Manufacturing Systems III

1
Manufacturing Systems III

• Chris Hicks MMM Engineering
• Email Chris.Hicks_at_ncl.ac.uk

2
Assessment

• End of year examination
• 2.5 hours duration
• Answer 4 questions from 6

3
Manufacturing Systems III

• Manufacturing Strategy
• JIT Manufacturing
• Manufacturing Planning and control
• Company classification
• Modelling Simulation
• Queuing theory (CFE)

4
Manufacturing Strategy
5

Reference

• Hill, T (1986),Manufacturing Strategy,
  MacMillan Education Ltd., London. ISBN
  0-333-39477-1

6
Manufacturing Strategy

• Long term planning
• Alignment of manufacturing to satisfy market
  requirements

7
Significance of Manufacturing

• Manufacturing often responsible for majority of
  capital and recurrent expenditure
• Long term nature of many manufacturing decisions
  makes them of strategic importance
• Manufacturing can have a large impact on
  competitiveness

8
Manufacturing Strategy

• Make / buy
• Process choice
• Technology
• Infrastructure, systems, structures
  organisation
• Focus
• Integration with other functions

9
Strategy Development

• Define corporate objectives
• Determine marketing strategies to meet these
  objectives
• Assess order qualifying and order winning
  criteria for products
• Establish appropriate processes
• Provide infrastructure

10
Identifying Market Requirements

• Order Qualifying criteria
• Order winning criteria
• Order losing criteria

11
Manufacturing Influences

• Costs
• Delivery
• Quality
• Demand flexibility
• Product range
• Standardisation / customisation

12
Profile Analysis

• Assess match between market requirements and
  current performance
• Identify changes required to manufacturing system

13
Market Requirements
Unimportant
V Imp.

• Price
• Quality
• Delivery
• CofOwn
• Customisation
• Other factors

14

Current Performance
Unimportant
V Imp.

Price Quality Delivery CofOwn Customisation O
ther factors
15

Market requirement
Achieved performance
V Imp.
Unimportant
Price Quality Delivery CofOwn Customisation O
ther factors

16
Process Choice

• Type of process project, jobbing, batch,line
• Flexibility
• Efficiency
• Robustness wrt product mix / volume
• Unique / generic technology?
• Capital employed
• How do processes help competitiveness?

17
Manufacturing Structure

• Layout functional or cellular?
• MTS / MTO
• Flexibility of workforce
• Organisation, team working etc.
• Breakdown of costs
• HRM issues

18
Products

• Relative importance, present and future
• Mix
• Complexity
• Product structure
• Concurrency
• Standardisation / customisation
• Contribution

19
Measures of performance

• What are they?
• Frequency of measurement
• Comparison with plan.
• Orientation product / process / inventory
• Integration with other functions

20
Infrastructure

• Manufacturing planning control
• Sharing information / knowledge
• CAD / CAM
• Accounting systems
• Quality systems
• Performance measurement

21
Case studies

• Heavy engineering
• PIP teams, simplification, value engineering,
  cellular manufacturing
• Automotive supplier
• world class but still relatively low
  productivity compared with Japanese sister
  company. Why?

22
Manufacturing is a business function
rather than a technical function. The emphasis
should be on supporting the market Terry Hill
(1996)

23
Just-in-Time Manufacturing
24
References

• APICS (1987),APICS Dictionary, American
  Production and Inventory Control Society, ISBN
  0-935406-90-S
• Vollmann T.E., Berry W.L. Whybark D.C.
  (1992),Manufacturing Planning and Control
  Systems (3rd Edition), Irwin, USA. ISBN
  0-256-08808-X
• Browne J., Harhen J, Shivnan J.
  (1988),Production Management Systems A CIM
  Perspective,Addison-Wesley, UK, ISBN
  0-201-17820-6

25
Just-in-Time Manufacturing

• In the broad sense, an approach to achieving
  excellence in a manufacturing company based upon
  the continuing elimination of waste (waste being
  considered as those things which do not add value
  to the product). In the narrow sense, JIT refers
  to the movement of material at the necessary
  time. The implication is that each operation is
  closely synchronised with subsequent ones to make
  that possible
• APICS Dictionary 1987

26
Just-in-Time

• Arose in Toyota, Japan in 1960s
• Replacing complexity with simplicity
• A philosophy, a way of thinking
• A process of continuous improvement
• Emphasis on minimising inventory
• Focuses on eliminating waste, that is anything
  that adds cost without adding value
• Often a pragmatic choice of techniques is used

27
Just-in-Time Goals

• Zero inventories
• Zero defects
• Traditional Western manufacturers considered Lot
  Tolerance Per Cent Defective (LTPD) or Acceptable
  Quality Levels (AQLs)
• Zero disturbances
• Zero set-up time
• Zero lead time

28
Just-in-Time Goals

• Zero transactions
• Logistical transactions ordering, execution and
  confirmation of material movement
• Balancing transactions associated with planning
  that generates logistical transactions -
  production control, purchasing, scheduling ..
• Quality transactions specification,
  certification etc.
• Change transactions engineering changes etc.
• Routine execution of schedule day in -day out

29
Benefits of JIT

• Reduced costs
• Waste elimination
• Inventory reduction
• Increased flexibility
• Raw materials / parts reduction
• Increased quality
• Increased productivity
• Reduced space requirements
• Lower overheads

30
Just-in-Time

• JIT links four fundamental areas
• Product design
• Process design
• Human / organisational issues
• Manufacturing planning and control

31

32
Product Design

• Design for manufacture
• Design for assembly
• Design for automation
• Design to have flat product structure
• Design to suit cellular manufacturing
• Achievable and appropriate quality
• Standard parts
• Modular design

33
Process Design

• Set-up / lot size reduction
• Include surge capacity to deal with variations
  in product mix and demand
• Cellular manufacturing
• Concentrate on low throughput times
• Quality is part of the process, autonomation,
  machines with built in capacity to check parts
• Continuous quality improvement
• No stock rooms - delivery to line/cell
• Flexible equipment
• Standard operations

34
Human / Organisational Elements

• Whole person concept, hiring people, not just
  their current skills / abilities
• Continual training / study
• Continual learning and improvement
• Workers capabilities and knowledge are as
  important as equipment and facilities
• Workers cross trained to take on many tasks
  process operation, maintenance, scheduling,
  problem solving etc.
• Job rotation / flexibility
• Life time employment / commitment?

35
Organisational Elements

• Little distinction between direct / indirect
  labour
• Activity Based Cost (ABC) accounting
• Visible team performance measurement
• Communication / information sharing
• Joint commitment

36
JIT Techniques

• Manufacturing techniques
• Production and material control
• Inter-company JIT
• Organisation for change

37
Manufacturing Techniques

• Cellular manufacturing
• Set-up time reduction
• Pull scheduling
• Smallest machine concept
• Fool proofing (Pokayoke)
• Line stopping (Jikoda)
• I,U,W shaped material flow
• Housekeeping

38
Group Technology / Cellular Manufacturing

• Improved material flow
• Reduced queuing time
• Reduced inventory
• Improved use of space
• Improved team work
• Reduced waste
• Increased flexibility

39
Set-up Time Reduction

• Single minute exchange of dies (SMED) - all
  changeovers lt 10 mins.
• 1. Separate internal set-up from external set-up.
  Internal set-up must have machine turned off.
• 2. Convert as many tasks as possible from being
  internal to external
• 3. Eliminate adjustment processes within set-up
• 4. Abolish set-up where feasible
• Shingo, S. (1985),A Revolution in Manufacturing
  the SMED System, The Productivity Press, USA.

40
Basic Steps in a Traditional Set-up Operation

• 1. Preparation, after process adjustments,
  checking of materials and tools (30).
• 2. Mounting and removing blades, tools and parts
  (5) Generally internal.
• 3. Measurements, settings and calibration (15)
  includes activities such as centring,
  dimensioning, measuring temperature or pressure
  etc.
• 4. Trial runs and adjustments (50) - SMED
• Typical proportion of set-up time given in
  parenthesis.

41
Set-up Analysis

• Video whole set-up operation. Use cameras time
  and date functions
• Ask operators to describe tasks. As group to
  share opinions about the operation.

42
Three Stages of SMED

• 1. Separating internal and external set-up
• doing obvious things like preparation and
  transport while the machine is running can save
  30-50.
• 2.Converting internal set-up to external set-up
• 3. Streamlining all aspects of the set-up
  operation

43
Separating Internal and External Set-up
44
(No Transcript)
45
ANDON

• A board which shows if any operator on the line
  has difficulties
• Red - machine trouble
• White - end of a production run
• Blue - defective unit
• Yellow - set-up required
• Line-stop - all operators can stop the line to
  ensure compliance with standards
• Flexible workers help each other when problems
  arise

46
JIT Material Control

• Pull scheduling
• Line balancing
• Schedule balance and smoothing (Heijunka)
• Under capacity scheduling
• Visible control
• Material Requirements Planning
• Small lot batch sizes

47
Pull Systems

• Work centres only authorised to produce when it
  has been signalled that there is a need from a
  user / downstream department
• No resources kept busy just to increase
  utlilisation
• Requires
• Small lot-sizes
• Low inventory
• Fast throughput
• Guaranteed quality

48
Pull Systems

• Implementations vary
• Visual / audio signal
• Chalk square
• One / two card Kanban

49
Material Requirements Planning / JIT

• Stable Master Production Schedule
• Flat bills of materials
• Backflushing
• Weekly MRP quantities with call off , a common
  approach

50
JIT Purchasing

• JIT purchasing requires predictable (usually
  synchronised) demand
• Single sourcing
• Supplier quality certification
• Point of use delivery
• Family of parts sourcing
• Frequent deliveries of small quantities
• Propagate JIT down supply chain, suppliers need
  flexibility
• Suppliers part of the process vs. adversarial
  relationships

51
JIT Purchasing

• Controls and reduces inventory
• Reduces space
• Reduces material handling
• Reduces waste
• Reduces obsolescence

52
Organisation for Change

• Multi-skilled team working
• Quality Circles, Total Quality Management
• Philosophy of joint commitment
• Visible performance measurement
• Statistical process control (SPC)
• Team targets / performance measurement
• Enforced problem solving
• Continuous improvement

53
Total Quality Management (TQM)

• Focus on the customer and their requirements
• Right first time
• Competitive benchmarking
• Minimisation of cost of quality
• Prevention costs
• Appraisal costs
• Internal / external failure costs
• Cost of exceeding customer requirements
• Founded on the principle that people want to own
  problems

54
JIT Flexibility

• Set-up time reduction
• Small transfer batch sizes
• Small lot sizes
• Under capacity scheduling
• Often labour is the variable resource
• Smallest machine concept

55
Reducing Uncertainty

• Total Preventative Maintenance (TPM) / Total
  Productive Maintenance
• 100 quality
• Quality is part of the process - it cant be
  inspected in
• Stable and uniform schedules
• Supplier quality certification

56
Total Preventative Maintenance (TPM)

• Strategy to prevent equipment and facility
  downtime
• Planned schedule of maintenance checks
• Routine maintenance performed by the operator
• Maintenance departments train workers, perform
  maintenance audits and undertake more complicated
  work

57
Implementation of JIT

58
Implementation of JIT

• Method
• 1. Lower inventory levels
• 2. Identify problems
• 3. Eliminate problems
• 4. Improve use of resources
• Inventory
• People
• Capital
• Space
• 5. Go back to step 1

59
JIT Circle

Standardisation Design - focus
TPM
JIT Purchasing
TQM
Visibility
JIT
Set-up reduction
Pull scheduling
Multi-skill Workforce
Plant Layout
Small machines
60
JIT Limitations

• Stable regular demand
• Medium to high volume
• Requires cultural change
• Implementation costs

61
Computer Aided Production Management Systems
(CAPM)
62
References

• Vollmann T.E., Berry W.L. Whybark D.C.
  (1992),Manufacturing Planning and Control
  Systems (3rd Edition), Irwin, USA. ISBN
  0-256-08808-X
• (Earlier editions just as good!)
• Browne J., Harhen J, Shivnan J.
  (1988),Production Management Systems A CIM
  Perspective,Addison-Wesley, UK, ISBN
  0-201-17820-6

63
Computer Aided Production Management (CAPM)
Systems

• All computer aids supplied to the manager
• Specification - ensuring that the manufacturing
  task has been defined and instructions provided
• Planning and control - scheduling, adjusting
  resource usage and priorities, controlling the
  production activity
• Recording and reporting the status of production
  and performance

64
Computer Aided Production Management (CAPM)
Systems

• Information systems responsible for
• Transaction processing - maintaining, updating
  and making available specifications, instructions
  and production records
• Management information - for exercising
  judgements about the use of resources and
  customer priorities
• Automated decision making - producing production
  decisions using algorithms

65

66
CAPM Systems

• Planning
• Control
• Performance measurement

67
Planning Modules

• Master Production Scheduling (MPS) - high level
  production plan in terms of quantity, timing and
  priority of planned production
• Materials Requirements Planning (mrp) /
  Manufacturing Resources Planning (MRP)
• Capacity Planning

68
Control Modules

• Inventory control - keeping raw material, work in
  process (WIP) and finished goods stocks at
  desired levels
• Shop floor control (Production Activity Control)
  - transforming planning decisions into control
  commands for the production process
• Vendor measurement - measuring vendors
  performance to contract, covering delivery,
  quality and price

69
Material Requirements Planning (mrp)

• Material requirements plannning originated in
  the 1960s as a computerised approach for planning
  of materials acquisition for production. These
  early applications were based upon a bill of
  materials processor which converted demand for
  parent items into demand for component parts.
  This demand was compared with available inventory
  and scheduled receipts to plan order releases
  Browne et al (1986)

70
Manufacturing Resources Planning (MRP)

• The combination of planning and control modules
  was termed closed loop MRP. With the addition
  of financial modules an integrated approach to
  the management of resources was created. This was
  termed Manufacturing Resources Planning.
• Material Requirements Planning (mrp / MRPI)
• Manufacturing Resources Planning (MRP/MRPII)

71
Material Requirements Planning

• Dependant demand
• Time phased planning
• Inputs
• Master Production Schedule
• Bill of Materials
• Inventory status
• Assumptions
• Infinite capacity
• Fixed lead times
• Fixed and predetermined product structure

72

73
MRP Record Card

74
MRP Conventions

• MRP time buckets
• Scheduled receipts at start of period
• Projected available balance at end of period
• Planned order releases at the start of period
• Planned orders vs. scheduled receipts
• Number of buckets planning horizon

75
Representation of Product

76
Linked MRP Cards

77
Backwards Scheduling
78
Forwards Scheduling
79
MRP Domain

• Steady state systems
• Low levels of uncertainty
• Shallow / medium or deep product structure
• Stable demand
• Predominantly make to stock
• Manufacturing orientation

80
MRP Parameters

• Planning horizon
• Size of time bucket
• Lot sizing rules
• Regeneration vs.. net change

81
Validity of MRP Assumptions

• Infinite capacity vs. capacity planning
• Fixed lead times / varying load
• Lead times are a result of the schedule
• Integration of planning levels requires
  feasibility at high and low levels

82
Typical Control Parameters

• Safety stock
• Safety lead time
• Yield
• Order quantity category
• Min/max order levels
• Max. days supply
• Min. days between orders

83
Lot sizing

• Lot-for-lot
• Economic Order Quantity (EOQ)
• Complex optimisation algorithms

84
Uncertainties in MRP

• Environmental uncertainty
• Customer orders
• Suppliers
• System uncertainty
• Product quality
• Scrap / rework
• Process times
• Design changes
• MRP nervousness / instability

85
Dealing with uncertainty in MRP

• Safety stocks
• Safety lead times
• Safety due date
• Hedging
• Over-planning
• Yield factors

86
Appropriate approaches

• Timing uncertainty safety lead time
• Quantity uncertainty safety stock

87
MRP Nervousness

• Significant changes in plans due to minor changes
  in high level plans
• Frequent changes in plans make the MRP system
  lose crdibility

88
Causes of Nervousness

• Demand uncertainty
• Product structure characteristics
• Incorrect lot-sizing rules

89
Nervousness Solutions

• Stable MPS
• Carefully change any parameter changes
• Use different lot sizing rules at the high and
  low levels of the product structure

90
MRP Problems

• Quality of the model
• Bill of materials structure
• Non-material activities
• Validity of the assumptions
• Lack of 2 way time analysis
• Quality of data
• Regeneration / computational effort
• Poor visibility
• Operational aspects

91
How to implement MRP

• Get accurate data
• Make sure you have accurate data
• Have good procedures to make sure that the data
  is always accurate
• Remember approximately 75 of MRP implementations
  fail
• Unsuccessful MRP costs nearly the same as
  successful MRP

92
Capacity Planning
93
References

• Vollmann T.E., Berry W.L. Whybark D.C.
  (1992),Manufacturing Planning and Control
  Systems (3rd Edition), Irwin, USA. ISBN
  0-256-08808-X
• Plossl G.W. Wight O.W. (1973), Capacity
  Planning and Control, Production and Inventory
  Management, 3rd quarter 1973 pp31-67

94
Capacity Planning

• The function of establishing, measuring and
  adjusting limits or levels of capacity.
• Capacity planning in this context is the process
  of determining how much labour and machine
  resources are required to accomplish the tasks of
  production.
• Open shop orders and planned orders in the MRP
  system are input to CRP which translates these
  into hours of work, by work centre, by time
  period
• APICS Dictionary 1987

95
Capacity Planning

• Plossl bath tub
• Lead-time queuing time set-up time
  processing time transfer time
• Queuing time is dependant upon the level of
  backlog in the system
• Three reasons why queues go out of control
• Inadequate capacity
• Erratic input
• Inflated lead time estimates

96
Plossl Bath Tub

97
Lead-time Syndrome

• Vicious circle which can occur when queuing
  conditions change
• Increased demand may increase backlog
• Increased backlog increases demand
• If the planned lead times are changed, more
  orders are likely to arrive to meet requirements
  during the increased lead time.
• This further inflates lead times etc. etc.

98
Capacity Control

• Input-output control ensure that the demand
  never exceeds capacity
• In MTO, backlogs act as buffers against workload
  variations. In this case its a trade off between
  maintaining resource utilisation and minimising
  lead-times and inventory

99
Capacity Planning Approaches

• Infinite loading assume infinite capacity,
  disregarding capacity constraints
• Finite loading work to capacity constraints

100
Infinite Loading
Backlog
101
Finite Loading
102
Infinite Loading

• Easier - less computation required
• Identifies and measures scheduled over and under
  loads
• Shows how much capacity is required to meet the
  plan (finite loading does not)

103
Finite Loading

• Capacity of each resource specified in terms of
  standard and maximum capacity
• Jobs loaded onto each work centre in priority
  order
• When resources are full, jobs are rescheduled
• Horizontal vs. vertical loading
• The only way to revise a finite loading schedule
  is to start from scratch, rearranging jobs in a
  new priority sequence

104
Capacity Planning

• A prerequisite to having an effective capacity
  planning system is to have an effective priority
  planning system.
• If the due dates, or lead times are incorrect,
  the schedule, the priorities and the projection
  of when the load will hit the resources will be
  fiction. The system will not work
• Plossl Wight 1973

105
5 Levels of Capacity Planning

• Resource planning highly aggregated, longest
  term level of capacity planning
• Rough-cut capacity planning uses MPS data
• Capacity Requirements Planning (CRP)
• Finite loading
• Input / output control

106
(No Transcript)
107
Rough-cut Capacity Planning

• Capacity Planning Using Overall Factors (CPOF)
  calculates the overall direct labour requirements
  for the MPS and identifies load based upon
  historic data
• Capacity Bills, uses BOM and planning data
• Resource profiles, same as capacity bills, but
  time phased
• See Vollmann et al for details

108
Capacity Requirements Planning

• CRP utilises MRP information such as lot sizing
  and inventory data
• Shop floor control provides information of the
  current status of items only the capacity
  required to complete items is considered
• CRP is based upon the infinite loading approach

109
Company Classification
110
References

• Woodward J. (1965), Industrial Organisation
  Theory and Practice, Oxford University Press,
  England
• New C.C. (1976), Managing Manufacturing
  Operations, British Institute of Management,
  Report No. 35.
• Barber K.D. Hollier R.H. (1986),The Effects of
  Computer Aided Production Management Systems on
  Defined Company Types, Int. J. Prod. Res. 24(2)
  pp311-327

111
References

• Barber K.D. Hollier R.H. (1986),The Use of
  Numerical Taxonomy to Classify Companies
  According to Production Control Complexity, Int.
  J. Prod. Res. 24(1) pp203-22

112

Company Classification

• Classification groups like items together
• Dependent upon classification variables
• Enables similarities and differences between
  companies to be identified
• Identify appropriate planning control method
• Identify appropriate technology

113

Classification Approaches

• General company classification
• Joan Woodward (1965) used Ministry of Labour
  categories for investigating organisational
  structure issues
• Sector based classification commonly used by
  financial institutions (e.g. FT classification)
• DTI - SMEs
• Classification of manufacturing
• Mode of production e.g. Burbidge (1971), volume
  of production jobbing, batch, flow
• Goldratt (1980) VAT analysis based upon pattern
  of material flow
• Production control complexity New (1976), Barber
  Hollier (1986)

114
Colin New Classification

• Survey of 186 companies to investigate
  manufacturing management practice
• Five classification areas
• Market - customer environment
• Relationship between cumulative lead time and
  delivery lead time e.g. make to stock or
• make to order
• Product range and rate of product innovation
• Product complexity - number of components per
  product, depth of product structure
• Organisation of manufacturing system, functional
  vs. group layout
• Cost structure of products

115
Market / Customer Environment

• Make to stock v/s make to order
• Marucheck McClelland (1986)
• Continuum from pure ETO - pure MTS
• Positioning of company usually a strategic issue
• Effects competitive factors - customisation vs.
  lead time and cost
• Position effects inventory
• Hicks (1994) Business process based description

116
Product Complexity

• Depth of product structure
• effects co-ordination of assembly processes
  (phasing), uncertainties, lead times etc.
• Number of components in product
• Source of components (make / buy)
• Standardisation / modular design vs. pure ETO
• Concurrent engineering also increases control
  complexity

117
Organisational Structure

• Type of layout (process / cellular)
• Management style
• Company culture
• Flexibility

118
Barber Hollier (1986)

• Worked aimed establish suitability of computer
  aided production management techniques for
  different types of company
• Based upon production control complexity
• Developed work of Colin New (1976)
• Used numerical taxonomy to identify clusters of
  common companies
• Work identified 6 groups of company

119
Chris Voss (1987)
120
(No Transcript)
121
(No Transcript)
122
(No Transcript)
123
Modelling Simulation
124
References

• Kreutzer W. (1986), System Simulation
  Programming Languages and Styles, Addison-Wesley
• ISBN 0-201-12914-0
• Mitrani I (1982),Simulation Techniques for
  Discrete Event Systems, Cambridge University
  Press
• ISBN 0-521-23885-4

125
Modelling

• Systems identification
• System representation
• Model design
• Model coding
• Validation
• (last two points relate to simulation modelling)

126
Types of Model

• Iconic models e.g. a globe is an iconic model of
  the earth
• Analytical models general solutions to families
  of problems based upon some strong theory (close
  form solutions)
• Analytical models represent systems through some
  abstract notion of similarity
• Symbolic models use of symbols to describe
  objects, relationships, actions and processes
• Churchman 1959

127

• Induction deducing a general principle from
  particular instances
• Deduction deducing a particular instance from a
  general law

128
Descriptive Model

• Descriptive models offer some symbolic
  representation of some problem space without any
  guidance on how to search it. The use of
  descriptive models is an inductive, experimental
  technique for exploring possible worlds
• Kreutzer 1986

129
Simulation

• The term simulation is used to describe the
  exploration of a descriptive model under a chosen
  experimental frame
• Kreutzer 1986
• Simulation is partly art, partly science. The
  art is that of programming a simulation should
  do what is intended. One should also know how to
  answer questions about the system being
  simulated
• Mitrani 1982

130
Limitations of Simulation

• Expensive in terms of manpower and computing
• Often difficult to validate
• Often yields sub-optimum results
• Iterative problem solving technique
• Collection, analysis and interpretation of
  results requires a good knowledge of probability
  and statistics
• Difficult to convince others
• Often a method of last resort

131
When to use Simulation

• The real system does not exist, or it is
  expensive, time consuming, hazardous or
  impossible to experiment with prototypes
• Need to investigate past, present and future
  performance in compressed, or expanded time.
• When mathematical modelling is impossible or they
  have no solutions
• Satisfactory validation is possible
• Expected accuracy meets requirements

132
Simulation Methodology

• System identification
• System Representation
• Model design
• Data collection and parameter estimation
• Program design
• Program implementation
• Program verification
• Model validation
• Experimentation
• Output analysis

133

System Identification

A system is defined as a collection of objects,
their relationships and behaviour relevant to a
set of purposes, characterising some relevant
part of reality Kreutzer (1986)
134

System Representation

Symbolic images of objects, relationships and
behaviour patterns are bound into structures as
part of some larger framework of beliefs,
background assumptions and theories of the
problem solver Kreutzer 1986
135

Model Design

A model is an appropriate representation of some
mini-world. Models can very quickly grow to form
very complicated structures. Control and the
constraint of complexity lie at the heart of any
modelling activity. Care must be exercised to
preserve only those chracteristics that are
essential. This depends upon the purpose of the
model Kreutzer 1986
136

It is necessary to abstract from the real system
all those components (and their interactions that
are considered to be important Mitrani 1982

137

Model Coding

This stage exists when computers are being used
as the modelling medium. This stage seeks a
formal representation of symbolic structures and
their transformations into data structures and
computational procedures in some programming
language Kreutzer 1986
138
Types of Simulation Model

• Monte Carlo
• Quasi-continuous
• Discrete event
• Combined simulation

139
Monte Carlo Simulation

• Derives name from roulette
• Static simulation
• Distribution sampling
• No assumptions about model
• Only statistical correlation between input and
  output explored
• Results often summarised in frequency tables
• Used for complex phenomena that are not well
  understood, or too complicated and expensive to
  produce other models

140
Quasi- Continuous Simulation

• Dynamic simulation. The clock is sequenced by a
  clock in uniform fixed length intervals. The size
  of the increment determines the resolution of the
  model
• Kreutzer 1986

141
Discrete Event Simulation

• Asynchronous clock
• Assumes nothing interesting happens between
  events
• Queuing networks in which the effects of capacity
  limitations and routing strategies often studied
  using DES
• This type of simulation most frequently used for
  simulating manufacturing systems

142
Types of Discrete Event Simulation

• Event scheduling
• Process interaction
• Object orientated
• Activity scanning

143
Event Scheduling Approach

• Event scheduling binds actions associated with
  individual events into event routines.
• The monitor selects event for execution,
  processing a time ordered agenda event notices.
• Event notices contain a time and a reference to
  an event routine.
• Each event can schedule another event, which is
  placed in the correct position of the agenda.
• The clock is always set to the time of the next
  immanent event

144
Process Interaction Approach

• Focuses on the flow of entities through the model
• Views system as concurrent, interacting processes
• Life cycle for each class of entities
• Monitor uses agenda to keep track of pending
  tasks
• Monitor records activation times, process
  identities and state that the process was last
  suspended

145
Object Orientated Programming

• Process records the values of all local variables
• Object contains, attributes (data), activities
  (processes) and lifecycle
• Communication between objects only through well
  defined interfaces provided by messages which an
  object is programmed to respond to
• Classes / sub classes
• Instances
• Inheritance

146
Activity Scanning Approach

• Each event is specified in terms of the
  conditions that need to apply for the event to
  start and finish
• Each event has a set of actions that take place
  when it finishes
• Model execution is cyclic, scanning all
  activities in the model testing which can start /
  finish.
• Clock only moves when whole cycle leaves status
  unchanged
• 3 phase structure computationally expensive
• Conditional Sequencing since programmer only
  states start and end conditions

147
Types of Simulation

• Deterministic - no random component
• Stochastic - represents uncertainties

148
Stochastic Simulation

• Sampling experiments
• Standard statistical approaches such as design of
  experiments used
• Random processes based upon pseudo random number
  generators

149
Pseudo-Random Number Generators

• Seed based algorithm produces random number
  from seed. Repeated execution gives same streams
  of random numbers
• Non-seed based, random number generated using
  time, or status of computer

150
(No Transcript)
151
Validation
Model qualification
CONCEPTUAL
Analysis
MODEL
REALITY
Programming
Computer
Simulation
Model
Model
verification
validation
Computer
Model