Content uploaded by François B Vernadat
Author content
All content in this area was uploaded by François B Vernadat on Mar 19, 2017
Content may be subject to copyright.
1
dimensions
of
Integrated
The
Manufacturing Systems Engineering
P.
Ladet1,
F.
Vemadat2
haboratoire
d'Automatique de Grenoble (lAG), Institut de
Ia
Production
Industrielle, ENSG/-INPG,
46
avenue Felix Viallet, F-38031 Grenoble, France
Phone: +33 76
57
48
32; Fax: +33 76
57
43 17; e-mail: joyaux@imag.fr
2INRIA RhOne-Alpes,
46
avenue Felix Viallet, F-38031 Grenoble, France
Phone: +33 76
57
47
77; Fax: +33 76
57
47
54;
e-mail: Francois. Vemadat@inria.fr
Abstract
Modern, integrated manufacturing systems need to be engineered in a systematic way like any
other complex dynamic systems. Due to the extreme complexity and interdisciplinarity nature
of
manufacturing system design, analysis, reengineering and continuous improvement, and due to
the trend for internetworking
of
enterprises, a new discipline called Enterprise Engineering is
emerging. Different aspects or dimensions
of
Enterprise Engineering in the context
of
integrated
manufacturing systems engineering are reviewed in the paper.
Keywords
Enterprise Engineering, integrated manufacturing systems, interdisciplinarity aspects,
integration, system life cycle, complexity
1.
INTRODUCTION
Manufacturing enterprises play an essential role in industrialised countries both in terms
of
employment and revenues. Central to these are manufacturing systems, which have become
significantly complex systems to design and to control. They can be defined as socio-economic
discrete event dynamic systems.
Nowadays, the economic environment
of
most manufacturing enterprises is drastically
changing. The economy
of
scale is being replaced by an economy
of
scope. Customisation, i.e.
product adaptation to specific customer requirements,
is
driving the demand. Globalisation, i.e.
the necessity to be present on world-wide markets, implies a timely deployment strategy
in
terms
of
product manufacturing and distribution, sometimes forcing strategic alliances with
partner companies. Fierce competition with emerging countries forces industrialised countries to
produce at lower cost, with higher quality and in shorter delays. Finally, the consequence
of
widespread automation and the need to gain on productivity are pushing manufacturing
enterprises to lean their management and manufacturing operations, therefore employing less
and less people.
P. Ladet et al. (eds.), Integrated Manufacturing Systems Engineering
© Springer Science+Business Media Dordrecht 1995
4 Part One Introduction
Proper design
or
reengineering
of
their manufacturing systems is now a must for most
industrial companies to face international competition. Modern manufacturing systems must be:
-flexible/agile
-reactive
-integrated and
-cost-effective
Manufacturing systems, like any other type
of
complex systems, need to be designed and
engineered in a systematic way by means
of
structured approaches relying on sound principles
and supported by efficient tools and methods. We call such an emerging approach Enterprise
Engineering.
This book only provides a first step in this direction and deals with:
-the concept
of
the Extended Enterprise
-some Enterprise Engineering approaches under development
-techniques for business process modelling and reengineering
-formal approaches for (integrated) manufacturing systems specification
-Petri net techniques for modelling and analysing the physical part
of
the manufacturing system
as well as production planning aspects such as process planning
-techniques for manufacturing system coordination and integration (both at the plant level and
at the enterprise level), and
-pre-normative and standardisation issues in the areas
of
enterprise modelling and integration
2. ENTERPRISE ENGINEERING
Enterprise Engineering is the art
of
designing, implementing, maintaining and continuously
improving enterprise systems (processes and components) so that the enterprise can fulfil its
mission according to its business objectives.
Enterprise Engineering embraces under one term several engineering disciplines such as
industrial engineering, systems engineering, information systems engineering and production
engineering.
It
also relies
on
techniques from management sciences, organisation sciences,
applied mathematics (especially, simulation and Operations Research)
or
human resource
management.
Traditionally, these various disciplines have been applied separately and in
ad
hoc ways
when designing
or
reengineering some
of
the enterprise business processes
or
any part
of
the
enterprise. The overall process remains essentially sequential (performed step by step,
"throwing the project over the wall" from one engineering domain to the next one) and technical
barriers, mainly due to protected "islands
of
competence", are typical attitudes encountered in
manufacturing systems engineering departments. Organisation and human resource
management aspects remain clearly separated from technical aspects such as functional design,
information system design, manufacturing plant layout design
or
computer network system
design. Today, most companies realise that management, organisational and technical issues are
closely inter-related, especially in the context
of
integrated systems.
There is a need to break down these barriers and develop structured methodologies to
be
supported by models and computer tools to quickly and efficiently engineer/reengineer modern
manufacturing systems. Such methodologies should help and guide system designers to:
-improve integration
of
enterprise components to increase enterprise productivity and efficiency
on the basis
of
enhanced communication, cooperation and coordination
of
enterprise operations.
This is the aim
of
CIM (Computer-Integrated Manufacturing);
-simplify and lean management and manufacturing procedures to gain on costs and delays;
Integrated manufacturing systems engineering
5
-parallelise work and operate according to the Just-In-Time (JIT) philosophy to reduce time-to-
market and reduce backlogs and work-in-process inventories; and
-capitalise on previous knowledge and know-how to learn from past experience and build the
enterprise memory, as required by Continuous Process Improvement (CPI).
Methodologies supporting Enterprise Engineering must lead to solutions which help the
enterprise to compete on quality, costs and delays
(QCD)
and to face management
of
change
because
of
the fierce competition on markets in a rapidly changing world.
Several aspects or dimensions characterising Enterprise Engineering must be considered. In
this paper, we focus on:
-the dimension
of
interdisciplinarity
-the dimension
of
system life cycle
-the dimension
of
integration
-the dimension
of
complexity
3.
THE DIMENSION OF INTERDISCIPLINARITY (OR VIEWS)
The complexity
of
manufacturing systems,
as
systems created by humans for their own needs,
has progressively increased to the point
of
being comparable to the complexity
of
natural
systems. There are two reasons for this:
-their management and control functions are themselves becoming more and more complex as
well
as
their integration needs; and
-the role
of
human operators
in
the control loop, recently rehabilitated after an era
of
over
automation, brings a part
of
uncertainty into the control process. This is a dimension which is
not easy to formalise.
The control
of
production systems in general, and
of
manufacturing systems
in
particular, relies
on a representative modelling
of
these systems as well as the control and management policies
to be applied to them. However, each function and each control level, due to its nature and the
concepts handled, requires a priori a different modelling tool. For instance, a scheduling
function, a machine control function, resource sharing and synchronisation, flow modelling on
the shop floor or human operator integration, all call for different modelling and analysis tools.
An interdisciplinary approach for Enterprise Engineering can therefore be defined on a three-
layer basis, corresponding to three levels
of
granularity
in
system modelling:
I . The first layer concerns the elementary level: A given function, a view
of
a manufacturing
process, a system component are often themselves very complex. Their analysis assumes the
availability
of
methods and modelling tools able to take into account their detailed
specificities and intricacies to a level as close to reality as possible. This regularly leads
to
advanced enhancement
of
existing tools as well as to the definition
of
new tools, usually
derived from previous ones. Advances
in
Petri nets provide a good example (see the
Manufacturing System Analysis section
of
this book for more details).
2. The second layer concerns the level of tool cooperation: The tools used to represent a given
function or a given mechanism must be able to exchange information, to be synchronised or
to cooperate with one another
as
well as to take into account the function or mechanism
environment, i.e. the other functional elements
of
the system. This is the problem
of
structured representation
of
an
application perceived from different approaches and
disciplinarity angles at different levels
of
abstraction, each one relying on a different
perception
of
the system modelled
in
terms
of
a given tool. The problem
is
to be able
to
6
Part One Introduction
interface these tools by the definition
of
consistent information systems or even a unique
information system to support collaborative decision making and engineering thinking.
3. The third layer concerns the global level: It consists in developing an all-embracing
modelling environment able to represent the set
of
all
facets
of
a manufacturing system or
more precisely to federate tools which, although remaining different and specialised to the
aspects they have been developed for, have been designed in a perspective for integrated
analysis and design. This is an area still under development and
it
is a long way far from
complete achievement. Relevant papers related to this area can be found in the Enterprise
Engineering section
of
this book.
For each
of
these layers and throughout the system development life cycle, an interdisciplinary
approach must be used for Enterprise Engineering in general, and for integrated manufacturing
systems engineering
in
particular. This concerns engineering sciences on one hand and
management sciences and human sciences
on
the other hand.
Too often, the design and reengineering
of
manufacturing systems
is
primarly considered as
a matter
of
industrial engineering and application domain engineering (such as mechanical
engineering, electrical engineering, or food industry engineering for instance), i.e. as a technical
problem. In fact, it is also, and sometimes most
of
all,
an
organisational problem, a
management problem, a human problem and
an
economic problem. In the case
of
integrated
systems, it also becomes a problem for computer scientists and information system designers.
The challenge in Enterprise Engineering
is
to develop a framework supported by
methodologies and tools to provide these different areas
of
expertise or viewswith
-a
way to understand the "language" or viewpoint
of
other disciplines
- a way to indicate where and how each one fits
in
the framework
- a way to federate the viewpoints and expertises
of
each
of
them
so that they can operate in synergy to develop/reengineer enterprise systems faster, better and
at
a reasonable cost to meet specified business objectives.
4 THE DIMENSION OF SYSTEM LIFE CYCLE
The system life cycle covers the set
of
activities going from enterprise system inception to
engineering, implementation, operation, improvement and finally system dismantlement.
In
addition to the interdisciplinarity dimension, it
is
now widely understood that this process
is
not strictly sequential (it does not follow a "waterfall" approach) but involves a number
of
iterations, redoings and revisions at all levels. Furthermore, emerging principles
of
Concurrent
Engineering (CE) principles can be applied to the design and engineering
of
enterprise systems
(i.e. for an entire plant or just a subsystem
of
it). Moreover, since there
is
always something
changing within the enterprise or in its environment, operational and management systems must
be regularly analysed to be improved or corrected. These are the areas
of
strategic management
on the administrative side and Continuous Process Improvement (CPI) on the technical side,
both defining the needs for Business Process Reengineering (BPR).
New issues in enterprise systems engineering come from the pressure to pay more attention
to environmental and energy issues in product and production system design and operation
all
over the system life cycle. Recycling
of
product components
as
well as production system
dismantlement become important issues
in
systems engineering.
The major phases
of
enterprise system life cycle have been identified and documented by the
Purdue Enterprise Reference Architecture (PERA) and the European pre-norm ENV
40
003 -
Framework for Enterprise Modelling produced by Comite Europeen de Normalisation (CEN).
They involve:
Integrated manufacturing systems engineering 7
-Business Objectives and
Mission
Definition:
This concerns strategic planning defined
by
top
management and must answer such questions as: what will be produced, for which market
segments, in what quantities and where.
-Requirements Definition: In this phase, business users must express in detail what has to be
done to achieve business objectives.
-
Design
Specification:
This phase covers preliminary design and detailed design
of
business
processes, resource needs and management, information systems and infrastructure and
organisation structures. Technology aspects as well as human factors must be considered in
terms
of
required capabilities. Formal description and simulation techniques are extensively
used. Exception handling mechanisms must
be
planned and analysed.
-
Implementation
Description:
Decisions on resource selection and layout, data distribution,
computer network configuration, etc. are made and documented in this phase.
-
Installation:
Components and programs
of
the system are implemented and tested.
-
Operation:
This phase is concerned with actual exploitation
of
the system in its business
environment.
-
Management
of
change/Continuous
Process
Improvement:
This concerns the identification
of
shortcomings
or
defficiencies in the system and their correction.
-Dismantlement
or
shut down.
5.
THE
DIMENSION OF INTEGRATION
Integration means putting system components together to create a synergistic whole, the
capabilities
of
which encompass the capabilities
of
each
of
its components alone. The benefits
of
integration rely on improved communications, cooperation and coordination (C3)
of
the
business processes
of
the system.
In the case
of
integrated manufacturing systems, integration can be achieved at several levels:
-Computer systems integration: This is concerned with physical systems integration in terms
of
systems interconnections
by
means
of
computer networks. The 7-layer ISO-OSI
architecture has been the technical reference but Ethernet is dominating the market
of
local
area networks. In the near future, ATM is supposed to be the new standard both for local
and wide area networks because
of
its high-speed and multimedia capabilities.
-Shop-floor integration: This is concerned with manufacturing systems integration and was
originally initiated
by
the
MAP
(Manufacturing Automation Protocol) initiative in the
US
followed by the
CNMA
ESPRIT project in Europe, both based on the ISO-OSI architecture.
The MMS (Manufacturing Message System) language is an important result
of
this work as
well as fieldbus definitions (such as
FIP
and Profibus), which are simplified and faster
computer network architectures dedicated for shop-floor applications.
-Plant integration: This form
of
integration goes one step beyond shop-floor integration and
may involve the exchange
of
engineering data such as drawings and bills
of
materials. Thus,
standard data exchange formats (such as IGES
or
STEP/EXPRESS for exchange
of
product
and process data and EDI for administrative electronic data interchange) as well as common
services are required to make applications communicate or even intemperate. Different kinds
of
integration platforms
or
integrating infrastructures known as middleware platforms are
being proposed. Examples
of
generic integration platforms are OSFIDCE and
OMG/CORBA. CCE-CNMA is an integration platform dedicated to CIM environments.
-Enterprise Integration: Enterprise Integration concerns the integration
of
the various business
processes
of
the enterprise to facilitate information, material and decision flows and therefore
increase reactivity and productivity.
It
deals with integration at the business level
of
the
enterprise.
It
requires a model
of
the business processes (as provided by Enterprise
8
Part
One
Introduction
Modelling) and an information infrastructure to monitor execution
of
this model and
coordination
of
the business processes. These are the ideas promoted by CIMOSA, the
European open system architecture for CIM.
-Extended Enterprise or inter-enterprise integration: Finally, the last level
of
integration
concerns the Extended Enterprise. While CIM mostly concerns intra-enterprise integration
(i.e. integration
of
the business processes
of
a given enterprise), the Extended Enterprise is
concerned with inter-enterprise integration (i.e. internetworking
of
enterprises interacting
along a common supply chain).
MMS and CIMOSA are covered
in
the Enterprise Engineering section
of
the book while the
coordination, CCE-CNMA and integrating infrastructure issues are covered
in
the
Manufacturing System Coordination and Integration section. The Extended Enterprise has a
section
of
its own.
6.
THE
DIMENSION
OF
COMPLEXITY
Due to requirements for flexibility/agility, reactivity, integration and automation, manufacturing
systems, and especially discrete manufacturing systems, become extremely complex systems
to
design and to control. Their complexity often largely exceeds the capabilities
of
one person. For
instance, let us think about a manufacturing system producing gear-boxes for cars
or
trucks or
an automated system producing electronic chips. Such systems are made
of
10 to 20
workcenters, can produce over a hundred
of
different product types and may involve dozens
of
different concurrent business processes.
The complexity may come from the large number
of
system components and their
interactions. It may also come from the sophistication
of
the system.
In terms
of
systems engineering, we have outlined
the
interdisciplinary aspects and the need
for different types
of
models to correctly assess the systems. The mixing
of
competences
(engineering, organisational, socio-economic, etc.) is also a factor
of
complexity in the design
of
such
systems
as well
as
in
the
design
of
relevant Enterprise Engineering methodologies.
Another aspect related
to
complexity concerns scalability
of
models. Currently, various
techniques are proposed to model manufacturing systems, their business processes or their
information systems. They are usually presented on small toy-examples
in-
the literature.
However,
in
industry people must deal with large models representing real systems.
It
is
therefore important to consider
the
scalability aspects when proposing a new modelling
approach.
7. CONCLUSION
Enterprise Engineering as defined
in
this paper
is
a collection
of
many different activities put
together to engineer enterprise systems, and especially integrated manufacturing systems. The
challenge
of
Enterprise Engineering
is
to provide industry with methods and guidelines for
building efficient enterprise systems (processes and components) faster and better. There is
therefore a need for developing advanced tools (models and computer tools) as well as suitable
methodologies.
Development
of
these tools and methodologies must take into account the interdisciplinarity
nature
of
such an engineering process, must cover the entire enterprise system life cycle, pay
special attention to integration issues and provide mechanisms
to
face the complexity dimension
of
the problem.
The aim
of
this book is to provide a first step
in
this direction. Especially, it
is
shown how
recent developments
in
Enterprise Modelling and Enterprise Integration can be combined with
Business Process Reengineering (BPR) and manufacturing system specification and analysis
Integrated manufacturing systems engineering 9
techniques (such as formal description techniques and Petri nets)
to
achieve some
of
the goals
of
Enterprise Engineering.
More developments remain to be done to bring Enterprise Engineering to the level
of
maturity
of
Software Engineering for example, from which it borrows many ideas and adopts a
similar approach. Especially, it is the authors' opinion that more work on formal techniques
should be pursued to design robust and reliable systems so that their qualitative and quantitative
properties can be more precisely analysed before they are installed and operated.
8 BIOGRAPHY
Pierre Ladet is a French citizen and lives in Grenoble. He got his Doctorat d'Etat degree in 1982
from the Institut National Polytechnique de Grenoble (INPG) in Physical sciences (Automatic
control). He his currently a professor at INPG. He has been a research officer at the Automatic
Control Laboratory
of
Grenoble (1982-1995) where he created a research team on discrete
events systems.
His research interests include discrete events systems and decision rules modelling using
Petri nets. With his research team, he developed a Petri nets methodology and automatic control
architecture including scheduling, decision and control aspects.
Professor Pierre Ladet is currently head
of
the national programme
of
CNRS « Conception
des systemes de production
»,
an interdisciplinary research action which puts together
disciplines from Ingineering Sciences and Human Sciences.
He also a mission officer for the French Ministry
of
Research and Education.
Fran¥ois Vemadat is a French and Canadian citizen. He got his Ph.D. degree in
1981
from
the University
of
Clermont, France in Electronics and Automatic Control. From 1981 till
1988, he has been a research officer at the Division
of
Electrical Engineering
of
the National
Research Council, Ottawa, Canada. In 1988, he joined INRIA, a French research institute in
computer science and automatic control.
His research interests include CIM database technology and information systems,
enterprise modeling and integration, knowledge representation, formal description
techniques, Petri nets and manufacturing plant layout design. He has been involved
in
several ESPRIT projects and was one
of
the chief architects
of
CIMOSA. He has published
over 90 scientific papers and is co-author
of
three
books. He is
the
European editor for
the
International Journal
of
CIM
and serves the scientific committee
of
several journals.
Dr. Vemadat served several times as a technical expert for the CIM program
of
the
Commission
of
the European Communities (EU DG Ill). He acts as a French expert for
national, European and international standardisation bodies on Enterprise Modelling and
Integration (AFNOR, CEN TC 310, ISO TC 184). He is a member
of
the IEEE Computer
Society, ACM and SME.
