Progress of Discrete Event Simulation in manufacturing

and logistics
Bernd Noche
Department of Mechanical Engineering
Transport Systems and Logistics
University of Duisburg - Essen
Lotharstraße 1-21
47057 Duisburg, Germany
b.noche@uni-duisburg.de

1. Introduction

Discrete event simulation is a powerful mean for the support of the planning and
operation of manufacturing and logistic systems. Their importance is growing though
European industry is not yet fully aware of the relevance of simulation.

The United States of America are leading in the process of evolving the technology
of simulation. They are organising the most important simulation conferences (e.g.
Winter simulation conference) and in the research field they are persecuting an
ambitious research program (e.g. IMTI – Integrated Manufacturing Technolgy
Initiative) which is accompanied by leading companies, especially in the field of the
production of military equipment and aircraft industry. The situation in Europe is
completely different. At present there is no decisive commitment of these industries
to simulation, their application and the development of software. Major trends are
supported mainly by the automotive industry which is merely applying software than
funding the developments. The chances, however, of European industry lie in their
strong application of simulation models for challenging task. Large and complex
models can be found in a variety of application fields, from chemical plants or mail
distribution units to production facilities and supply chain networks.

2. Simulation needs continuous developement

Research, however, is still part of the development strategies of the vendors, a

general guided industrial and governmental development concept is missing. This
situation is alarming since most of the vendors are small companies which are
unable to invest in long term oriented software products. The uncertainity of industry
decisions and attitudes as well as the lack of standardisation of models, interfaces
1
and basic data acquisition makes every high volume invest decision of vendors to an
explosive high risk event.

But there are good chances for European industry to benefit from their strong and
challenging applications. Through many branches and through different company
sizes in different stages of product life cycles simulation technology is acknowledged
and applied. Why isn’t it possible to use simulation everywhere and as a general tool
for investment decisions and optimisation of daily operations?
It is difficult to fancy all obstacles and provisos. But some of them may be:

- The technology is still too complicated and expensive

- Simulation people have to be experts in the technology
- The integration of the simulation tools into general application software is still
very limited
- Adaption of simulation to knowledge domains is poor
- New application fields seem to be too risky
- The benefit of simulation is not credible enough and convincing
- Data for feeding models is difficult to acquire
- Large models need much time for modelling and simulating
- The modelling and result interpreting is not standardized
- Result generation and explication of effects is often intransparent
- Control strategies are hard to be developed and implemented

And additionally project expenses and times for project developments are shortened,
simulation technology, however, is seen as a delaying factor. The challenge is to find
new ways and solutions for strengthening the application of simulation in Europe.
Therefore a roadmap has to be developed which addresses the following activities:

- Identify major needs and causes of impediment

- Form interest groups for the elaboration of suggestions for improvement
- Establish platforms of vendors and customers with common interests
- Find relevant fundings for the most important developments
- Try to publish success stories and define standards
- Spread the knowledge to vendors and simulation users

2
 - Establish a permanent information concept proving benefits and cost
reductions

The list of measures may be larger and must be elaborated more detailed. In a
preliminary stage, however, a European common understanding of the needs and
benefits of simulation has to be established which finally leads to improvements. It
has to be ensured that these improvements are really increasing the acceptance of
simulation. Thus, the market of the technology will be enlarged leading to a
reasonable return on investment of the efforts.

Learning

Raw materials
Unfinished goods
Products Finished goods
Article of trade
Information
Funtions
Business processes Cost model
Strategies

Worker
Staff Schedulers
Employees

Systems
Operation processes Facilities
Controls

Finance Customer

Fig. 1: Environment for the development of simulation technology in manufacturing

and logistics

The figure illustrates the environment in which simulation will progress:

- Customer: direct fulfilment to customer needs and demands

- Finance: aspect which has to be elaborated more clearly showing directly
cost/benefit ratios
- Learning: improving tools and spreading the simulation technology

The simulation technology will progress if it is possible to focus more precise on the
elements of the logistic chain: Processes on the operation and business level,

3
organisational aspects with control strategies for personnel and resource allocation
and products which have to reach in time.

3. Application examples

There where simulation is used in Europe we can find trained communities with
extraordinary skills and experience. The reasons for this situation might be the fact
that
- industry in Europe is investing in highly automated and complex logistic
systems
- control strategies for efficient use of capacities and resources require
sophisticated algorithms
- process orientation is a major perspective of development of supply chains
- operation aspects are considered more and more leading finally to fast and
integrated solutions

Some areas where simulation experience is directly available can be found in the
design of
- distribution systems
- manufacturing systems and
- chemical plants

For the design of those systems the typical modelling elements have been defined
and a variety of questions which should be answered within the projects are
available.
Figure 2 explains the basic elements of modelling of distribution systems. It is clear
that not only the modelling elements define the quality of the simulation study. Fast
reaction to customer demands are essential. This leads to a general view on
modelling accuracy. It shows that in many cases a mere modelling of conveyors and
stores is not yet satisfying. The models contain strategies and control rules which
help to evaluate just in time concepts and how time tables are met under difficult
circumstances.

4
 Flowchart Breakdowns Shift models
Organisational control rules Specification

Delivery of customer
Customer orders orders

Pallets
Goods receipt Parcels

Distribution system

Article master data Processing time

Route table Conveyor data Initial storing situation

Fig. 2: Essential data base for simulating distribution centers

The design of modern distribution systems is devoted to mission oriented processes

where promised delivery dates are a crucial part of control and operation strategies.
The mission to be accomplished is to meet promised delivery dates.

It is interesting to see that in manufacturing modelling elements do not differ

considerably from distribution systems although some specific requirements are
obvious. Figures 3 gives an impression of some interesting modelling aspects in
some application fields of manufacturing. Here it is not only necessary to model
machines or production lines. Main focus has to be directed onto the design of whole
factories offering the chance to strengthen the specific structures of a factory in order
to achieve a higher competitiveness by relying on their own history, staff skills,
customer structure and scope of products.

Simulation in chemical industry has evolved very much slower than in other sectors.
The reasons might be that the available simulation tools never fitted really to the
application fields. The tools had been developed in a manufacturing environment and
were adapted more or less skilful to the needs of chemical plants. The nature of
chemical products is completely different. This leads to specific modelling
techniques, control strategies and production programs.

5
 xafjasöfkjda xafjasöfkjda
kffajdöfa kffajdöfa
Customer orders
Organisational List of articles Forecasts Pool of employees
instructions

Transport

xafjasöfkjda xafjasöfkfaj
kffajdöfa döfa

Working plan Parts list Breakdowns x

y Operation Shift models
Repair
times
Assembly plan

Fig. 3: Essential data base for modelling manufacturing systems

Figure 4 gives an overview about modelling elements in chemical industry. The goals
in the simulation studies are generally the same as in manufacturing. One of the big
differences, however, is the investment in machinery. In chemical plants the role of
persons is often restricted to supervision functions whereas in the manufacturing
process the interaction between persons and machines are an essential part of the
questionnaires.

Data about installations Raw materials

- Capacity of tanks, silos List of
- Connexions articles
- Production facilities
- Warehouse

Shift models

Process flow
Operation
times

Delivery of Pallets Customer orders

Organisational customer orders Big bags
instructions Breakdowns

Fig. 4: Essential data base for modelling chemical plants

6
4. Conclusions

The simulation technology in Europe has an own history and has led to specific
European concepts. The reason is simple: Considerable amount of money has been
invested into the technology and will continuously be invested leading to a higher
integration into existing design and operation environments.

European industry should take this as a basis and develop own concepts. These
concepts should strengthen the applicability of the simulation technology by:
- integrating knowledge about simulation, validation and verification, modelling
etc. into the simulation environments
- improving explainability of simulation results not only by more and better
visualisation and animation
- expanding the use of the models from the design step into the operation level
by focussing more on operation needs
- strengthening accessablity of tools, models and methodology
- penetrating more into new field such as cost simulation, test environments etc.

Customers must be convinced that simulation is helpful and that the value of benefits
is exceeding by far expenses. But this benefit is limited today to some extraordinary
projects with a certain magnitude of investments and complexity. The devotion of
developers to models which can be used for operations will offer the chance to
improve the performance of European industry in general.

The improvements may be funded either by research programs or investments of

vendors or large customers. These investments, however, should comply with
general goals defined by appropriate working groups and made available to the
simulation community by dissemination or direct access. And in the near future other
application fields are needing support from discrete event simulation technology such
as traffic systems, flow of persons, document flow, health care, distributed computer
(software) networks.
Many results from the traditional application fields of logistics can be transferred or
adapted in these additional fields. It is not necessary to invent the wheel twice.