PROCESS

1
PROCESS

• Reading Assignment
• Supplementary module 5
• PROCESSES, EVENTS TEMPORAL RELATIONSHIPS

2
DISCRETE CHANGE

• ATTRIBUTE VALUES RELATIONSHIPS CHANGE IN
  RESPONSE TO DISCRETE EVENTS
• CONSTRAINTS ON ENTITIES CHANGE IN RESPONSE TO
  DISCRETE EVENTS

• RELATIONSHIPS INVOLVE A FINITE SET OF ENTITY
  INSTANCES IN FINITE AND DISCRETE POINTS IN TIME

3
TEMPORAL RELATIONSHIPS

• The sweep of time makes temporal relationships
  special
• Time is a (dense) ordinal domain
• Time cannot be reversed
• The ordinality of a temporal relationship
  increases the information content of its
  non-temporal parent(s)
• The moving finger, having writ, moves on
• States of past time slices are frozen
• Temporal relationships recognize the
  irreversibility and direction of time

CARDINALITY CONSTRAINTS ON TEMPORAL RELATIONSHIPS
4
CARDINALITY CONSTRAINTS ON TEMPORAL RELATIONSHIP
Examples of constituent Object Classes
Examples of temporal combinations (which might be
constrained)
retailer
customer
product
Time Slices
Past
Flow of time
HIGH ORDER TEMPORAL RELATIONSHIP
Present
Sold to..
buys thru..
Examples of Components of a tuple
possible combinations of time slices for which
constraints may be defined over three state
changes
m..n
Temporal Matrix
m..n
possible combinations for which constraints may
be defined in a three-way relationship
5
TEMPORAL RELATIONSHIPS

• A relationship infused with time is a process
• Because it conveys before and after information
  Resources come before products
• This is the basis of causality The meaning of
  cause and effect is created thus
• Processes inherit the properties of relationships
• And create their polymorphisms
• Just as objects and relationships created
  polymorphisms of the basic semantics of Pattern
  in step with added information
• In this section we will discuss these properties
  of processes

RELATIONSHIP
Object
Object
Subtype of
Subtype of
precede succeed
Resource
Product
Subtype of
Subtype of
Subtype of
used by 0 or more
produced by 0 or more
PROCESS
use 1 or more
Produce 1 or more
6
PROCESS

• Resources may be consumed or referenced by a
  process
• Referenced resources are catalysts or
  facilitators
• Consumption is equivalent to reducing the
  capacity of the resource to engage with the
  process
• Consumption is a temporal polymorphism of
  capacity
• Capacity to engage is context sensitive
• May apply to all or some processes.
• Capacity to engage, that depends on the process
  engaging the resource, is a polymorphism of
  generic capacity to engage
• Resource life may be conceived as consumption of
  a resource by an idempotent process
• Eg drugs expire, limited time offers, perishable
  items, agreements with term limits
• Events may also be resources
• A trigger for a process is a role of event. The
  triggering event is a resource.

A.
Resource Expiry time for process
Beginning of Resource life for process
RESOURCE LIFE
Used within 0 or more valid Valid for 1 or more
use
B.
Subset of
Used by 0 or more (an event) use 1 or more
Produce 1 or more produced by 0 or more
PROCESS
Trigger 0 or more triggered by 1 or more
Event
7
ENQUIRY/OBSERVATION

• The enquiry is a process
• Synonym for observation
• The work product of observation is information
  The state of an object
• The state of the object instance will change from
  unobserved to observed
• States of the instance observed may also change
  from Unknown to a known state with specific
  values for observed properties
• The observation may even change the states of
  known properties
• Eg Management attention might reduce defect
  rates, even if the management makes no changes to
  the observed process

8
PROPERTIES OF PROCESSES Cycle time

• Cycle Time The time lapse between the two ends
  of a temporal relationship
• Must all processes begin and end?
• Processes may be sagas
• Patterns of infinite extent in one or both
  directions in time
• Eg the process of routinely loading and
  dispatching trucks from a factory, radioactive
  decay
• Start and finish may be Unknown values

No
A few surprises
Could also be aggregations or compositions

• Interrupt

Events too have history
(re)

• Cancel

9
PROPERTIES OF PROCESSES Inverse

• Temporal inverses, reversibility and reversion
• Temporal polymorphism of the inverse of a
  relationship
• The flow of time adds a new dimension to the
  inverse the concept of reversibility of a
  process
• Eg we cannot unbake a cookie to produce the
  dough it was baked from
• The process is irreversible
• We can freeze the water we melted
• The melting process for water is reversible
• Although the inverse reverses a relationship, it
  still has to go forward in time and cannot erase
  the history of the objects it relates.
• Reversion may not be absolute, accurate,
  reliable, complete or valid
• Extent to which the reversion recovers consumed
  resources may vary by resources consumed
• Usually less, may be equal or more (maps to
  completeness of reversal with respect to a
  resource), may produce equivalent but not
  identical resources (extent of validity of
  reversion)
• Polymorphism of cardinality constraints
• Number of reversals permitted (non-temporal
  polymorphism)
• Number of simultaneous reversals permitted
  (temporal polymorphism)
• Number of consecutive reversals permitted
  (temporal polymorphism)
• Cardinality constraints may involve temporal and
  non-temporal dimensions of the relationship

10
PROPERTIES OF PROCESSES Recursion, Reflexivity,
Idempotency

• Temporal recursion, temporal reflexivity and
  temporal idempotency
• The properties of recursion, reflexivity and
  idempotency may apply independently to an
  instance of the object, to time slices of
  objects, or simultaneously to both
• A class recursive temporal relationship also
  loops back to the same class, but perhaps
  different time slices of the same, or different
  object instances
• If it is also time slice reflexive, it may (but
  does not always have to) connect the same time
  slice
• If time slice irreflexive, it cannot do so
• If time slice idempotent, then it must do so.
• A class irreflexive temporal relationship cannot
  loop back to the same object instance.
• A time slice irreflexive temporal relationship
  cannot loop back to the same time slice.
• An idempotent temporal relationship loops back to
  the same object instance, but
• May be time slice irreflexive it may have to
  connect different time slices of the same object
  instance
• Concept of temporal order, ie how far into the
  future do the influence of events reach, is a
  polymorphism of this
• May be time slice reflexive, i.e., it may be
  permitted to connect the same time slice of the
  same object, but is not required to always do so.
• When a relationship loops back to the same time
  slice, there is no passage of time it cannot be
  a process because there is no before and
  after.
• A totally idempotent temporal relationship must
  always connect the same time slices of the same
  object instance.
• It cannot be a process
• A time slice idempotent temporal relationship
  must always relate concurrent time slices of
  object instances
• The object instances could belong to the same or
  different object classes.
• Cannot be a process
• A reflexive relationship may or may not loop back
  to the same time slice of the same object
  instance, and may or may not loop back to the
  same object instance, provided it loops pack to
  an instance of the same class (for example, a
  person may be his own counsel in a court of law
  at one time, and be represented by someone else
  at a diffent time the representation
  relationship loops back to the class of persons
  in both cases).

11
Polymorphisms of Relationships
RECURSIVE
RECURSIVE
RELATIONSHIP
RELATIONSHIP
RETURN
OBJECT
OBJECT
12
PROPERTIES OF PROCESSES Symmetry

• A symmetrical process is a meaningless concept
• Time is asymmetrical
• Processes always connect resources to products
  over a forward span of time
• All processes must be asymmetrical
• Reversion of the process is its inverse
• Symmetry requires the inverse and the
  relationship be identical
• This is meaningless for processes Reversion of a
  process can never be the same as the process it
  reverses.
• That a process does not change over time is a
  form of temporal symmetry
• The property of stationarity
• Non-stationary processes will change their
  characteristics (properties) over time

However..
PROPERTIES OF PROCESSES Mutability

• Temporal mutability When mutability changes over
  time
• Eg for a limited time, the temporary and
  permanent license plates of a new car are legally
  mutable.
• Temporal mutability may involve combinations of
  time slices

• A polymorphism of non-stationarity

13
PROPERTIES OF PROCESSES Order

• Temporal order How remote a history will affect
  behavior
• Becomes quite complex for processes
• How many time slices a relationship spans for
  each object class
• How many time slices a relationship spans for
  each object instance

• Combinations of time slices Which time periods
  are combined
• Object instances
• Object classes
• Both

• Overall Temporal order Related time slice
  farthest back in history
• If only the current state impacts state
  transitions , temporal order 1
• If neither current state, nor history matter,
  temporal order 0

14
PROPERTIES OF PROCESSES Degree

• The temporal degree of a relationship with
  respect to a participating object instance is the
  number of distinct time slices of that object
  instance the relationship involves
• Temporal degree of an idempotent relationship
  No. of times a relationship loops back to the
  same object instance
• Overall temporal degree of a relationship is the
  total number of time slices of all objects that
  the relationship involves
• The relationship can only loop forward in time
• May loop across object instances or classes

Recursive relationship temporal degree 2.
Idempotent relationship Temporal degree with
respect to object 3 2
Past

• Constraints on temporal degree limits the length
  of tuples of time slices

Instance 1
Instance 3
Instance 2
V1
V1
V1
V2
V2
V2
Time
Time
V3
V3
V3
V4
V4
V4
Present
OBJECT CLASS

• Combinations of time slices How many are
  combined over which time periods
• Object instances
• Object classes
• Both

Relationship across classes temporal degree 2.
OBJECT
15
PROPERTIES OF PROCESSES Degree (2)

• Can distinct time slices be concurrent?
• Yes!

THE PROCESS

• Sign

• Sign

• Sign

• Sign

• Provided they are time slices of the same object
  in different states
• Each state is a polymorphism of the check

Payable
Check

• Same check seems to imply 1st degree, but a
  2-tuple is involved two instances of the
  signature process
• Second temporal degree, idempotent relationship

• Parallel process signatures may or may not be in
  the same time slice
• Idempotent relationship of the 2nd degree, but
  both loops occur in parallel

• Scenario Both the CEO and CFO must sign a check
  for it to be payable

16
See Case Study on reusing and modifying process
knowledge in supplementary Module 5
Are the sub-processes that constitute the check
signing process also subtypes of the check
signing process when considered individually?
PROCESS MAP
17
possible combinations of time slices for which
constraints may be defined over three state
changes
Class 2
Class 1
Class 3
Instance 1
possible combinations for which constraints may
be defined in a three-way recursive relationship
across Instances
Instance 2
m..n
Instance 3
PRESENT
m..n
PAST
m..n
m..n
possible combinations for which constraints may
be defined in a three-way relationship across
classes
18
PROPERTIES OF PROCESSES Cardinality Ratios

• Temporal Cardinality Cardinality of combinations
  across time slices
• Batch Processes The cardinality ratio of a
  single cell describes how many items of an object
  class were simultaneously involved in an instance
  of the process at the time

• Cardinality ratio of 2 or more on produce
  batch process
• Batch Size If the object class is a product of
  the process, the cardinality rato of the cell
  that represents the contemporary time slice tells
  us how many items of the product a single
  instance of the process produces at the time
• This is the cardinality ratio of the produce
  relationship below
• Resource requirement If the object class is a
  resource, it tells us how many items of the
  resource are required at a time by an instance of
  the process

Cardinality Ratio
Cardinality Ratio
RELATIONSHIP
Object
Object
Subtype of
Subtype of
precede succeed
Resource
Product
Subtype of
Subtype of
Subtype of
used by 0 or more
produced by 0 or more
PROCESS
use 1 or more
Produce 1 or more
19
Concurrency vs Repetition
PROPERTIES OF PROCESSES Cardinality
FLOW OF TIME
Repetition (cardinality of the process across
time slots)
Bake Cookie
Bake Cookie
Bake Cookie

• PAST

• FUTURE

(2)
(3)
(4)
Concurrency at given moments (Concurrency
cardinality of the process at a given time)

• Non-temporal cardinality does not care about the
  (unknown) flow of time
• total cardinality

20
The meaning of Cascade
PROPERTIES OF PROCESSES Cardinality
FLOW OF TIME
Cardinality Ratio
Bake Cookie
Bake Cookie
Bake Cookie

• PAST

• FUTURE

(1)
(4)
Number of successors per predecessor (Cascade
ratio Number of processes triggered by each
predecessor)
21
PROPERTIES OF PROCESSES Efficiency/Productivity

• Efficiency with respect to a given resource is
  the ratio between the cardinality ratio of the
  product and the cardinality ratio of the resource
• For higher order relationships, the cardinality
  ratios of other resources are held at a fixed
  level

Divide by
Product
Resource
Efficiency
use 1 or more
Produce 1 or more
used by 0 or more
produced by 0 or more
PROCESS
Cycle Time
Productivity
Divide by
22
PROPERTIES OF PROCESSES Capacity

• Temporal capacity similar to non-temporal
  capacity
• Cardinality upper bounds Limited capacity to
  produce product or use resource
• May vary by object class, instance or time slice
  or any combination of these
• Eg an instance of a razor blade engaged in a
  shaving process cannot be used by any other
  process concurrently (capacity for concurrent
  engagement)
• Eg A Person may concurrently be engaged in a
  maximum of 4 projects at a time (capacity for
  concurrent engagement)
• Eg Razor blade may be good for only 5 shaves
  (Capacity for repetition)
• How much of a resource is engaged or consumed is
  normalized by the use relationship between the
  resource and the process
• What kind of capacity is normalized by the
  produce relationship?
• Is a process a polymorphism of a product?
• Is the product a polymorphism of a goal?

23
PROCESS DECOMPOSITION
RESOURCE
Consumed by
produce
EXTERNAL EVENT
WORKPRODUCT
trigger
BAKE COOKIE
A.
ANOTHER EXTERNAL EVENT
B.
precedes
precedes
precedes

• A process is a polymorphism of a succession
  constraint
• Succession constraint forges the meaning of
  precede
• Subprocess is different from the subtype of a
  process
• Subprocess is a part of a composite process
• Together, the subprocesses imply (are transitive
  with) the process
• Eg Arrange Dough Glob is a subprocess in Bake
  Cookie
• Eg A subtype of Bake Cookie might have been
  Bake sugar free cookie

24
CONDITIONAL EVENTS MUTUAL INCLUSION
Parallel Processes
Precedence Diagramming Method (PDM Notation from
PMI)
COMMONLY USED NOTATIONS
Topos of collaboration
PICK ITEMS
TAKE ORDER
SHIP ITEMS WITH INVOICE
Universal Modeling Language (UML notation from
OMG)
RAISE INVOICE
25
Predecessor
Predecessor
Cardinality Ratio 1
Successor
(with respect to successor) Every instance of its
successor is triggered by a single instance of
the conjunction
26
CONDITIONAL EVENTS MUTUAL EXCLUSION
TOPOS OF CONFLICT

• Mutually exclusive relationships define the topos
  of conflict
• The conflict here is not between employer and
  candidate, but between making or refusing an offer

MAKE OFFER
INTERVIEW CANDIDATE
MUTUALLY EXCLUSIVE SUCCESSION OF EVENTS
X
ISSUE REGRET LETTER
27
Example of mutually exclusive temporal
compositions
TOPOS OF COLLABORATION
Topos of conflict
TOPOS OF COLLABORATION
28
MAKE OFFER
COMPOSITION
ACCEPT CANDIDATE
INTERVIEW CANDIDATE
MUTUALLY EXCLUSIVE SUCCESSION OF EVENTS
X
REJECT CANDIDATE
COMPOSITION
ISSUE REGRET LETTER
29
CONDITIONAL EVENTS SUBTYPING OF SUCCESSION
Belongs to the topos of collaboration or not?

• Topos of asymmetric collaboration

OPEN NEW PROBLEM
Succeeded by 0 or more succeed 1
ANALYZE ISSUE
TAKE SERVICE CALL
SUBTYPE OF
Succeeded by 0 or more succeed 1
Severe new problem
ALERT MANAGEMENT
30
MORE COMPLEX CONDITIONAL EVENTS

• Complex occurrence constraints can stem from
  constraints on the degree and order of the
  relationship in three dimensions
• Complex occurrence constraints may also tie
  states of instances and classes of processes
  together with occurrence, precedence and
  succession constraints
• Some simple examples are

• A processes that must be suspended if another one
  has started
• A process that must be started if another one is
  cancelled, etc.

• Interrupt

• Cancel

• Automate validation of constraints and
  interdependencies in large and complex temporal
  or causal networks
• Check for and eliminate unintended side effects
• Deadly embraces
• Validate null spaces

31
ADD LATENCY CONSTRAINTS TO THE MIX
BEGIN
END
BEGIN
END
BEGIN
END
Delay
Delay
Cycle Time
Successor
Delay
Delay
Delay
Delay

• More complex themes
• Waiting period (with each kind of occurrence
  constraint)
• Eg 3 days grace period for a contract
• Optional or mandatory mutual exclusion, mutual
  inclusion with or without time delay or within a
  time slot defined in terms of the start or finish
  of a process
• Processes that must start when one or more of its
  predecessors end without time delay or within a
  deadline (inclusion constraint)
• The subtyping constraint Processes that may
  start only after its predecessor ends without
  time delay or deadline (inclusion constraint)

Cardinality Ratio 1

• A few Common themes

32
EXAMPLE OF LATENCY
Cycle Time
Cycle Time
Delay
Cycle Time
Delay
BEGIN
END
BEGIN
END
BEGIN
END

• FLOW OF TIME

33
EXAMPLE OF LATENCY
BEGIN
END
BEGIN
END
Cycle Time
Cycle Time
Delay
Delay
Cycle Time
(Begin Delay)
BEGIN
END

• FLOW OF TIME

34
EXAMPLE OF LATENCY
BEGIN
END
BEGIN
END
Cycle Time
Cycle Time
Delay
Delay
Cycle Time
(Begin Delay)
BEGIN
END

• FLOW OF TIME

35
EXAMPLE OF LATENCY
BEGIN
END
BEGIN
END
Cycle Time
Cycle Time
Delay
Delay
Cycle Time
(Begin Delay)
BEGIN
END

• FLOW OF TIME

36
CONDITIONAL EVENTS SUBTYPING OF SUCCESSION
OPEN NEW PROBLEM
Succeeded by 0 or more succeed 1
ANALYZE ISSUE
TAKE SERVICE CALL
SUBTYPE OF
Succeeded by 0 or more succeed 1
Severe new problem
ALERT MANAGEMENT
37
OPEN NEW PROBLEM
Example of a transitive process
Succeeded by 0 or more succeed 1
ANALYZE ISSUE
TAKE SERVICE CALL
SUBTYPE OF
Severe problem

• With non-temporal transitive sets of
  relationships, any one of the relationships in
  the transitive set may be removed to normalize
  information
• When temporal information is included, the last
  process in the transitive set should be
  eliminated, because its predecessors will imply
  its occurence

Succeeded by 0 or more succeed 1
Succeeded by 0 or more succeed 1
TRANSITIVE TRIAD
ALERT MANAGEMENT
trigger
38

• Is the product a polymorphism of a goal?
• Is a process a polymorphism of a product?

PURPOSE OF BAKE COOKIE
BAKE COOKIE
RESOURCE
Subtype of
A.
EXTERNAL EVENT
WORKPRODUCT
BAKE COOKIE
Supplementary Module 5 Essence of a Process and
goals of a business
EVENT COMPOSITION

(Change scope?)
39
PRODUCT OF A PROCESS

• Work Product The purpose of the process
• By Product A lower priority goal being achieved
  as a part of the process
• Waste product An item of no relevance or worse
  (undesirable item) produced as a part of
  achieving the purpose of the process
• An arbitrary classification determined by how the
  enterprise defines its purpose.
• Process engineering starts when the enterprise
  starts defining how it will achieve its overall
  purpose
• Processes are polymorphisms of this purpose

See Supplementary Module 5 Box 56
STRUCTURE AND GOVERNANCE

• The resources and work products of a governing
  process are properties of governed processes
• A governing process is a polymorphism of a high
  order pattern
• Governing processes may be governed by governing
  processes, which may in turn be governed by other
  governing processes and so on
• The concept of governance order is a polymorphism
  of the order of a pattern
• Like constraints that constrain other constraints

40
INPUT OUTPUT PROCESSES
INPUT PROCESS
OUTPUT PROCESS
TRANSFORMATION PROCESS
CORE BUSINESS PROCESS OF THE COMPOSITION (normaliz
es its purpose. Is also repository of rules for
creating products with resources)
May be split only if each output process is an
independent irreducible fact
May be split only if each input process is an
independent irreducible fact
Input process normalizes rules about feeding
resources to the business process
Output process normalizes rules about the flow of
products from the business process

• Cannot be split if there are complex cardinality,
  latency and other constraints (eg a
  synchronization condition) between input
  processes for different resources
• Eg feeding of dough globs arranged on a cookie
  sheet implies the cookie sheet and the dough must
  be input simultaneously
• Then the input process for each resource cannot
  be split because it is a single irreducible fact

• Eg The act of unloading the cookies from the
  cookie sheet produces a used sheet and a batch of
  cookies
• A single event and irreducible fact
• We cannot split the output process of the cookie
  from the output process of the cookie sheet

41
CONSIDERATIONS

• Load Balancing
• Allocation of limited resources
• Affects latency
• Breaks out a new state Enabled (subtype of
  Suspended) with respect to a resource
• Latent processes may
• Be cancelled
• Wait for the resource
• May be deadlocked
• Substitute the needed resource with a mutable
  resource
• Acquire the needed resource (partially or fully)
• A polymorphism of conflict if this causes
  mutual exclusivity constraints to kick in
• the resource becomes unavailable to other
  processes
• Governing process may determine which option is
  used
• Parallel execution may decrease overall cycle
  time, but is prone to chaos as complexity
  increases
• Must be well governed

42
ADDITIONAL PROPERTIES OF PROCESSES

• Marginal Differences
• Consider differences between current and proposed
  states for
• Cycle time
• Not additive in a the aggregate
• Additive over a critical path
• Economic value added
• May not be additive in the aggregate
• Each instance may add value simultaneously to
  multiple compositions through reuse
• Activity Cost
• Direct Line Activity Cost
• Overheads and are normalized by the aggregate
• A composition of processes may have conditional
  activities
• Affects each item

Supplementary Module 5 Process Maps, Supply
chains and Business Process Engineering
43
PROCESS OWNERSHIP

• RResponsibility

• AAuthority

• WWork

• CConsult

• FFacilitate (a weak form of governance)

44
RAWCF RULES

• FFacilitate (a weak form of governance)

Subtype of

• CConsult

Subtype of

• RResponsibility

Subtype of

• Good Practice
• Assign to a single individual when possible
• Not always possible Eg Tug-of-war
• The irreducible fact cannot be a topos of
  conflict if this practice must be followed

• Must be assigned by execution time
• May be assigned at execution time

• AAuthority

Subtype of

• WWork

Must be assigned at task level

• Implications
• Each role is transitive with constituent
  subprocesses
• When the same person or organization plays
  multiple roles, it will suffice to specify the
  lower level role

45
AGILITY STRUCTURED vs. UNSTRUCTURED PROCESSES
(Most Structure)

• Unknown values must be instantiated at execution
  time
• Agile processes will instantiate at execution time

Unknown values of properties of a process
contribute to its lack of structure
Less certain (Least Structure)
(All properties of processes have not been shown
in the figure)
Supplementary Module 5 Box 59 and Processes that
gain or lose structure
Governance processes must instantiate and
regulate these Cannot be ad-hoc

• What effect will this have on business
  operations?
• Governance and management style?
• Automation?

46
AUDITABILITY

• Universal attributes of every temporal object
• Who made the change (All the dimensions of
  process ownership)
• When the change was made
• The instance of the process that caused the
  change and the (instances of resources) that were
  used
• Why it was made (the causal chain that led to the
  process)
• How long it took to make the change (cycle time)

47
Process Reengineering

• Change goals
• Substitute Mutable goals change scope degrees
  of freedom
• Change RAWCF
• Horizontal or vertical integration
• Change Process dependencies
• Substitute (mutable) resources
• Change any of the other process parameters we
  have discussed
• Automate to achieve the above goals Remember W
  is transitive A person is still responsible for
  execution

48
SUPPLY CHAINS
Reading Assignment Supplementary Module 5
Integrating Businesses through Process
Reengineering

• PROJECT Patterns of e-commerce

• PROJECT Map the Aris, SCOR, CPFR, Netmarket and
  Rosettanet supply chains to the patterns in your
  text book. Show how they may all be integrated

49
Product ReengineeringQuality is derived from the
Voice of the Customer
Shareholder Value
Financial
Stakeholders
FOCUS OF ALL VALUE
CSF How are we performing
financially?
Compliance Quality
Goals
Customer Value
Features
Metrics
Inovation/adaptation Quality
Process Quality

• All requirements and quality drivers must focus
  on creating value for the business
• Value for the customer
• Value for the shareholder

• Other stakeholders needs are enablers of these
  goals
• Regulators, Process owners, innovators and the
  community of knowledge

50
PRODUCT REENGINEERING

• See figure 2.18 in your text book, Box 62 in
  supplementary module 5

• A feature adds information
• May satisfy, hinder satisfaction of, or be
  irrelevant to satisfying need
• A feature is a constraint in information space
• Product innovation involves
• Adding features that enhance satisfaction
• Removing features that hinder satisfaction
• Eg An electronic check may be signed
  simultaneous by requisite signatories
• The paper check was constrained to be in one
  place at a time feature inherited from physical
  object
• Removing irrelevant features if it makes economic
  sense

51
EXAMPLE OF AUTOMATED PRODUCT INNOVATION
BEFORE
AFTER