Soft System Structures
John Boarder
Cartref Consulting Systems
Beckley,
OXON, UK, 0X3 9SS
Tel. / Fax.: (44)-1-865-351475
E-mail: JohnBoarder@ cartref.com
Web Site: www.cartref.com
Abstract.
In this paper we adapt the usual definitions on which soft system approaches are based
in favour of those that allow us to unfold a ‘closed’ enterprise into an ‘open’,
dimensioned, layered, class-based and valued system of influences derived from
expressions of interest or concems of stakeholders. The approach is formal but we
present the structures informally using a simple, novel, unfolding technique; the method
of our soft system approach is developed elsewhere. A fictional example, based on
airport security, is used to illustrate the structures and an example system archetype is
interpreted. The intention is to make soft system structures more mechanical and
therefore perhaps more acceptable to wider system engineering audiences.
INTRODUCTION
Soft approaches to systems engineering based on General Systems Theory (Skyttner,
1996) led to the methods of (Checkland, 1981), (Hitchins, 1992) and (Senge, 1996). In
effect, these methods interpret system stakeholder viewpoints and develop responses
that lead to system improvements. The approaches are structured and semiformal.
However, the procedures and structures are based on differing definitions of concepts
such as boundary, environment, process, etc. These definitions led us to ask:
Is there a common structured approach for system development, a structured approach
that integrates soft and hard methods and which facilitates wider acceptance of “ soft”
concepts and their application in Systems Engineering?
Our practice suggests a common structured approach exists; it is novel but readily
apparent. The structures parallel and overlay supply chains in business. The approach
formalises interventions using common, structured, simple questions. From captured
stakeholder responses, ‘valued’ systems of influences over the supply chains are derived
and interpreted. The same simple questions and structures develop stakeholder
requirements and a structured method statement for their achievement. While we do not
present the complete model, we only address the structures here, we hope to stimulate
discussion and we present the formal relational model in the appendix. To support our
development, we briefly highlight some essential concepts. Then, we present our
altematives and illustrate them using an example fictional case study from Airport
Security and an interpretation of a system archetype.
Our common structured approach is to be defined formally elsewhere. Here, we use a
simple demonstration technique that “unfolds” our enterprise structures from
stakeholders viewpoints. Without loss of generality, the technique “unfolds”
multidimensional and multi-layered, supply chain structures. These structures are
relations over literature-based classes of resources and their properties, processes and
their characteristics. Directed and valued influences, on which representations of the
state of real world enterprise are based, overlay the structures
ON SOFT SYSTEM STRUCTURES
We highlight aspects of soft system structure definitions that led us to feel that there is
something not quite right. We have used the methods referred to and they have
benefited our students, our clients and ourselves. However, definitions of system,
environment, boundary, emergence, hierarchy, communication and control, etc., that
support the methods, conflict with our experience and increase intervention costs and
risk.
On system environment and boundary
In General System Theory, (Skyttner, 1996), system environment is separated from
system. Indeed systems exist within an environment separated by a boundary.
“In order to define the system’s environment its boundary must be defined. The
boundary surrounds the system in such a way that the intensity of interactions across
this line is less than that occurring within the system. ... boundaries possess a coding
and decoding property ... As systems do not always exist boundary to boundary, the
concept of interface is necessary to denote the area between the boundaries of
systems.”
For us these definitions raise at least two questions:
* How do we meaningfully explore a system boundary and why should we? If
interactions at the system boundary are relatively low there may be little of interest
there. Of greater interest are interactions across high traffic interfaces. But that
raises the question.
¢ What is an interface? If an interface exists between boundaries of systems,
according to the definition, it lies in the environment. Is an interface then another
system of the environment? If so, how do systems communicate through such
interface systems? There appears to be a regression of isolated systems.
On holons and emergence
(Checkland, 1981) characterises Soft System Methodology as an enquirer using a
‘holon’ in a system of ‘holons’. Holon is an abstract term used to convey that a system
“has emergent properties, has a layered structure and processes of communication and
control which in principle enable it to survive in a changing environment”. The term
‘holon’ is said to be misused when intended to convey wholeness about a system rather
than a system being conceived as a whole by an enquirer. To us the term ‘holon’ masks
an essential duality of system, that of the relation between the system as a whole, as
attributed by an observer, and relations among system’s parts that integrate it with parts
of surrounding systems.
On emergence, emergence is said to be “the principle that whole entities exhibit
properties which are meaningful only when attributed to the whole, not to its parts”. In
the examples given, properties, claimed to be emergent, are characteristics of human
bodily processes. To desensitised individuals such objects do not have these properties.
Thus, it is attribution by an observer that assigns properties from the human frame of
reference onto remote objects. When we consider the partial viewpoints in stakeholder
expressions, this explains the difficulty in attempts to construct emergent properties of
systems from emergent properties of their parts.
On terminology and influences
(Hitchins, 1992) explores system interactions and potential interactions with newly
introduced systems-of-interest. The N-Square chart of (Lano, 1979) is used to explore
influences among system components. In using the chart we experience difficulties with
inconsistent terminology in real systems. It is difficult to transfer experience from one
investigation to another because of terminological inconsistency. Perhaps this is useful.
It requires reconsideration of terminology. It can reduce the risk of erroneous transfer of
concepts between applications. However, concepts derived from one intervention
should be reusable and guidance on terminology should be provided.
On generic structures and system archetypes
In (Senge, 1996) ‘Systems archetypes or generic structures embody the key to learning
to see structures in our personal and organisational lives’. The essential feature of
archetypal systems is the influence concept; some aspect of a system influences,
perhaps after a delay, some other aspect of a system. Influences cause some system
properties to be reinforced and others to be eroded. Understanding relatively few
generic influence structures can be a useful management learning technique.
For us, system archetypes suffer from a lack of regularity and inconsistent terminology.
Archetypes are explained as influence processes operating over a mixture of entities,
processes, properties, conditions, resources within a broader system. No consistent basis
for archetypal structures is given.
On Thinking, Practice and Engineering
Summarising the above, we find that while soft systems structures are extremely useful,
their conflicting, inconsistent terminology, their lack of regularity, reuse and
repeatability across applications, and the difficulties of attribution, reduce their strengths
and introduce weaknesses so that opportunities for integrated approaches are often seen
more as threats to existing thinking, practice and engineering than as aids to competitive
advantage.
A ‘SOFT’ VIEW OF SYSTEM ENGINEERING OR ENTERPRISE
DEVELOPMENT
Stakeholders, Entropy and Enterprise Development
We start our definition of soft system structures with the stakeholders to an enterprise.
This allows us to retum to more mechanical definitions of structures. Instead of
separating observers from enterprises under observation, we involve observers in order
to reduce the effects of attribution. We set aside the definition of system, and discuss the
issues of system engineering under the heading enterprise development. We offer the
following.
¢ Anenterprise is a name for some entity that is of benefit to stakeholders
¢ Stakeholders are themselves enterprises who benefit from a common enterprise
¢ Stakeholders have varying degrees of interest or concern in their common enterprise
¢ Enterprise development is the satisfaction of stakeholder interests or concems in
their common enterprise
These four statements suggest that enterprise development is guided by two overall
aims:
1. elaboration of the common enterprise according to the needs of its stakeholders
2. elaboration of a means to develop the common enterprise to satisfy the needs of its
stakeholders
We discuss the means for satisfying these aims elsewhere where we present our
approach but both aims introduce the concept of an ‘open’ versus a ‘closed’ enterprise
and allow us to discuss the structures of our approach. Thus:
¢ An enterprise is either ‘open’ or ‘closed’ according to viewpoints of its stakeholders,
their level of abstraction and the level of disorder they are prepared to accept.
Disorder
Enterprise 'open'
Stakeholder Unacceptable level of disorder
Intervention
Acceptable level of disorder
Enterprise 'closed'
I Time
Figure 1: Entropic basis for Intervention
From this definition, an enterprise can be ‘closed’ with respect to stakeholders; see
osed’, we mean stakeholders have no interests or concems for the
enterprise. As far as stakeholders are concerned, the goods and services provided by
their common enterprise satisfy their needs.
Altematively, an enterprise can be ‘open’ with respect to stakeholders. By ‘open’, we
mean stakeholders have interests or concerns. The goods and services provided do not
satisfy stakeholder needs. Such definitions create a Principle of Normality and its
opposite a Principle of Abnormality:
¢ A normal enterprise is a ‘closed’ enterprise
¢ An abnormal enterprise is an ‘open’ enterprise
Hence, enterprise development is itself an enterprise for retuming some common
enterprise in an abnormal, or relatively disordered state, to an enterprise in a normal, or
relatively ordered state. Our structures are therefore ‘entropic’.
By the definition of abnormality, stakeholders only have interests or concems in ‘open’
enterprises. Such enterprises indicate that, at least, there is ‘something not quite right’
there is possibly even something quite wrong, and something needs to done.
Altematively, the ‘something not quite right’ might indicate an advantage to be gained
by developing the enterprise; again something needs to be done. ‘Open’ or ‘closed’
states are not mutually exclusive states. The state of a enterprise is relative to its
stakeholders, their levels of abstraction and their flexibility, their adaptability, in the
face of disorder.
Closed
Enterprise
e { aa »\ Stakeholder
Influences y Enterprise
Open
Enterprise f
ei Unfolding
Influences y Supply
>< Chains
{ Influences 1 tnftuences a Local
‘a ae “™
Figure 2: Unfolding an Enterprise
For example, provided airline passengers, as goods, arrive at the correct gate, in the
correct numbers, of the correct type, with minimal variation in arrival time and of the
correct class, the gate handling staff treat the service of preceding enterprises as closed
enterprises, as normal enterprises. Such enterprises are processing passengers
appropriately such that its outputs satisfy the interests or concems, the needs, of gate
handling staff. However, if gate handling is disturbed by the late or unannounced non-
arrival of passengers, then the preceding enterprises, in the supply chain of passengers,
become ‘open’ for analysis and subject to expressions of interest or concern; something
is not quite right!
Unfolding the ‘something ...’
When enterprises are ‘opened’ to their stakeholders, an unfolding takes place. An
enterprise is unfolded as a duality; differentiated into an ‘environment’ and an ‘activity’.
The unfolding defines enterprise environment and activity as:
* Environment comprises those resources which support enterprise activity
* Activity comprises those processes which support enterprise environment
In the passenger handling example, the environment comprises the resources needed to
support the activity for processing passengers from the departure lounge onto a flight.
The activity comprises the processes supporting the transformation of potential
passengers into actual passengers.
In unfolding the closed enterprise, as well as revealing its partial structure, expressions
of stakeholders reveal potential influences among resources and processes that are the
possible causes of disorder and dissatisfaction; see Figure 2. Also, unfolding possible
influences between the common enterprise and that of a stakeholder illuminates
commercial supply chains and supply chain influences.
A definition of enterprise boundary is offered:
¢ The boundary of one enterprise with another is the collection of resources and
processes the enterprises hold in common
By unfolding, rather than the environment of an enterprise being separated from its
surroundings, as in the system view, the environment of an enterprise is integrated with
stakeholder environments. Integration is a partial function of enterprise activity.
Conversely, enterprise activity is integrated with stakeholder activity. Integration is a
partial function of enterprise environment. Enterprises interact through resources and
processes held in common. In effect, waves of activity sweep along supply chains
supporting waves of change in the environment. Conversely, waves of change in the
environment support further waves of enterprise activity. The issues for enterprise arise
from the influences particular resources in an environment have on particular processes
in an activity and vice versa. Such issues cause disturbance and disorder in the supply
chains.
Management
Support
Historical
Operational
e Resultant
Dimension =
Development
Figure 3: Layers, Dimensions and J ohnston Vectors
Multiple Dimensioned and Layered Enterprises
Refining supply chains, we note that, without loss of generality, there are at least five
supply chains interacting in and being integrated by any one enterprise and that each
supply chain defines a dimension. We propose that the five dimensional supply chains
are:
* operational dimension of material and money transformations as goods and
services
* management dimension of strategic decisions and tactical responses
¢ support dimension of maintenance demands and service responses
¢ development dimension of requirement formulation and project management
¢ historical dimension of defining and auditing the state of the enterprise
Along each of these dimensions environment resources and activity processes mutually
influence each other and therefore the state of disorder and disorganisation.
We used the term ‘level of abstraction’ with regard to stakeholders. By this we mean a
stakeholder viewpoint is with respect to their relative position in a supply chain. Those
stakeholders immediately involved in a local enterprise have an immediate level of
abstraction, they are aware of immediate resource and process issues. Those
stakeholders who benefit directly in extemal and sub level enterprises have a direct
level of abstraction, they are aware of issues in relation to goods and services. Those
stakeholders who are more remote in wider and base-level enterprises have indirect
levels of abstraction; they only have a direct view of issues conceming their own direct
stakeholders and can only infer issues for indirect stakeholders. Indirect stakeholders
only view the enterprise through interactions with their own direct stakeholders. Levels
of abstraction therefore provide a layered view of enterprise; our model is multi-layered
and multidimensional; see Figure 3.
In our structures, influences derived from stakeholder views are directed parallel to
supply chain dimensions and, according to the relative frequency and significance of
those views, such influences have magnitude. (Johnston, 2000), found it beneficial to
study the implication of such influence vectors in developing strategy for engineering
departments. Figure 3 refers to these vectors and their resultant as Johnston vectors.
Since the direction and magnitude of Johnston vectors are derived from stakeholder
views, subject to review and audit, their attribution as emergent states of enterprise is
meaningful.
Resource-Property and Process-C haracteristics classes
Influences at work in a dimension as perceived by stakeholders are influences of
resources over processes and processes over resources. More specifically, influences are
due to the properties of resources and the characteristics of processes. Resources,
properties, processes and characteristics have been classified loosely in the literature.
Jenkins, (Checkland, 1981) defined systems as ‘groupings of men and machines with an
overall objective and characterised by an economic criterion which measures success’,
provides us with three classifications; manpower, machines and money. To these we
would add methods and materials. Enterprise environment comprises collections of
resources in the classes
¢ Manpower: human resources needed to motivate the enterprise
* Materials: resources transformed by the enterprise into Money
* Machines: organisational and other devices used to aid the transformation of
materials
¢ Methods: procedures, definitions and descriptions for the activity of the enterprise
¢ Money: resources transformed by the enterprise into Materials
The operations management discipline (Slack et al, 1998) discusses characteristics that
distinguish operations. We reinterpret these characterisations for properties of resources.
Thus we qualify resources by properties in classes:
Volume: properties that quantify resources
Variety: properties that classify resources
Variation: properties that stabilise resources
Value: properties that qualify resources
Venue: properties that localise resources with respect to subject, place, etc.
Action learners, quality gurus, etc., define enterprise activity using process sequences.
In (Deming, 1986) for example, the Shewhart cycle is given for transformation
processes which we abbreviate to Observe, Study, Analyse and Act; other authors
attribute the terms differently. We reinterpret this cycle to highlight the difference
between preparing an understanding of what we want to act on(our first aim) and
preparing to act (our second aim). We also recognise that in an arbitrary enterprise,
individual processes are concurrent. Therefore enterprise activity comprises collections
of processes in classes:
¢ Observing: capturing and recording data on the enterprise environment and activity
¢ Modelling: assessing observed data and creating information in support of the
decision process
¢ Defining: deciding the resources and processes needed to sustain the enterprise as a
closed enterprise
¢ Organising: planning, scheduling resource availability and process initiation
¢ Interpreting: acting out the organised plans for the enterprise environment and
activity
Again in (Slack et al, 1998) performance objectives are discussed. We use such
objectives to characterise processes. Thus, we define five characteristic classes for
processes:
Quality: characteristics for the qualification of processes
Speed: characteristics for the quantification of processes
Dependability: characteristics for the availability of processes
Flexibility: characteristics for the adaptability of processes
Productivity: characteristics for the capability of processes
In (Slack et al, 1998) the fifth performance objective, is Cost. We have deliberately
adopted the term Productivity to eliminate an ambiguity. Cost is a property attributable
to a resource; a Volume property attributed to a Money resource. Money resources are
exchanged between stakeholders for Materials with appropriate properties. Cost
therefore indicates a direct relation between stakeholders, whereas we want to express
the immediate relationship between input and output resources, the transformation of
Materials and Money, and the capability of transforming processes; Productivity.
Using our definitions we define:
* enterprise activity as:
* The interpretation by manpower and machines of methods for observing,
modelling, defining, organising and acting in the enterprise environment, to
support the transformation of materials into money and money into materials
* enterprise environment as:
* The manpower, machine and method resources that support processes in the
enterprise activity for observing, modelling, defining, organising and acting to
transform materials into money and money into materials
* enterprise as:
* environment resources integrated with activity processes communicating with
stakeholder enterprises through changes in the properties of common resources
and changes in the characteristics of common processes.
Influences at work in an enterprise that cause stakeholders to express interests or
concems are relations between instances of resource classes and properties classes and
instances of process classes and characteristic classes, the instances being rooted in
terminology of real-world enterprise.
Managerial
Operational
Support
Developmental
Figure 4: Tetrahedral View of Enterprise
Finally, in our structure definitions, management and marketing use the technique of
SWOT (Strengths, Weaknesses, Opportunities and Threats) analysis for assessing
business. We classify our influences according to SWOT values. Strengths and
weaknesses are immediate, positive and negative influences. Opportunities and Threats
are direct and indirect, positive and negative influences. Using SWOT values and our
definitions, we have a relational model for real-world enterprise that is structured and
ordered according to dimensioned, layered, class based, directed and valued influences.
In general our valued influences have the structure:
¢ dimension, layer, resource, property, influence, dimension, layer, process,
characteristic
¢ dimension, layer, process, characteristic, influence, dimension, layer, resource,
property
Using relative frequencies and levels of significance of stakeholder viewpoints, the
relational model, see appendix, allows a classified form of N-Square Chart to be used to
capture enterprise state. From this chart, representations of enterprise state can be
displayed. Figure 4 illustrates dimensional state and can be adapted to include layered
states. Ideally, a closed enterprise N is positioned instantaneously and historically at the
equilibrium point, equidistant from the four vertices. Figure 5 is a SWOT matrix setting
immediate against direct and indirect influences for a given supply chain dimension.
Ideally, the Centre of Value is at the origin so that the ideal enterprise has no
superfluous strengths, all weaknesses have been addressed, there are no unsatisfied
opportunities and no unmitigated threats. If the Centre of Value is at the origin for each
of the five dimensions, the stakeholders have no interests or concems for their common
enterprise and it is effectively closed. However, some residual strength, some residual
opportunity, is beneficial in allowing the enterprise to weather transient weaknesses and
threats.
Strength
Center
of
Value
Threat Support
Weakness
Figure 5: SWOT Matrix representation of influences
THE AIRPORT SECURITY EXAMPLE
We outline an example enterprise, illustrative of those on which our model is based.
The example is airport passenger security and only illustrates structural definitions.
Passengers, on arrival at an airport, check in at an airline desk and proceed to the
departure lounge via airport security. At this point, passengers form orderly queues and
proceed one at a time to deposit hand luggage on the X-ray scanner conveyor and their
personal metal items in a visible tray. Passengers proceed through archway scanners and
their hand luggage is X-rayed. If alarms trip, randomly, according to passenger
profiling, or because suspicious material is detected, passengers or their luggage
undergo detailed scrutiny. Secure passengers proceed to the departure lounge, insecure
passengers are detained. Uncleared passengers form a common resource for check-in
lounge and security check. Secure passengers are a common resource for security check
and departure lounge. Insecure passengers are a common resource for security check
and police services. The operational supply chain boundary for the security check
enterprise is therefore well-defined.
Non-exclusively, resources for this localised enterprise include:
* Manpower: operations staff, management staff, maintenance staff
* Material: passengers, electrical power, luggage
* Machine: archway-scanners, X-ray scanners, hand-scanners, conveyors
¢ Method: security procedures, equipment operations procedures, standing orders,
search techniques
¢ Money: departure-lounge space, slots in the check-in queue for security, security
incidents
Properties of such resources include:
* Volume: passenger numbers, staff numbers
* Variety: passenger types, airport security state
* Variation: passenger size, queue size, passenger density, emotional state, (e.g. staff
fatigue)
* Value: contributions to airport security, airport use
* Venue: ground side, air side locations
Airport security environment is a mix of such resources and properties.
The activity processes for this localised enterprise include:
* Observing passenger shapes and sizes, luggage contents, tray contents
Modelling displayed shapes, sizes and colours against prohibited items
Defining real and potential security breeches
Organising and allocating staff
Interpreting procedures and conducting searches
Characteristics of these processes include:
* Quality: securing the correct passengers and securing them correctly
¢ Speed: rate at which passengers are processed
¢ Dependability: availability of the airport security service, their effectiveness in
detecting security breeches
¢ Flexibility: ability to react to incidents
¢ Productivity: number of passengers processed in unit time, the number of incidents
Airport security activity is a mix of such processes and their characteristics.
Using the defined influence forms:
¢ dimension, layer, resource, property, influence, dimension, layer, process,
characteristic
¢ dimension, layer, process, characteristic, influence, dimension, layer, resource,
property
We interpret possible stakeholder comments:
¢ staff fatigue influences passengers processed per hour
* operational, local, manpower, variation, weakens, operational, local, defining,
productivity
passengers per hour influences departure-lounge-queue size
* operational, local, defining, productivity, strengthens operational, local, money,
variation
large departure-gate- queue sizes influence security decision accuracy
* operational, local, money, variation, weakens, operational, local, defining,
dependability
reduced numbers of passengers to be security checked influences security decision
accuracy
* operational, local, material, volumes, strengthens, operational, local, defining,
quality
improved security decision accuracy influences security staff fatigue
* operational, local, defining, quality, strengthens operational, local, manpower,
variation
reduced staff fatigue influences the number of passenger processed per hour
* operational, local, manpower, variation, strengthens, operational, local,
interpreting, productivity
We note that, while the example draws only from the operational dimension, similar
examples apply in other dimensions. The example illustrates:
Resources, properties etc., drawn from general classes
Transitive influences: influences form chains and loops
Symmetric influences: resources influence processes and the same processes can
influence the same resources
Anti-reflexive influences: resources do not immediately influence resources,
processes do not immediately influence processes
Influences as instances of a general class of statements in a language with a formal
structure
We note that this example illustrates a supply chain that conveys passengers from
airport portals to seats on aircraft. It involves operational, managerial and support
aspects which affect the airport at different layers. Airport security is the subject of
development according to current national security requirements and has a history that
has contributed to and influenced its development. Further,
Operational stakeholders include check-in enterprises and departure lounge
enterprises
Managerial stakeholders include operations, strategic and corporate management of
airport security and management of staffing schedules and availability tactics
Support stakeholders include equipment maintenance, human resource, financial
resource planning, airline security, police co-ordination, etc. and preventative
maintenance, financial accounting and staffing, power supplies and distribution,
health and safety, etc.
Developmental stakeholders include passenger representation groups, airlines,
planning groups and equipment suppliers, human factors engineering, power
distribution, etc.
¢ Historical stakeholders include all those agencies that having served and been
served by airport security and all those passengers, equipment suppliers, previous
staff, have created the preconditions for its current state.
ON SYSTEM ARCHETYPES
We give one example, the Fixes that Backfire System Archetype, in the form of our soft
system structures. In (Senge et al, 1996), the archetype is described as representing the
situation in which ‘whoever makes the most noise gets the most attention’ regardless of
real need. A diagram of a similar archetype is given in Figure 6. Interpreting the
diagram, using influence statements, we have:
* Problem Symptom influences Process Fix;
¢ Process Fix influences Problem Symptom
¢ Process Fix influences Side Effects;
¢ Side Effects influences Problem Symptom
S Delay
Side 2
Effects
Figure 6: Fixes that Backfire
However, there is no information to distinguish what Problem Symptom, Fix or Side
Effects are. We therefore make the following definitions. A Problem Symptom is a
resource property. Some resource is exhibiting a problem. A Fix is a process
characteristic. By adjusting characteristics of processes we seek to solve problem
symptoms; to return the resource exhibiting this property to a normal state. Side effects
are unintentional consequences of changing the characteristics of processes. Side effects
are therefore either characteristics of processes that arise immediately or they involve
chains of resources and processes that deliver the unintended consequences later. We
therefore interpret the archetype in the form of Figure 7, where the side effect sequences
can be of arbitrary length.
¢ Resource Properties influence Process Characteristics
¢ Process Characteristics influence Resource Properties
¢ Process Characteristics influence Side Effect Resource Properties
¢ Side Effect Resource properties influence Side Effect Process Characteristics
¢ Side Effect Process Characteristics influence Resource Properties
We note that delay is represented by cumulative process execution times. Generalising,
and interpreting Problem Symptom and Process Fix as an Enterprise, such
interpretations lead us to ask, before any change is undertaken:
¢ What will my Enterprise be?
¢ Who will my Enterprise benefit? Who are its stakeholders?
¢ What will my Enterprise do? Both intentional and unintentional
¢ What will my Enterprise use? What resources and their properties will be affected?
¢ How will my Enterprise be transformed?
¢ How will the intentionally and unintentionally transformed Enterprise influence
other Enterprises?
Finally, we note that these questions are the basis of structured and ordered enterprise
developments.
Neo
Side Effects
Figure 7: Structured System Archetype
CONCLUSIONS
Our approach to soft system structures started with the feeling that there is something
not quite right with the definitions and examples underlying system thinking, system
practice and system engineering. We sought and, we believe, found a rational and more
formal basis for interventions in enterprise development. Our concems led us to a soft
system structure that helps overcome issues of conflicting, inconsistent terminology,
lack of regularity, reuse and repeatability across applications, and difficulties raised by
attribution. In consequence, in our own soft system practice, enterprise development, we
have increased our strengths and reduced our weaknesses so that our soft system
approach offers more opportunities for enterprise integration, is seen less as a threat to
existing thinking, practice and engineering and as a greater aid to competitive
advantage, risk reduction and cost reduction.
REFERENCES
Checkland, P. 1981. System Thinking, System Practice. John Wiley & Sons
Deming, W. E.1986. Out of the Crisis. MIT CAES
Hitchins, Derek, K. 1992. Putting Systems to Work. John Wiley & Sons
Johnston, M. 2000. Enterprising Processes - The future of Engineering Departments
across Scottish Airports, MBA Thesis, Surrey University
Lano, RJ.A. 1979. Techniques for Software and System Design. TRW Series on
Software Technology. Vol. 3; North Holland Publishing Co.
Senge, P., Ross, R. Smith, B., Roberts, C., Kleiner, A. 1996. The Fifth Discipline
Fieldbook. Nicholas Brealey
Slack, N., Chambers, S., Harland, C. Harrison, A. Johnston, R. 1998. Operations
Management. 2"™ Ed. Pitman Publishing
APPENDIX: A RELATIONAL ENTERPRISE MODEL
Defined bottom-up:
Resources R = {Manpower, Materials, Machines, Methods, Money}
Properties V = {Volume, Variety, Variation, Value, Venue}
Environment E.=R %V
Processes P = {Observing, Modelling, Defining, Organising, Interpreting}
Characteristics C = {Quality, Speed, Dependability, Flexibility, Productivity}
Activity A. =P %C
Layers Z = {Wider, External, Local, Sub, Base}
Layered Environment Zp. =Z %(R %V)
Layered Activity Zy.=Z %(P %C)
Dimensions X = {Operational, Managerial, Support, Developmental, Historical}
Dimensioned, Layered, Environment Xz. =X % Zz.
Dimensioned, Layered, Activity X,.=X %Za.
Influences I = {Threaten, Weaken, Normal, Strengthen, Support}
Resource Influenced, Dimensioned, Layered, Activity
Kea =(1 % (Xe. %Xa-))
Process Influenced, Dimensioned, Layered, Environment
Kaz =(1 % (Xa. % Xe)
Enterprise N = Union (Kg-4, Ka-z)
BIOGRAPHY
John Boarder is an independent consultant. A BCS Chartered Engineer and member of
the IEE, John contributes, to many system engineering events. Operating a research
based consultancy, he lectures and supervises research in Computer Science, System
Engineering and Operations Management. Material presented here is derived from
practice and CMS, DMS, MBA courses. A co-author of the IEE Draft Guide to the
Practice of Systems Engineering, John is a member of INCOSE UK.
END NOTE
This paper ‘Soft System Structures’ was first presented at the Spring Symposium and
Tutorials of INCOSE UK, 2001. This is a reformatted version of that paper with minor
corrections.
CC BY-NC-SA 4.0