilding supply
chain management is to integrate these distributed
and heterogeneous systems as a whole or
information sharing system, which is urgent problem
that needs to be resolved. But the current system is
mainly based on the traditional organization
management mode and do not really realize business
process integration from the scope of the entire
supply chain.
RESEARCH ON INNOVATION OF SHIPBUILDING SUPPLY CHAIN MANAGEMENT BASED ON COMPLEXITY
ADAPTIVE SYSTEM
391

3.2 Mechanism of Complexity of
Shipbuilding Supply Chain
3.2.1 Interaction and the Occurrence of
Emerging of Shipbuilding Enterprises
in the Supply Chain
The essence of emerging is the process from small to
big and from simple to complex. It is a kind of
interaction of before and after association. These
interactions and the system coming from the
interaction are nonlinear. The interaction is its center
and is more complex than the simple sum of
individual behaviour. The stable mergence
phenomenon could be an integral part of more
complex emergence phenomena.
3.2.2 Mutual Adaptation between the Agents
and Environment
The agents in the supply chain are the economic
entities which are those independent or half
independent. These agents have the supply and
demand relation that is full of competition,
cooperation and dynamic state etc. Operation unit,
business process, member’s enterprise, supply chain
system, the whole action environment constitute
agents in different layers. And each agent has its
own target, conduct strategy, internal structure and
existence motive.
These agents interact through gathering to expect
to continuously adapt to environment. Towards the
whole supply chain, there is no concentrated control
centre to guide the behavior of each agent, but the
whole supply chain is still in good order. Because
this activity and the repeated and mutual gathering
function with environment, which are basic move for
development and evolution of the whole supply
chain system.
3.2.3 Dynamic State and Non-line of
Shipbuilding Supply Chain
Modern enterprises exist in the environment which is
full of severe competition and indetermination.
Under CAS, although a shipbuilding enterprise
makes very small decision, it will largely influence
on related enterprises (Forrester, 1961). In the
meantime, environment also influences on modes,
various rules or standards of enterprises. The agent
will change its standard which indicates its ability to
adapt to environment because of the change of
mode, target appearance, basic function standard and
the ability to adapt to the environment etc.
3.2.4 Complexity of Shipbuilding Supply
Chain
The intertexture of positive and negative feedback
results in the complexity of supply chain and the
positive feedback exerts a great influence on the
system. Shipbuilding supply chain faces an
environment that is full of complexity, the random in
agent, which plays an important role in process that
system looks for various possibilities. Some
important whole characteristics come out in the
mentally dense and the random process.
3.2.5 Coordination of Shipbuilding Supply
Chain
On the one hand, how to manage and control the
logistics flow, information flow, knowledge flow,
the funds flow between suppliers and demanders and
availably lower inventory, accelerate turnover of
logistics and information flow, raise the efficiency of
circulates in enterprise production and commodity,
quickly respond to market opportunity, which
become urgently problems needing to be worked
out; On the other hand, because of market
competition which is turning worse, the pursue of
enterprises in the whole supply chain will result in
some antinomies. Enterprises should know that the
whole supply chain can achieve the state which is
called Pareto Optimality, but it can not achieve the
state of the optimization of every enterprise.
Therefore, an enterprise should understand how to
cooperate with the enterprises which are in upper
and lower class. This can make sure that the
efficiency of the whole logistics chain is the biggest
and the efficiency of the enterprise reaches the best
state (Chandra and Kumar, 2001).
4 SIMULATION MODEL FRAME
OF SHIPBUILDING SUPPLY
CHAIN ACCORDING TO
MULTI-AGENT
According to CAS, the paper adopts the software
terrace-SWARM which was developed by SFI in
order to establish the model for CAS. The Swarm
model defines the frame of model. With the help of
this frame, we can build simulation model based on
Multi-agent.
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
392

4.1 Introduction of Swarm Model
Swarm model can make the customer build up
his/her own application according to this frame; at
the same time, it considers factors such as the
characteristics of different agents, and the details of
space distribution, the characteristics of network
connections and so on. It mainly carry out
interaction towards a series of independent
individuals through independent affairs carry on
handing over to with each other, it will help study
behavior of the complicated system which is
composed of many agents.
Lin, FuRen and others have studied the process
restructuring strategies towards different orders
under the application of Swarm, the strategies have
affected on the performance of the whole supply
chain. They obtained ideal results (Lin and Shaw,
1998).
4.2 Simulation Model of Supply Chain
based on Multi-agent
Table 1 shows mapping relationship between supply
chain model and Swarm model. We can see the four
similarities between them. These similarities are
very important for supply chain management.
According to this kind of corresponding relation, it
calls Swarm 21.1 class library in Jbuilder3.0
platform to build simulation model of supply chain
based on Multi-Agent (figure 4) shown. The
individual details are omitted for the limited length
of the paper.
Table 1: Mapping relationship between supply chain
model and Swarm model.
Supply chain model Swarm model
Different role agents in
different levels
A series of agents which
have different inner states
and behavior rules
Information interaction
among different agents
Information transmission
among different agents
Logistics flow and
machining process
The agent under discrete
event simulation and time
debugging triggering
The agents get the whole
supply chain
performance through
gathering and interaction
The performance of all
agents decides the whole
performance of system
In this simple model, system includes supplier
agent, manufacturer agent, customer agent, order
class and credit standing class, they are all
autonomic agents. Logistics, information, fund and
knowledge flows of the whole system will run under
demand pull. First, Model Swarm produces orders
and distributes orders for customers and the
customers consult price and delivery time from
manufacturers. Then manufacturers will inquire
supplier and supplier will quote in response. After
getting quotation, manufacturers will tell the
customers quotation and then the customers will
choose the best manufacturers. Similarly, the
manufacturers will also choose the best customers.
Finally, it will evaluate the credit of manufacturers
(suppliers) and update the prestige historical data.
5 RELATED STRATEGIES OF
SHIPBUILDING SUPPLY
CHAIN MANAGEMENT BASED
ON CAS
Supply chain is a CAS. Shipbuilding supply chain
management should fully consider the characteristics
of CAS and adjust management mode and theory.
We can put forward suggestions as follows (Wu,
2006).
SC_Observer
Observer
Swarm
Supply chain
Swarm
Manufacture
Agent
Supplier
A
g
ent
Customer
A
g
ent
Credit
Order
Figure 4: Simulation model of shipbuilding supply chain
management.
5.1 Self-management towards
Adaptability
Supply chain is a self-adaptive system, if one
enterprise wants to develop in the market; it need set
up a self-management thought. On the one hand, it
should coordinate with other enterprises in supply
chain; on the other hand it should pay attention to
itself-management. The kind of self expresses the
choices of “internal mode”.
RESEARCH ON INNOVATION OF SHIPBUILDING SUPPLY CHAIN MANAGEMENT BASED ON COMPLEXITY
ADAPTIVE SYSTEM
393

When the inner and outer environment is
changed, the shipbuilding enterprises should rapidly
readjust their goal and fundamental structure and
establish a rapid reaction mechanism for them to
change and optimize their own product
characteristics, management process, production
method, enterprise culture, even employee career
development. Meanwhile, supply chain network is
self-organization. Shipbuilding enterprises should
actively adjust their position in the supply chain and
reorganize themselves quickly according to the
dynamic change of the environment.
5.2 Paying Special Attention to
Adjustment of the Internal Model
Shipbuilding supply chain pursues the condition of
Pareto Optimality, this shows that enterprises in the
whole supply chain are unable to maximize their
own goals and interests. Based on this, the
enterprises need to establish their own
characteristics, such as product and service,
management, technology, enterprise culture and so
on, in order to meet diverse and dynamic market
demand. Under the optimization of the whole supply
chain, the enterprise realizes the target and
maximized benefit.
5.3 Strengthening the Management of
“Flow”
Towards supply chain management, we should do
well in managing logistics, information, knowledge
and fund flows among enterprises or between
enterprise and external environment. First, the
enterprises should realize that only they transmit
“flow” completely among themselves, it will be
possible for them to enhance their adaptive and anti-
risk ability. Enterprises should closely cooperate in
functional layer, operational layer and flow layer etc,
then eliminate the bottleneck of whole supply chain,
and raise the whole operation performance of supply
chain; Secondly, the enterprise shall establish
information system and strengthen the ability to
exchange information with external environment in
order to make enterprises effectively complete the
production plan and in or out goods scheduling.
Through this process, the enterprises can avoid
unnecessary waste.
5.4 Paying More Attention to Micro
Management
The good or bad operation performance of the enter-
prise depends largely on the national or social
policies and laws. The enterprise must recognize the
supply chain type which it is located in and
determine its strategic position in the shipbuilding
supply chain reasonably. Through the research, the
paper finds that the evolution and development of
the supply chain system is a collective dynamic
process. This can change strategic errors of some
enterprises that they blindly position themselves in
“core position” of supply chain. Only they clearly
establish strategic concept, they can reasonably
allocate resources and realize the target.
5.5 Establishing the Concept of
Continuous Learning
Shipbuilding supply chain system is the complex
adaptive system that has the ability to learn, which
means that supply chain is in change every minute,
including the change of external environment and
internal management pattern of the enterprise. Only
enterprises have the ability to learn fast, they can
deal with business environment which is changed
fast. Therefore, enterprises in shipbuilding supply
chain must establish the concept of learning
successively and make themselves become learning
organizations.
6 CONCLUSIONS
Shipbuilding supply chain is a CAS, so we need to
convert thinking mode and use the methods provided
by complexity science. Based on shipbuilding supply
chain operational mode of CAS, it can be more
objective to describe the behavior of complexity
system and provide more scientific operational
strategies of shipbuilding supply chain for policy
makers; Meanwhile, the paper puts forward the
corresponding innovation strategies of shipbuilding
supply chain management, such as self-management
toward adaptability; paying special attention to
adjustment of the internal model; strengthening the
management of “flow”; paying more attention to
micro management and establishing the concept of
continuous learning, which can reasonably interpret
complex phenomenon and resolve problems in
current shipbuilding supply chain. For the limited
time, simulation model of shipbuilding supply chain
hasn’t been simulated through the computer, which
will be the research direction in the future.
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
394

REFERENCES
Choi, T. Y., & Dooley, K. J., 2001. Supply networks and
complex adaptive systems: control versus emergence.
Journal of Operations Management, 351-366.
Chandra, C. & Kumar, S., 2001. Enterprise architectural
framework for supply-chain integration. International
Journal of Operation &Production Management,
21(4):461-475.
Chen, Y., 2001. Complex adaptive system theory and its
application: foundation, contents and implication.
Journal of Systemic Dial Entices, 9(4):35-39.
Dong, X., 2006. Research on enterprises’ innovation
system based on the CAS theory. Master Degree
Paper of Hebei Industry University, 10-14.
Forrester, J. W., 1961. Industrial Dynamics. Cambridge,
MA: MIT Press.
Lin, F., & Shaw, M. J., 1998. Reengineering the order
fulfilment process in supply chain networks. The
International Journal of Flexible Manufacturing
Systems, 197-229.
Li, B., 2006. Supply chain management lectures. Beijing:
electronic industrial press.
Qiu, S., 2003. Research on coordination theory, method
and application of complex adaptive systems, Doctor
Degree Paper of Tianjin University, 25-30.
Ren, W. & Zhao, H., 2008. Review of shipbuilding supply
chain management. Science Technology and
Engineering, 8(15):64-67.
Wu, J., 2006. Complex Adaptive System of Supply Chain
Management Strategy. Aerospace Industry
Management, 17-19.
Tan, Y., & Dong, H., 2001. Research on complexity
adaptive system theory and its application. System
Engineering, 19(5):1-6.
Zhang, T., 2003. Analysis and model of operation mode of
supply chain. System Engineering Theory and Practice,
8-9.
RESEARCH ON INNOVATION OF SHIPBUILDING SUPPLY CHAIN MANAGEMENT BASED ON COMPLEXITY
ADAPTIVE SYSTEM
395