Content uploaded by Robert Bucki
Author content
All content in this area was uploaded by Robert Bucki on Mar 26, 2017
Content may be subject to copyright.
The Method of Logistic Optimization in E-commerce
Robert Bucki
(The College of Informatics and Management in Bielsko-Biała
Bielsko-Biała, Poland
rbucki@wsi.edu.pl)
Petr Suchánek
(Silesian University in Opava, School of Business Administration in Karviná
Karviná, Czech Republic
suchanek@opf.slu.cz)
Abstract: Rapidly changing business environment requires new approaches and methods for
supporting management systems in all types of companies. Modern companies doing business
use e-commerce systems by default. One of the key areas of e-commerce systems is logistics
and the supply chain. The optimal way to ensure the success of logistics and supply chains is to
use the methods of modeling and simulation based on appropriate models and especially its
mathematical representation. In this paper, authors highlight the customer-oriented model of the
e-commerce system and deal with logistic optimization and simulations. As an example,
a sample logistic structure which requires the adequate control approach is presented. This is
realized by means of heuristic algorithms which are responsible for meeting the set criterion.
Moreover, the criteria to either maximize the production output or minimize the lost flow
capacity of the logistic system or minimize the tool replacement criterion are introduced.
Equations of state are given in order to represent the flow of material through the logistic
system.
Keywords: e-commerce system, logistics, simulation, logistic process, production
maximization criterion.
Categories: C.4, H.1.1, H.4.2, I.6.8
1 Introduction
E-commerce development has acquired great importance in the contemporary
business. It has helped business organizations, businessmen and end users to
overcome the barriers of time and distance to sell, buy and carry out other business
transactions across the globe. The development of e-commerce makes it necessary to
search for new management methods and techniques designed to optimize the entire
e-commerce system. The emergence of e-commerce has changed the relationship
between customers and retailers and created a new need to restructure the entire
supply chain and logistic system. The main objective remains the maximization of the
whole system effectiveness which is to be achieved at the lowest possible costs. E-
commerce systems are by their nature considered as systems with low operating costs.
This must be provided by means of the adequate optimization approach. The area,
which could significantly affect the costs associated with running an e-commerce
system, is either logistics or the supply chain (SC). The initial point for the
Journal of Universal Computer Science, vol. 18, no. 10 (2012), 1238-1258
submitted: 30/6/11, accepted: 25/5/12, appeared: 28/5/12 © J.UCS
implementation of all types of systems is a suitable model with its own architecture
and mathematical description. Different types of models in conjunction with
a suitable mathematical representation allow us to perform the simulation process
whose outputs can help managers make suitable decisions. The major support of
simulation is provided by advanced computer systems. In this context, we often talk
about the so-called computational logistics. Computational logistics involves planning
and implementation of large and complex tasks using computations and higher
mathematics. Computational logistics is implemented in many areas, including the
flow and storage of goods, services, and related information from their point of origin
to the point of consumption. In computational logistics, optimization models and
algorithms must be developed and verified for the plan and execution of complex
logistics and supply chain systems. The main objective of this paper which emerges
from the above discussion is to present the importance of logistic optimization in
customer-oriented e-commerce systems as well as a sample logistic structure which
requires the adequately specified control approach. These days it is evident that
production activities are treated as sub-processes of trade processes. Should it be
thought that an e-commerce system is directly oriented on the order of the specified
product e.g. furniture, it is necessary to properly emphasize the problem of order
realization as well as supply of a ready product. Moreover, it is also indispensable to
optimize logistics within realizing the determined products. The main goal of the
paper is to present how the order manufacturing process is controlled by means of
heuristic algorithms while implementing the required criterion.
2 E-commerce system
The structural model describes an e-commerce system as a set of functionally
connected components (Figure 1). The main basic components of e-commerce
systems are:
customers and generally business environment,
the Internet,
the web server,
LAN (Local Area Network),
CRM (Customer Relationship Management) characterized among others in
[Pradeep, 10],
ERP (Enterprise Resource Planning) [Bredford, 09],
payment system [White, 08],
delivery of goods [Chopra, 09],
after-delivery (after-sales) services,
information systems (CRM/ERP) of cooperating suppliers.
Important and integral parts of the whole system are hardware, software, people,
co-operative suppliers, legislation and generally information and communication
technologies (ICT). All shown parts of e-commerce systems are supported and
controlled by the management at all control levels (tactical, operational, strategic).
Most important for control and support functions are SCM - Supply Chain
Management, FRM - Financial Resource Management, HRM - Human Resource
1239
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
Management, IBP – Integrated Business Planning and Information
system/Information technology administration (IS/ICT management).
Figure 1: Basic structure of e-commerce system Source: after [Suchánek, 10]
3 Customer-oriented e-commerce system
The e-commerce system should be seen as the part of a complete business
environment. Here, the main decisive part of the environment is understood as the
customer whose orientation is one of the main prerequisites of an effective
management system.
Based on its own research, Simulation Systems Ltd. [Simulation Systems, 11]
presented the fundamental definition and description of customer-oriented systems.
They say that each customer-oriented system results from the individual work with
the client taking into account specific details of his business processes. The customer-
oriented system is made according to the principle emphasizing the well-known truth
which reads that the system must be created for the user, avoiding the need to adjust
the user for the system.
The development of customer-oriented systems includes the full cycle of tasks,
such as: collection and analysis of data needed for automation, definition of the
requirements, designing, implementation, introducing, testing and maintenance of the
system. The specific features of the approach to designing of customer-oriented IS are
formulated as follow:
1240 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
fundamental analysis of the customer’s business as a base for working out the
solutions corresponding to the actual goals and problems of the customer;
detail elaboration and coordination with the customer of all stages of the project
development, control points and required resources;
support of the convenient mode of maintenance, modification and extension of the
system;
provision of openness, mobility and scalability of the system.
Customer-oriented functionalities of e-commerce comprise business-to-customer
(B2C) applications such as remote shopping, banking and infotainment-on-demand.
To study the possibilities of e-commerce functionality improvement, the structure
model presented in Figure 1 does not seem to be sufficient. Any company strives to
accomplish its business strategy. Following the business strategy, long-term and
short-term objectives and targets are set.
After measuring the results achieved in a certain period of time, the differences
from the target values are evaluated and corrective actions for the next term are
specified by the company management. Thus, in a very general way, any business
system can be looked upon as a specific form of a control loop (Figure 2). In our
opinion, this holds true also for e-commerce systems.
Let us set an example in which a customer faces a billboard showing a car. Then
the customer decides to buy this car immediately. There is a credit card in his pocket
with the necessary financial coverage and the telephone number of the seller on the
billboard. The customer contacts the dealer and tells him he wants to buy a car
immediately specifying required parameters and the geographical localization (the
customer does not have to call, but he can use the Internet via his mobile phone or
another mobile device). The dealer should be able to determine immediately where
the desired car is available and how long it will take them to deliver the car to the
customer. Let us suppose that using the information system the dealer determines that
the car can be delivered to the customer’s destination in e.g. half an hour and
subsequently makes a quote. If the customer agrees, he can sit on a bench next to the
billboard and wait. Within this period of time, a responsible person will bring him the
required car. The customer makes a payment by means of a mobile credit card reader
and takes over the car. This example presents another possible direction of e-
commerce development. The scheme in Figure 3 fully complies with the need of
direct orientation to the customer. However, it requires to solve a wide range of
technical, personnel and legal issues. The discussed model seems to be realistic. In
order to make this model fully operational in the real environment, the following
conditions have to be met:
ensuring unambiguous identification of the customer (the customer who has made
the order);
defining the legal framework for this type of business in particular;
defining principles of the legal protection of the dealers (for example it is
necessary to ensure that the customer could not deny the ordered goods – click-
wrap and browse-wrap contracts);
ensuring technical support of the mobile credit card readers;
defining and implementing the logistic system (especially the supply chain).
1241
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
Figure 2: Generic e-Commerce model Source: after [Suchánek, 11]
The controlled subsystem consists of the actual e-Commerce system itself. The
inputs constitute the management decisions on the e-Commerce parameters like
prices, investments, marketing decisions, safety rules etc. The outputs include sales
value, profits and margins, return on investment (ROI), stability, security and other
key performance factors. The outputs are measured in the measuring element and
compared with the targets in the differential component.
There is a general rule on the basis of which the customer wants to find products
quickly and easily and get products in the shortest time as well as to pay for goods in
the selected way and to have the longest possible warranty. One way of meeting this
rule is to accept the basic philosophy of orientation on the customer and use all
available technologies and business and management practices in company
operations. The general business pattern following the stated conditions can be
presented as transformation of Figure 1 into Figure 3. Using the pattern shown in
Figure 3, it is possible to advance much further in the considerations.
1242 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
Figure 3: Main activities focused on the customer
4 Logistic and supply chain and its optimization in e-commerce
For the discussed purpose, namely the area of electronic commerce, logistics is
defined as a business planning framework for the management of material, goods,
services, information and capital flows. It includes the increasingly complex
information, communication and control systems required in today's business
environment [Logistics, 96], [Langevin, 10]. Especially, in business, the logistics has
the meaning of the way of the supply of goods by which the customers’ requirements
can be solved. It determines the path of transporting goods and a lot of important
information exchanged between the consumer, production or source points. It has the
special term production logistics which means that the supply of the high quality
products in time along with the security must be assured (for tax related products: like
iphone, iPod, etc. which are for sale). Logistics and the entire system must be
controlled. In this context, we are talking about the management of logistics and
supply chain management.
Logistics management is that part of the Supply Chain Management process
which is responsible for planning, implementing and controlling the efficient,
effective forwarding and reversing the flow and storage of goods, services, and
related information between the point of origin and the point of consumption in order
to meet customers' requirements. Supply Chain Management (SCM) encompasses the
planning and management of all activities involved in sourcing and procurement,
conversion, and all Logistics Management activities. Importantly, it also includes
coordination and collaboration with channel partners which can be suppliers,
intermediaries, third-party service providers and customers. In essence, Supply Chain
Management integrates supply and demand management within and across companies
[Cantrell, 04].
One of the key areas of e-commerce system is to optimize the whole e-commerce
system. It is clear that to be optimized, the entire e-commerce system and all of its
subsystems must be optimized. The rules defined by [Ratliff, 03] as the general
principles of logistic (as a subsystem of the e-commerce system) optimization can be
1243
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
adopted:
Objectives - must be quantified and measurable;
Models - must faithfully represent required logistic processes;
Variability - must be explicitly considered;
Data - must be accurate, timely, and comprehensive;
Integration - must support fully automated data transfer;
Delivery - must provide results in a form that facilitates execution, management
and control;
Algorithms - must intelligently exploit individual problem structure;
People - must have the domain and technology expertise required to support the
models, data, and optimization engines;
Process - must support optimization and have the ability to continuously improve;
Return on Investment (ROI) - must be provable considering the total cost of
technology, people and operations.
Optimization is a highly timely topic in the field of manufacturing and supply
chain management. Optimal control of a substitutable inventory system, structured
assemble-to-order systems and the impact of advance demand information on various
production-inventory control mechanisms are the key factors which must be taken
into account while planning order realization procedures [Shanthikumar, 03].
A deterministic system does not involve any randomness in the development of
subsequent states of the logistic system. Therefore, such a model will always produce
the same output from a given initial state. Stochastic ordering is a fundamental guide
for decision making under uncertainty. It is also an essential tool in the study of
structural properties of complex stochastic systems [Shaked, 07]. Supply chain
optimization is the application of processes and tools to ensure the optimal operation
of a manufacturing and distribution supply chain. This includes the optimal placement
of inventory within the supply chain, minimizing operating costs (including
manufacturing costs, transportation costs, and distribution costs). This often involves
the application of mathematical modelling techniques using computer software.
The general objective is to find optimal logistic network design. In the logistic
network design problem (LNDP), decisions must be made regarding the selection of
suppliers, the location of plants and warehouses, the assignment of activities to these
facilities, and the flows of raw materials and finished products in the network
[Cordeau, 08]. Optimal network design/redesign minimizes inventory carrying,
warehousing, and transportation costs while satisfying customer response-time
requirements. Specifics include the network’s distribution levels and centers, location
and mission of each facility, assignment of supplier and customer locations to each
center, and inventory deployment [Frazelle, 06]. To create an optimal network
design/redesign, Frazelle recommends a 10-step of logistics network design process
shown in the Table 1 (analogous approach can be found for example in [Groothedde,
11]).
1244 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
10-step of logistic network design process
1. Assess/evaluate the current network
2. Design and populate the network optimization database.
3. Create network design alternatives such as more or fewer hierarchies,
multi-commodity flows, pooling opportunities, merge-in-transit, direct
shipping, cross docking, and supply-flow optimization concepts.
4. Develop a network optimization model.
5. Choose a network optimization tool.
6. Implement the network model in the chosen tool.
7. Evaluate alternative network designs.
8. “Practicalize” the recommended network structure.
9. Compute the reconfiguration cost.
10. Make a go/no-go decision.
Table 1: The 10-step of logistic network design process
Simulations can be effectively used in order to support optimization procedures in
the state of uncertainty. Simulations are based on modelling and in particular
mathematical models. Computer simulations are used extensively as models of real
systems to evaluate output responses. Applications of simulation are widely found in
many areas including supply chain management, finance, manufacturing, engineering
design and medical treatment [Fu, 95], [Kim, 05], [Semini, 06]. Supply chain
simulations [Chang, 03]:
help to understand the overall supply chain processes and characteristics by
graphics and/or animation;
are able to capture system dynamics: using probability distribution, user can
model unexpected events in certain areas and understand the impact of these
events on the supply chain.
could dramatically minimize the risk of changes in planning process: By what-if
simulation, user can test various alternatives before changing plan.
Traditionally, the formal modelling of systems has been carried out by means of
a mathematical model which attempts to find analytical solutions to problems which
enable the prediction of the system behaviour on the basis of the set of parameters and
initial conditions. There are many methods for the purposes of simulation and
especially simulation-based optimization techniques. Kleijnen claims that supply
chain simulations can be carried out by means of the use of spreadsheet simulation
methods, system dynamics, discrete event simulation and/or business games
[Kleijnen, 03]. E-commerce systems can be classified into the category of intelligent
systems. Evolutionary optimization can be effectively used for the optimization of
intelligent systems. Evolutionary optimization is becoming an omnipresent technique
in almost every process of intelligent system design. Just to name few, engineering,
control, economics and forecasting are some of the scientific fields that take
advantage of an evolutionary computational process that supports engineering
systems with intelligent behavior [Nedjah, 08]. Detailed breakdown of the simulation-
1245
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
based optimization techniques is presented in Table 2 [Ashram, 96], [Silva, 05].
Simulation-based Optimization Techniques
Deterministic search techniques
Heuristic search technique
Complete enumeration and
random choice
Response surface search
Pattern search techniques
Conjugate direction search
Steepest ascent (descent)
Tabu search technique
Hooke and Jeeves type
techniques
Simplex-based techniques
Probabilistic search techniques
Random search
Pure adaptive and hit-and-run
search
Evolutionary Techniques
Simulated annealing
Genetic techniques
A short comparison
References and Further
Readings
Stochastic approximation techniques
Kiefer-Wolfowitz type
techniques
Robbins-Monro type
techniques
Gradient surface method
Post-solution analysis Rare Event Simulation
Table 2: Simulation-based Optimization Techniques
For the purpose of presenting a sample logistic manufacturing system,
deterministic search techniques and especially heuristic search techniques are
implemented. Professor Ashram further states that the heuristic search technique is
probably most commonly used in optimizing response surfaces. It is also the least
sophisticated scheme from the mathematical point of view and it can be thought of as
an intuitive and experimental approach. The analyst determines the starting point and
the stopping rule based on the previous experience with the system. Generally,
heuristic search methods generate and test algorithms in order to meet the set
manufacturing criteria.
In e-commerce, logistic optimization must be based upon business objectives to
find the schedule producing the product with the least cost or the shortest production
lead time. It all means that the planner ought to be presented with the best algorithm
enabling him to manufacture the best sequence of products. Traditionally simulation
models have been used effectively on a project-by-project basis to analyze changes to
a factory physical equipment. Simulation models contribute more effectively to the
design of new methods for production scheduling and new supply chain
configurations. By the nature of the control the simulation model is to assist with the
day-to-day scheduling of the company doing business.
Knowing that there is an order to be realized, we can easily find out in advance
what kind of charge will be necessary and what amount of it will be needed. In this
context we are talking about supply chain master planning presented, for example, in
1246 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
[Schröpfer, 09]. Running the simulation process for just a few hours, the results show
a completely different but more accurate story. This is why simulation is so important.
Simulation, and only simulation, takes into account the combined effect of variability,
uncertainty, and complex interdependencies between processes. Simulation is helping
companies improve their businesses and become more competitive. Increasingly
competitive markets bring new challenges and customers' demands are constantly
changing. As a result, manufacturers need to become more responsive, more quickly,
more efficiently, yet often within tighter budgets and timescales. Simulators are
helping them to visualize, analyze and optimize their processes to achieve these
business performance improvements [Bucki, 09].
Following the above remarks, in the next section we present the sample logistic
structure which requires the adequate control approach.
5 General assumptions of the sample logistic system
The problem itself consists in determining the sequence of elements of the order
vector which are to be realized subsequently. The proposed heuristic algorithms
choose the required element on which certain operations are carried out. The state of
orders decreases after each production decision which influences the state of the
whole logistic system at every stage. The given criteria are used on condition that
each of them is associated by adequate bounds. Assuming that the results of
calculations which are made for one chosen heuristic algorithm do not deliver
a satisfactory solution, there arises a need to test other algorithms.
Let us assume that the logistic system consists of I active logistic blocks arranged
in a series. Each logistic block is meant to carry out a production operation. There are
no buffer stores placed between logistic blocks in the discussed case. Each logistic
block consists of a certain number of production stands. If a production stand in the
logistic block does not have enough production capacity to accept all production flow,
another parallel stand carrying out the same production operation is activated in this
exact logistic block. If a production stand is not needed any more, it is deactivated in
the discussed logistic block [Bucki, 10].
Let us introduce the vector of charges:
l
wW , Ll ,...,1,
where: l
w - the l-th charge material.
Let us introduce the vector of orders:
n
zZ , Nn ,...,1
,
where: n
z - the n-th production order (given in units).
Having defined the vector of charges W as well as the vector of orders Z, the
assignment matrix of products to charges is proposed as follows:
ln,
, Nn ,...,1
, Ll ,...,1,
where: ln,
- the assignement of the n-th product to the l-th charge material.
Elements of the assignment matrix take the following values:
1247
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
0
1
,ln
if the n-th product is realized from the l-th charge,
otherwise.
The assignment matrix of products to charges is presented as follows:
We also assume that used charge vector elements are immediately supplemented,
which means that we treat them as the constant source of charge material. However,
for simplicity reasons, it is assumed that each n-th product is made from the universal
charge which enables realization of the given n-th product from any l-th charge
material element, Ll ,...,1. The determined charge material enters the production
system itself. Machines in each logistic block i
M, Ii ,...,1
in the production system
carry out autonomous operations on the specified material. The operations are
realized subsequently. If the logistic block i
M is able to accept the material from
which the n-th product is made, it is passed to this block and after carrying out the
adequate operation is passed to the block 1i
M. After leaving the I-th block, it fills
the elements of the order vector in the serial logistic system.
n
I
jI
JI
i
ji
Ji
j
J
lz
m
m
m
m
m
m
m
m
m
w
1,
,
,
1,
,
,
1,1
,1
,1
......
Let us introduce the vector of production stands corresponding with the logistic
blocks:
jii mM ,
, Jj ,...,1
,
where: ji
m, - the j-th production stand in the i-th logistic block.
The stage k, Kk ,...,1
is the moment at which the manufacturing process at any j-th
production stand in the i-th logistic block begins. We need to consider that decisions
are made at the stage 1k, Kk ,...,1.
It is assumed that operations in the stands in the system are realized subsequently.
The structure scheme of the serial logistic system is presented as follows:
1248 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
k
n
k
I
kjI
kJI
k
i
kji
kJi
k
kj
kJ
ln z
e
e
e
e
e
e
e
e
e
1,
,
,
1,
,
,
1,1
,1
,1
,.....
,
where: Ii,...,1,Jj ,...,1,Kk ,...,1,, Nn ,...,1
Ll ,...,1.
Elements representing production stands take the following values:
0
1
,
kji
e
if the n-th product is realized by the j-th production stand in
the i-th logistic block at the k-th stage,
otherwise.
Moreover, it is assume that 1
1,
k
i
ie
Having assumed the above, we can introduce the life vector of the logistic system
for a new brand set of tools:
in
gG ,
, Nn ,...,1
, Ii ,...,1
,
where: in
g,- the number of the n-th product units which can be realized in any
j-th production stand of the i-th logistic block before the tool in
this stand is completely worn out and requires an immediate
replacement.
The elements of the matrix G take the following values:
0
,
in
g
if the n-th product is realized in the j-th production stand
of the i-th logistic block,
,...,1
,
otherwise.
Moreover, we can assume that
in
Ii
g,
1
,
,...,1
.
If the number
is reached for the given n-th element of the vector Z in the j-th
production stand of the i-th logistic block, the tool in this exact stand has to be
replaced by a new one.
Let
1
,
1)( kji
k
nnsS be the matrix of state of the logistic system for the
n-th product realization at the stage 1
k where 1
,
)( kji
ns is the number of units of the
n-th product already realized in the j-th stand of the i-th logistic block with the use of
the installed tool. The matrix of state can be shown in the following extended form:
1249
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
1
1,
1
1,
1
1,1
1
,
1
,
1
,1
1
,
1
,
1
,1
1
)(...)(...)(
)(...)(...)(
)(...)(...)(
k
I
k
i
k
kjI
kji
kj
kJI
kJi
kJ
k
n
nsnsns
nsnsns
nsnsns
S
Let
1
,
1)( kji
k
nnpP be the matrix of the flow capacity of the logistic system
for the n-th product realization at the stage 1
k where 1
,
)( kji
np is the number of
units of the n-th product which still can be realized in the j-th stand of the i-th logistic
block.
The matrix of state can be shown in the following extended form:
1
1,
1
1,
1
1,1
1
,
1
,
1
,1
1
,
1
,
1
,1
1
)(...)(...)(
)(...)(...)(
)(...)(...)(
k
I
k
i
k
kjI
kji
kj
kJI
kJi
kJ
k
n
npnpnp
npnpnp
npnpnp
P
On the basis of the above assumptions we can determine the flow capacity of
the j-th production stand in the i-th logistic block for the n-th element of the order
vector Z at the stage 1k:
1
,,
1
,)()( kjijn
kji nsgnp
The manufacturing procedure consists in realizing orders in sequence
(manufacturing of the order may begin when the previously realized one leaves the
logistic system). Its disadvantage consists in the need of waiting for completing the
manufacturing process of a certain product before resuming it again for the next one.
This results in not using the available flow capacity of the whole production system.
Moreover, during the production course tools must be replaced. When another
element of the order vector Z enters the production system, the state of the system has
to be recalculated.
Let us define the production times for the n-th product, Nn ,...,1
in the j-th
production stand of the i-th logistic block in the matrix form:
prIN
priN
pr
N
pr
In
pr
in
pr
n
pr
I
pr
i
pr
pr
T
,,
1,
,
,
1,
,1,11,1
......
......
......
If the n-th product is not realized in the j-th production stand of the i-th logistic
block, then 0
,
pr
in
(in this case the implemented technology excludes production
operations in certain logistic blocks).
Let us define the vector of replacement times for the tools in the logistic system:
repl
I
repl
i
replrepl
T
......
1
,
1250 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
where: repl
i
- the replacement time of the tool in the j-th production stand of the i-th
logistic block.
If pr
in
pr
in 1,
,
, then the j-th production stand of the i-th logistic block becomes
blocked while manufacturing the n-th product so there is the need to activate the
production stand 1.1 ji
m which is based on the assumption that the number of active
production stands in the i-th logistic block is increased by 1 (not to block the
production process in the logistic block 1i
M) on condition that the number i
J is not
exceeded, where i
jis the number of the j-th production stand in the i-th logistic
block. It is justified by the fact that the production of the subsequent part of the n-th
order from the vector
Z
can be started instantly without having to await for
completing the current production operation in the logistic block i
M. The structure of
such a system guarantees continuing the production activity in any stand 1j in the
i-th logistic block if the stand i
j requires tool replacement.
Let us introduce the production rate vector
in
vV ,
. Its element in
v, is the
number of units of the n-th product made in the time unit.
Let us calculate the total manufacturing time of all elements from vector Z:
K
k
I
i
repl
i
k
i
pr
in
I
i
N
nyT 0
,
1
The coefficient k
i
y takes the following values:
0
1
k
i
y
if the replacement procedure of the tool in the i-th stand is carried out,
otherwise.
6 Production criteria
The criteria presented hereby are to either maximize the production output or
minimize the lost flow capacity of the production stands or minimize the tool
replacement time. Let us propose production criteria for the logistic system along with
the necessary bounds:
6.1 The production maximization criterion
Let us introduce the production maximization criterion:
max
11111
K
k
N
n
k
n
K
k
kxqQ ,
where: k
n
x - the number of units of the n-th element realized at the k-th stage.
The tool replacement bound: cy repl
i
I
i
k
i
1
,
where: c - the maximal allowable tool replacement time,
1251
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
repl
i
- the replacement time of the used tool in the j-th stand of the i-th logistic
block.
The flow capacity bound:
in
I
i
kin
k
igpy ,
1,
,
where: kin
p, - the lost flow capacity of the j-th stand of the i-th logistic block
at the k-th stage.
The production-maximizing criterion is reduced to the replacement time of tools
and flow capacity bounds.
6.2 The lost flow capacity criterion
Let us introduce the flow capacity criterion:
min
1,
11122
I
j
kin
K
k
I
i
k
i
K
k
kpyqQ
The tool replacement bound: cy repl
i
I
i
k
i
1
The order bound: n
N
n
k
nzx
1
The lost flow capacity criterion is reduced to the replacement time of tools and
order bounds.
6.3 The minimal tool replacement time criterion
Let us introduce the flow capacity criterion:
min
11
3
repl
i
K
k
I
i
k
i
yQ
The flow capacity bound: in
I
i
kin
k
igpy ,
1,
The order bound: n
N
n
k
nzx
1
The minimal tool replacement time criterion is reduced to the flow capacity
bound and the order bound.
7 Equations of state
The state of the discussed serial logistic system for the n-th element changes in the
production course as follows:
K
n
k
nnn SSSS ......
10
The state of the j-th production stand of the i-th logistic block in case of the n-th
product manufacturing changes consequently:
1252 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
Kji
kjijiji nsnsnsns ,,
1,
0
,)(...)(...)()(
which can be written in the following form:
k
n
kji
kji
kji xns
ns
ns
1
,
1
,
,
)(
)(
)(
if no n-th product is realized in the j-th stand of
the i-th logistic block at the k-1 stage,
otherwise.
Let i
be the tool to be replaced with a new one, I
i
1 . The state of the j-th
production stand of the i-th logistic block in case of replacement of tools changes in
the way shown below:
0
)(
)(
1
,
,
kin
kin ns
ns if
i at the stage k-1,
if
i at the stage k-1.
If tools in all stands are totally worn out, then GS
and they need an immediate
replacement procedure to resume the production process.
The order vector Z changes after every production decision:
Kk ZZZZ ......
10
The order vector is modified after every decision about production:
1
1
k
n
k
n
k
n
k
nz
xz
z if the number of units k
n
x of the n-th order is realized at the k-th stage,
otherwise.
8 Heuristics
In order to control the logistic process we need to implement heuristics for
determining elements from the vector Z for the production process. The following
control algorithms are put forward:
8.1 The algorithm of the maximal order
This algorithm chooses the biggest order vector element characterized by the biggest
coefficient 1k
n
in the state 1k
S.
To produce the element a, the following condition must be met:
1
1
1max
)( k
Nn
k
a
kn
aq
,
where: 11 k
n
k
nz
.
1253
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
The above approach is justified by avoiding excessive bringing the production
line to a standstill in order to change an element to be manufactured. If in state 1k
S
only minimal orders were chosen, in consequence the number of orders might be
reduced. Such control is favorable because the serial production line is blocked and
must be stopped only in order to replace the tools in certain stands (on condition that
the replacement process disturbs the flow of the material).
8.2 The algorithm of the minimal order
This algorithm chooses the smallest order vector element characterized by the
smallest coefficient 1k
n
in the state 1k
S.
To produce the element a, the following condition must be met:
1
1
1min
)( k
Nn
k
a
kn
aq
,
where: 11 k
n
k
nz
.
The above approach is justified by the need to eliminate the elements of the order
vector Z which could be sent to the customer just after the n-th product leaves the
production line on condition that the customer sets such a requirement.
8.3 The algorithm of the relative order
This algorithm chooses the order element characterized by the maximal relative order
coefficient 1k
n
in the state 1k
S.
To produce the element a, the following condition must be met:
1
1
1max
)( k
n
Nn
k
a
kaq
,
where: 0
1
1
n
k
n
k
nz
z
.
It is assumed that the orders are realized one after another that is to say each order
element zn in the state 1k
S is reduced partly. Such control is advantageous when
some parts of the order are needed earlier and the rest can be manufactured later.
9 The block diagram of the logistic process
In order to create the block diagram of the logistic manufacturing system (Figure 4)
criteria
Q, B,..,1
must be implemented where:
the production maximization criterion → 1
Q
the lost flow capacity criterion → 2
Q
the minimal tool replacement time criterion → 3
Q
1254 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
Moreover, a heuristic algorithm
h,
,...,1
is responsible for choosing
the n-th element of the order vector where:
the algorithm of the maximal order → 1
h
the algorithm of the minimal order → 2
h
the algorithm of the relative order → 3
h
Figure 4: The block diagram of the discussed logistic manufacturing system
10 Conclusions
The problem presented in the paper discusses the issue of the serial production system
with no buffer stores. The system delivers ready products corresponding with the
elements of the order vector. The main goal is to fulfil the task by the set criteria.
There is also a possibility to implement a two- or three-criterion model. Such models
may lead to delivering a solution which would satisfy criteria included in the
discussed model only partly as there should be bounds added. Heuristic algorithms
1255
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
proposed in the paper enable the operator to choose the satisfactory production
sequence on the basis of which a certain element of the order vector is determined to
be realized. The use of one specified algorithm does not mean that we will achieve the
result satisfying the given criterion. It is advisable to implement other algorithms and
decide which one minimizes the order realization time, the loss of residual capacity,
and the total replacement time or satisfies a hypothetical customer’s demand not
specified hereby. Another idea already used in available works is to simulate the
combination of heuristic algorithms. By means of this method, we are able to combine
two or more algorithms. It also seems reasonable to draw products for manufacturing.
To achieve a satisfactory result a big number of simulations must be carried out. In
conclusion, it must be admitted that a simulator imitating real environment should be
built to continue this work. Simulation experiments carried out in the synthetic
environment may deliver an answer which heuristic approach is the expected one for
a certain criterion. The idea of time scaling by means of the simulation method on
condition a satisfactory number of simulation experiments is carried out seems to able
to deliver the results minimizing the total order realization time.
In conclusion, we can say that the interactive software combines real-life images,
with a step-by-step immersion process, which successfully replicates the experience
of controlling a real system. By providing the user with both the flexibility to operate
the whole system and the support of the data files, which can be easily downloaded, it
is possible to fit the simulator into any given system. Although the approach is
flexible, it is far from being an easy option. The design of the simulator ensures its
reliability at every stage of the technological process. Even in situations where there
can immerse a dose of uncertainty, the user will be able to count on the intuitive
support of the program. By using a simulator there will be an impact on further
production performance. The advantages far outweigh the costs of building
a simulator in a very short period of time proving its market value, which will
increase with every production unit. Operators will gain self-confidence as well as
improve their communication skills and learn how to contribute more effectively in
a plant activity, which is required of them in return.
Acknowledgements
This paper was supported by project ,,Strengthening of Competitive Advantages in
Research and Development of Information Technology in the Moravian-Silesian
Region" Nr. CZ.1.07/2.3.00/09.0197 within the EU Operational Program Education
for Competitiveness.
References
[Arsham, 96] Arsham, H. Systems Simulation: The Shortest Route to Applications. 9th Edition.
National Science Foundation, Baltimore, Maryland, USA, 1996.
[Bălan, 09] Bălan, A. The Knowledge Management – Necessity for the Modernization of the
Organizations. In Journal of Applied Economic Science, Volume IV, Issue 4(10), 2009.
[Bora, 09] Bora, C. Advantages and Disadvantages of Electronic Commerce (e-commerce).
From Advantages and Disadvantages of Electronic Commerce (e-commerce), 2009. Retrieved
1256 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
December 13, 2009. URL: <http://www.buzzle.com/articles/advantages-and-disadvantages-of-
electronic-commerce-e-commerce.html>
[Bredord, 09] Bredford, M. Modern ERP: Select, Implement & Use Today's Advanced Business
Systems, Chattanooga, H&M Books, 2009.
[Bucki, 10] Bucki, R. Managing Flexibility of the Production System with Interoperation Store.
Mezinárodní vědecký seminář: Nové trendy v ekonometrii a operačním výzkumu. Katedra
ekonometrie Fakulty informatiky a statistiky VŠE v Prahe a Katedra operačného výskumu a
ekonometrie Fakulty hospodárskej informatiky EU v Bratislave. Zbornik 15. – 17. december
2010.
[Bucki, 09] Bucki, R., Antonyová, A., Suchánek, P. Economic Control of the Distribution
System by Means of Heuristic Approach. Vol. 12, No. 2, National Academy of Sciences of
Ukraine, State Department on Communication and Informatization of Ukraine, State Scientific
and Research Institute of Information Infrastructure, 2009.
[Burd, 05] Burd, S.D. Systems Architecture. 5th edition. Course Technology, 656 p. 2009.
[Cantrell, 04] Cantrell, B. Logistics Vs SCM. Supply Chain Dynamics, Inc., Washington, 2004.
[Chang, 03] Chang, Y., Makatsoris, H. Supply Chain Modeling Using Simulation. In
International Journal of Simulation Modelling, Vol. 2 No. 1, 2003.
[Chopra, 09] Chopra, S., Meindl, P. Supply Chain Management, 4
th Edition. New Jersey,
Prentice Hall, 2009.
[Cordeau, 08] Cordeau, J.F., Laporte, G., Pasin, F. An iterated local search heuristic for the
logistics network design problem with single assignment. Elsevier, International Journal of
Production Economics, Volume 113, 2008
[Dictionary.com, 11] Dictionary.com. Logistics. Retrieved March 20, 2011. URL:
<http://dictionary.reference.com/>
[Frazelle, 06] Frazelle, E. How to design a logistics network in 10 steps. [On-line] Retrieved
May 10, 2011. Material Handling &Logistics - MH&L, 2006.
[Fu, 95] Fu, M.C., Hu, J. Sensitivity analysis for Monte Carlo simulation of option pricing.
Probability in the Engineering and Information Sciences, 9(3):417–446, 1995.
[Groothedde, 11] Groothedde, B. Logistics Network Design and Collaboration: Design,
Theory, and Practice. 1 st edition. Springer, Physica-Verlag HD, New York, 2011.
[Kim, 05] Kim, P., Ding, Z. Optimal engineering system design guided by data-mining
methods. Technometrics, 47(3), pp. 336–348, 2005.
[Kleijnen, 03] Kleijnen, J.P.C. Supply Chain Simulation: A Survey. Center for Economic
Research, Discussion Paper No. 103, Tilburg, 2003.
[Langevin, 10] Langevin, A., Riopel, D. Logistics Systems: Design and Optimization. Springer,
New York, 2010.
[Logistics, 96] Logistics Partners Oy. Helsinki, 1996.
[Nedjah, 08] Nedjah, N., Mourelle, L.M. Evolutionary Optimization for Intelligent Systems
Design. J.UCS Journal of Universal Computer Science, Vol. 14, no. 15, 2008.
[Pradeep, 10] Pradeep, I.K. et al. CRM System Using CM-AKD Approach of D3M.
International Journal of Engineering Science and Technology, Vol. 2(3), pp 237-242, 2010.
1257
Bucki R., Suchanek P.: The Method of Logistic Optimization ...
[Ratliff, 03] Ratliff, H.D. The 10 Rules for Logistics Optimization. Scientific Logistics, Inc.,
Atlanta, 2003.
[Schröpfer, 09] Schröpfer, A., Kerschbaum, F., Schütz, C., Pibernik, R. Optimizations for Risk-
Aware Secure Supply Chain Master Planning. J.UCS Journal of Universal Computer Science,
Volume 15, Issue 15, 2009.
[Semini, 06] Semini, M., Fauske, H. Applications of discrete-event simulation to support
manufacture logistics decision-making: A survey. In L. F. Perrone, F. P. Wieland, J. Liu, B. G.
[Shaked, 07] Shaked, M., Shanthikumar, J.G. Stochastic Orders. Springer Series in Statistics,
XVI, 2007.
[Shanthikumar, 03] Shanthikumar, J.G., Yao, D.D., Zijm, W., Henk, M. (Eds.) Stochastic
Modeling And Optimization Of Manufacturing Systems And Supply Chains. International
Series in Operations Research & Management Science, Vol. 63, 2003.
[Silva, 05] Silva, C.A., Sousa, J.M.C., Runkler, T., Palm, R. Soft computing optimization
methods applied to logistic processes. In International Journal of Approximate Reasoning
archive, Volume 40 Issue 3, November, New York, 2005.
[Simulation Systems, 11] Simulation Systems Ltd. Customer-oriented systems. [On-line]
Retrieved May 10, 2011. Simulation Systems Ltd., Obninsk, 2011. URL:
<http://www.ssl.obninsk.ru>
[Suchánek, 10] Suchánek, P. Business Intelligence as a Support of E-commerce Systems in
Connection with Decision Making and Cross-Border Online Shopping. In Journal of Applied
Economic Science, Volume V, Issue 1(11), Spring 2010.
[Suchánek, 11] Suchánek, P., Dolák, R., Miškus, M., Šperka, R. Intelligence Decision Support
Systems in E-commerce. In book Efficient Decision Support Systems: Practice and Challenges
– From Current to Future / Book 2, Open Access Publisher InTech, 2011.
[White, 08] White, J.J. Principles of Payment Systems. New York, West, 2008.
1258 Bucki R., Suchanek P.: The Method of Logistic Optimization ...
