from
Information Systems
Development
to
Enterprise Engineering
Jan L.G. Dietz
Delft University of Technology
Outline
Problems and analysis
The Enterprise Engineering Paradigm
The need for scientific rigor
Conclusions
Jan L.G. Dietz: Lecture ISD 2010 - slide 2
Outline
Problems and analysis
The Enterprise Engineering Paradigm
The need for scientific rigor
Conclusions
Jan L.G. Dietz: Lecture ISD 2010 - slide 3
Some persistent problems in enterprises
The complexity of organizations has become unmanageable.
The complexity of ICT applications has become unmanageable.
The ICT applications landscape is mostly a big mess.
ICT applications do not meet business requirements and expectations.
The maintenance costs of ICT applications are ever increasing.
There is still no (positive) correlation between ICT investments and
revenues.
Jan L.G. Dietz: Lecture ISD 2010 - slide 4
Analysis of the problems
The information systems professionals fall short in assisting enterprises
to implement change initiatives effectively.
The key reason is the low level of professionalism: they are not able to
deliver ICT applications that satisfy the real needs.
The organizational professionals also fall short in assisting enterprises to
implement change initiatives effectively.
The key reason is their predominant managerial approach (functionorientation, black-box thinking), whereas changing an enterprise requires
an engineering approach (construction-orientation, white-box thinking).
Jan L.G. Dietz: Lecture ISD 2010 - slide 5
Motivation for Enterprise Engineering (1)
The vast majority of strategic initiatives fail, meaning that enterprises
are unable to gain success from their strategy.
This figure is reported from various areas: Information Systems Development,
Business Process Management, Total Quality Management, Six Sigma …
Abundant research indicates that the key reason for strategic failures
is the lack of coherence and consistency among the various
components of an enterprise.
Jan L.G. Dietz: Lecture ISD 2010 - slide 6
Motivation for Enterprise Engineering (2)
Both organizations and software systems are complex systems, and
prone to entropy.
This means that in the course of time, the costs of bringing about similar
changes increase in a way that is known as combinatorial explosion.
Regarding (automated)
demonstrated.
information
systems,
this
has
been
Regarding organizations, it is still a conjecture.
Entropy can be reduced and managed effectively through modular design
based on atomic elements.
Jan L.G. Dietz: Lecture ISD 2010 - slide 7
The mission of Enterprise Engineering
Addressing the challenges mentioned before requires a paradigm
shift.
It is the mission of the discipline of Enterprise Engineering to develop
new, appropriate theories, models, methods and other artifacts for
the analysis, design, implementation, and governance of enterprises
by combining (relevant parts of) management and organization
science, information systems science, and computer science.
The ambition is to address (all) traditional topics in said disciplines
from the Enterprise Engineering Paradigm.
In addition, the results of our efforts should be theoretically rigorous
and practically adequate.
Jan L.G. Dietz: Lecture ISD 2010 - slide 8
The evolution of Enterprise Engineering
Information Systems Sciences
Form
Data
Data Systems
Engineering
Content
Information,
Communication
Information Systems
Engineering
Intention
Collaboration,
Cooperation
Organization
Sciences
Enterprise Engineering
Jan L.G. Dietz: Lecture ISD 2010 - slide 9
Outline
Problems and analysis
The Enterprise Engineering Paradigm
The need for scientific rigor
Conclusions
Jan L.G. Dietz: Lecture ISD 2010 - slide 10
The EE Paradigm: postulate 1
In order to perform optimally and to implement changes successfully,
enterprises must operate as a unified and integrated whole.
Unity and integration can only be achieved through deliberate
enterprise development (including design, engineering, and
implementation) and governance.
Jan L.G. Dietz: Lecture ISD 2010 - slide 11
The EE Paradigm: postulate 2
Enterprises are essentially social systems, of which the elements are
human beings in their role of social individuals, bestowed with
appropriate authority and bearing the corresponding responsibility.
The operating principle of enterprises is that these human beings
enter into and comply with commitments regarding the products
(services) that they create (deliver).
Commitments are the results of coordination acts, which occur in
universal patterns, called transactions.
Jan L.G. Dietz: Lecture ISD 2010 - slide 12
The EE Paradigm: postulate 3
There are two distinct perspectives on enterprises (as on all systems):
function and construction. All other perspectives are a subdivision of
one of these. Accordingly, there are two distinct kinds of models:
black-box models and white-box models.
White-box models are objective; they regard the construction of a
system. Black-box models are subjective; they regard a function of a
system. Function is not a system property but a relationship between
the system and some stakeholder(s).
Both perspectives are needed for developing enterprises.
Jan L.G. Dietz: Lecture ISD 2010 - slide 13
The EE Paradigm: postulate 4
In order to manage the complexity of a system (and to reduce and
manage its entropy), one must start the constructional design of the
system with determining its ontological model.
This is a fully implementation independent model of the construction
and the operation of the system. Moreover, an ontological model has
a modular structure and its elements are (ontologically) atomic.
For enterprises the meta model of such models is called enterprise
ontology. For information systems the meta model is called
information system ontology.
Jan L.G. Dietz: Lecture ISD 2010 - slide 14
The EE Paradigm: postulate 5
It is an ethical necessity for bestowing authorities on the people in an
enterprise, and having them bear the corresponding responsibility,
that these people are able to internalize the (relevant parts of the)
ontological model of the enterprise, and to constantly validate the
correspondence of the model with the operational reality.
It is a duty of enterprise engineers to provide the means to the people
in an enterprise to internalize its ontological model.
Jan L.G. Dietz: Lecture ISD 2010 - slide 15
The EE Paradigm: postulate 6
To ensure that an enterprise operates in compliance with its strategic
concerns, these concerns must be transformed into generic functional
and constructional normative principles, which guide the (re-)
development of the enterprise, in addition to the applicable specific
requirements.
A coherent, consistent, and hierarchically ordered set of such
principles for a particular class of systems is called an architecture.
The collective architectures of an enterprise are called its enterprise
architecture.
NOTE
The term “architecture” is often used (also) for a model that is the outcome of
a design process, during which some architecture is applied. We do not
recommend this homonymous use of the word.
Jan L.G. Dietz: Lecture ISD 2010 - slide 16
The EE Paradigm: postulate 7
For achieving and maintaining unity and integration in the (re-)
development and operation of an enterprise, organizational measures
are needed, collectively called governance.
The organizational competence to take and apply these measures on
a continuous basis is called enterprise governance.
Jan L.G. Dietz: Lecture ISD 2010 - slide 17
Outline
Problems and analysis
The Enterprise Engineering Paradigm
The need for scientific rigor
Conclusions
Jan L.G. Dietz: Lecture ISD 2010 - slide 18
Is it φτψ or bullshit?
Jan L.G. Dietz
Delft University of Technology
What is bullshit?
Bullshit is unavoidable whenever circumstances require someone to
talk without knowing what he is talking about.
(Harry Frankfurt)
Bulshitters retreat from searching for correctness to personal sincerity.
Since they cannot be true to the facts, they try to be true to
themselves. This is the core of bullshit.
(Harry Frankfurt)
Bullshit is a greater enemy of truth than lies are.
(Harry Frankfurt)
Jan L.G. Dietz: Lecture ISD 2010 - slide 20
What is φτψ?
The combination of the letters stands for rigorous science in general.
In isolation they stand for three theories:
φ-theory (FI): Fact - Information
A theory about factual knowledge, resolving questions like “How do
computers mimic human cognition?”, “What is the difference between
information and data?”, and “What is a fact?”.
τ-theory (TAO): Technology - Architecture - Ontology
A theory about the development of systems. It clarifies the difference
between function and construction and the role of technology, architecture,
and ontology in system development.
ψ-theory (PSI): Performance in Social Interaction
A theory about the construction and operation of organizations. It makes the
design of complex organizations (including ICT applications) intellectually
manageable.
Jan L.G. Dietz: Lecture ISD 2010 - slide 21
The φ-theory: ontological parallelogram
to
ms
for
nsi
object
on
con
na
xte
sig
se
to
de
subjective
ha
ers
name
denotes
type
instance of
ref
tes
fact
member of
Jan L.G. Dietz: Lecture ISD 2010 - slide 22
class
objective
Violation of the φ-theory
In order to intellectually manage the design of information systems, in
particular data bases, the (conceptual) data model needs to be verified
by exemplification (instantiation)!
Modeling approaches with verification:
NIAM: Natural Language Information Analysis Method)
ORM: Object Role Modeling
Modeling approaches without verification:
ER: Entity Relationship, as well as all derivates like:
UML: Unified Modeling Language
This is not φτψ but bullshit
Jan L.G. Dietz: Lecture ISD 2010 - slide 23
The τ-theory: white-box model
the mechanic's perspective
construction :
the components and their
interaction relationships
operation :
the manifestation of the
construction in the course of time
constructional composition
car
chassis
wheels
motor
Jan L.G. Dietz: Lecture ISD 2010 - slide 24
lamps
The τ-theory: black-box model
the driver's perspective
function :
relationship between
input and output
functional decomposition
Jan L.G. Dietz: Lecture ISD 2010 - slide 25
behavior :
the manifestation of the
function in the course of time
The Generic System Development Process
using system
construction
ontology
functional
principles
constructional
principles
object
system
function
function
design
design
constructional
requirements
requirements engineering can only become
complete and objective when it is based on
the ontological model of the using system
technology
ontology
implementation
implementation
functional
requirements
construction
object system
construction
technology
Jan L.G. Dietz: Lecture ISD 2010 - slide 26
engineering
reverse engineering
architecture
Violation of the τ-theory
Current Architecture Frameworks (which are nothing more than
development methods in disguise), are incoherent and
inconsistent collections of best practices, lacking any scientific
basis.
At best they may be helpful in project management.
This holds for TOGAF and for all proprietary frameworks.
This is not φτψ but bullshit
Jan L.G. Dietz: Lecture ISD 2010 - slide 27
ORGANIZATIONAL BUILDING BLOCK
initiator
transaction
executor
A0
T1
A1
Summary of
the ψ-theory
basic transaction process
rq
ontological
pm
actor role
COORDINATION (F)ACTS
rq
pm
O-phase
E-phase
st
ac
st
ac
PRODUCTION (F)ACTS
R-phase
customer has entered
into a commitment
ontological
a rental has been started
atomic process step
infological
customer has specified
what he wants
customer : request : employee : [rental] has been started
< performer > < intention > < addressee > < proposition >
datalogical
customer has expressed this
in an English sentence
infological
the amount to be paid
has been calculated
datalogical
performa
informa
forma
human abilities
Jan L.G. Dietz: Lecture ISD 2010 - slide 28
the rental form has been
filled out
ORGANIZATIONAL BUILDING BLOCK
executor
initiator
transaction
executor
A0
T1
A1
initiator
Summary of
the ψ-theory
initiator
executor
quit
allow
basic transaction process
rq
rq
COORDINATION (F)ACTS
re
fuse
ontological
pm
actor role
cancel
(rq)
de
cline
allow
pm
re
quest
E-phase
st
ac
ac
O-phase
st
pro
mise
cancel
(pm)
PRODUCTION (F)ACTS
R-phase
re
fuse
ontological
customer has entered
into a commitment
a rental has been started
atomic process step
infological
re specified
customer has
fuse he wants
what
datalogical
customer has expressed this
in an English sentence
allow
customer : request : employee : [rental] has been started
ac
cept
< performer > < intention > < addressee > < proposition >
cancel
(ac)
state
performa
cancel
(st)
infological
the amount to be paid
allow
has been calculated
datalogical
the rental form has been
filled out
re
fuse
re
ject
informa
forma
stop
human abilities
Jan L.G. Dietz: Lecture ISD 2010 - slide 29
The the ψ-theory: organization theorem
B-organization
Ontological Production:
deciding
judging
manufacturing
I-organization
Infological Production:
remembering
computing
reasoning
D-organization
Datalogical Production:
copying
storing
transporting
Jan L.G. Dietz: Lecture ISD 2010 - slide 30
The indefiniteness of process modeling
activity A
document
activity B
Is passing the document from A to B:
Only a datalogical act?
Example: A hands over the document to B to archive it.
Or (also) an infological act?
Example: A informs B about the content of the document.
Or (also) an ontological act?
Example: A requests B to do something.
Jan L.G. Dietz: Lecture ISD 2010 - slide 31
Violation of the ψ-theory
Current business process modeling approaches, like Flowchart,
BPMN, EPC, and Petri Net reduce business processes to sequences
of (observable) actions and results, thereby loosing the essential
deep structure (which is always a tree of transactions) and
neglecting all tacitly performed coordination acts.
Therefore they are useless (if not dangerous) for business process
re-design and re-engineering.
This is not φτψ but bullshit
Even worse are the function-oriented techniques (SADT, IDEF0)
since by definition they reflect the personal interpretation of the
modeler (black-box model)!
Jan L.G. Dietz: Lecture ISD 2010 - slide 32
Outline
Problems and analysis
The Enterprise Engineering Paradigm
The need for scientific rigor
Conclusions
Jan L.G. Dietz: Lecture ISD 2010 - slide 33
Conclusions
• There is an urgent need to broaden the scope of Information
Systems Development towards Enterprise Engineering
• There is an urgent need to base our work on solid scientific
foundations. Help fighting the bullshit!
• Only by adopting the Enterprise Engineering Paradigm can
substantial organizational changes (including Information Systems
Development) be made intellectually manageable.
Jan L.G. Dietz: Lecture ISD 2010 - slide 34
The CIAO! Network
The discipline of Enterprise Engineering is being developed by a
network of research institutes: the CIAO! Network.
Full members:
TU Delft (Prof. Jan Dietz, Prof. Alexander Verbraeck, Dr. Joseph Barjis,)
Polytechnic Utrecht (Prof. Hans Mulder)
TU Lisboa (Prof. José Tribolet, Prof. Pedro de Sousa, Dr. Jan Hoogervorst,
Dr. Martin Op ‘t Land)
University of Antwerpen (Prof. Herwig Mannaert, Prof. Hans Mulder, Prof.
Jan Verelst)
University of St. Gallen (Prof. Robert Winter, Dr. Antonia Albani)
Aspirant members:
HSE at Nizhniy Novgorod (Prof. Eduard Babkin)
Tokyo Institute of Technology (Prof. Junichi Iijima).
Jan L.G. Dietz: Lecture ISD 2010 - slide 35
Activities of the CIAO! Network
• For about five years we have run a CIAO! Workshop connected to
some host conference (CAiSE, OTM, DESRIST).
• In May/June 2011 the first independent EE Working Conference will
be organized (in Antwerpen).
• Since several years there is a special scientific book series in
Enterprise Engineering, published by Springer.
• Since several years there is a ‘sub series’ of LNBIP (published by
Springer), called “Advances in Enterprise Engineering”.
Jan L.G. Dietz: Lecture ISD 2010 - slide 36
Links and references
Reports on research projects can be found on:
www.ciaonetwork.nl
Reports on practical projects can be found on:
www.demo.nl
You can contact me at: j.l.g.dietz@tudelft.nl, jan.dietz@sapio.nl
Dietz, J.L.G., Enterprise Ontology – Theory and Methodology, Springer Verlag,
2006
Dietz, J.L.G., Architecture – building Strategy into Design, Sdu, 2008
Hoogervorst, J.A.P., Enterprise Governance and Enterprise Engineering,
Springer Verlag, 2009
Jan L.G. Dietz: Lecture ISD 2010 - slide 37