Manufacturing Systems II

1
Manufacturing Systems II

• Chris Hicks
• Chris.Hicks_at_newcastle.ac.uk
• http//www.staff.ncl.ac.uk/chris.hicks

2
Topics

• Group Technology (Cellular Manufacture)
• Inventory Management
• Material Requirements Planning
• Just-in-Time Manufacture

3
Cellular Manufacturing
4
References

• Apple J.M. (1977) Plant Layout and Material
  Handling, Wiley, New York.
• Askin G.G Standridge C.R. (1993) Modelling and
  Analysis of Manufacturing Systems, John Wiley
• ISBN 0-471-57369-8
• Black J.T. (1991) The Design of a Factory with a
  Future, McGraw-Hill, New York, ISBN
  0-07-005550-5

5
References (cont.)

• Burbidge J.L. (1978)
• Principles of Production Control
• MacDonald and Evans, England
• ISBN 0-7121-1676
• Gallagher C.C. and Knight W.A. (1986)
• Group Technology Production Methods in
  Manufacture
• E. Horwood, England ISBN 0-471-08755-6
• Hyde W.F. (1981)
• Data Analysis for Database Design
• Marcel Dekker Inc
• ISBN 8247-1407-0

6
Manufacturing Layout

• Process (functional) layout, like resources
  placed together.
• Group (cellular) layout, resources to produce
  like products placed together.

7
(No Transcript)
8
Scientific Management

• F.W.Taylor 1907
• Division of labour - functional specialism
• Separation of doing and thinking
• Workers should have exact instructions
• Working methods should be standardised
• Specialisation led to functional layouts

9
Process Layout

• Like machines placed together
• Labour demarcation / common skills
• Robust wrt machine breakdown
• Common jigs / fixtures etc.
• Sometimes high utilisation
• Components travel large distances
• High work in progress
• Long lead times
• Poor throughput efficiency
• Often hard to control

10
Group Technology (Cellular Manufacturing)

• Group Technology is a manufacturing philosophy
  with far reaching implications.
• The basic concept is to identify and bring
  together similar parts and processes to take
  advantage of all the similarities which exist
  during all stages of design and manufacture.
• A cellular manufacturing system is a
  manufacturing system based upon groups of
  processes, people and machines to produce a
  specific family of products with similar
  manufacturing characteristics (Apple 1977).

11
Cellular Manufacturing

• Can be viewed as an attempt to obtain the
  advantages of flow line systems in previously
  process based, job shop environments.
• First developed in the Soviet Union in 1930s by
  Mitrofanov.
• Early examples referred to as Group Technology.
• Promoted by government in 1960s, but very little
  take up.
• In 1978, Burbidge asked What happened to Group
  Technology?
• Involves the standardisation of design and
  process plans.

12
Group (Cellular) Layout

• Product focused layout.
• Components travel small distances.
• Prospect of low work in progress.
• Prospect of shorter lead times.
• Reduced set-up times.
• Design - variety reduction, increased
  standardisation, easier drawing retrieval.
• Control simplified and easier to delegate.
• Local storage of tooling.

13
Group (Cellular) Layout

• Flexible labour required.
• Sometimes lower resource utilisation due to
  resource duplication.
• Organisation should be focused upon the group
  e.g. planning, control, labour reporting,
  accounting, performance incentives etc.
• Often implemented as a component of JIT with team
  working, SPC, Quality, TPM etc.
• Worker empowerment is important - need people to
  be dedicated to team success. Cell members should
  assist decision making.

14
Characteristics of Successful Groups
Characteristic Description Team Specified team
of workers Products Specified set of products
no others Facilities Dedicated machines /
equipment Group layout Dedicated
space Target Common group goal for
period Independence Groups can reach goals
independently Size Typically 6-15 workers
15
Adapted from Black (1991)
16
Implementation of Cellular Manufacturing

• Grouping - identifying which machines to put into
  each cell.
• Cell / layout design - identifying where to put
  to place machines.
• Justification
• Human issues

17
Types of Problem

• Brown field problem - existing layout, transport,
  building and infrastructure should be taken into
  account.
• Green field problem - designers are free to
  select processes, machines, transport, layout,
  building and infrastructure.
• Brown field problems are more constrained, whilst
  green field problems offer more design choice.

18
Grouping Methods

• Eyeballing
• Classification of parts
• Product Flow Analysis
• Cluster Analysis
• Matrix methods (e.g. King 1980)
• Similarity Coefficient methods
• Layout generation without grouping
• Beware
• Different methods can give different answers
• There may not be clear clusters
• Cellular manufacturing not always appropriate

19
Classification of Parts

• Based upon coding.
• Many schemes available.
• Basic idea is to classify according to geometry,
  similar shapes require similar processes.
• Grouping codes together is synonymous with
  grouping together like parts.
• Very prevalent in 1960s and 70s.
• Many schemes aimed at particular sectors.

20
Coding issues

• Part / component population
• inclusive should cover all parts.
• flexible should deal with future parts /
  modifications.
• should discriminate between parts with different
  values for key attributes.
• Code detail - too much and the code becomes
  cumbersome - too little and it becomes useless.
• Code structure - hierarchical (monocode), chain
  (polycode) or hybrid.
• Digital representation - numeric, alphabetical,
  combined.

21
(No Transcript)
22
Product Flow Analysis

• Developed by Jack Burbidge (1979).
• Uses process routings.
• Components with similar routings identified.
• Three stages
• Factory flow analysis.
• Group analysis
• Line analysis
• (See Askin and Standridge p177-179)

23
Factory Flow Analysis

• Link together processes (e.g. machining, welding,
  pressing) and subprocesses (turning, milling,
  boring) used by a significant number of parts.
• Large departments are formed by combining all
  related processes.
• These are essentially independent plants that
  manufacture dissimilar products.

24
Group Analysis

• Breaks down departments into smaller units that
  are easier to administer and control.
• The objective is to assign machines to groups so
  as to minimise the amount of material flow
  between the groups.
• Small inexpensive machines are ignored, since
  they can be replicated if necessary.

25
Group Analysis

• Construct a list of parts that require each
  machine. The machine with fewest part types is
  the key machine.
• A subgroup is formed from all the parts that need
  this machine plus all the other machines required
  to make the parts.
• A check is then made to see if the subgroup can
  be subdivided.
• If any machine is used by just one part it can be
  termed exceptional and may be removed.

26
Group Analysis

• Subgroups with the greatest number of common
  machine types may be combined to get groups of
  the desired size.
• The combination rule reduces the number of extra
  machines required and makes it easier to balance
  machine loads.
• Each group must be assigned sufficient machines
  and staff to produce its assigned parts.

27
Process Plan Example
28
Applying Grouping

• Steps
• 1. Identify a key machine. Either E or F.
• Create a subgroup to D,E and F.
• 2. Check for subgroup division. All parts visit F
  and so subgroup cannot be subdivided. Only part 7
  visits machine D so it is exceptional and is
  removed.
• 1. Identify an new key machine for remaining 6
  parts. A is the new key machine with subgroup
  A,B,C producing parts 1,2 3.
• 2. Subgroup division - C only used for part 3,
  therefore exceptional and can be removed.

29
Applying Grouping

• 1. Identify next key machine. Only parts 4,5, 6
  remain as well as machines C and D.
• 2. All parts use all machines - no subdivision
  possible.
• 3. Cell designer can now recombine the three
  subgroups into a set of workable groups of
  desired size.
• 4. The final solution must provide adequate
  machine resources in each group for the assigned
  parts. If exceptional parts exist, or if groups
  are not self contained, then plans must be made
  for transport.

30
Rank Order Clustering
1. Evaluate binary value of each row. 2. Swap
rows over to get them in rank order.
31
Rank Order Clustering
Next apply same method to the columns
32
Rank Order Clustering
Next swap over columns to get in rank order.
33
Rank Order Clustering
ROC has got a solution close to a block diagonal
structure. The process can be repeated
iteratively until a stable solution is found.
34
(No Transcript)
35
Similarity Coefficients

• Consider a pair of machines I,j,
• ni number of parts visiting machine i
• nj number of parts visiting machine j
• nij number of parts visiting i and j.
• Define similarity coefficient as
• sij max(nij/ni,nij/nj)
• Values near 1 denote high levels of interaction.
• Values near 0 denote little or no interaction.

36
Similarity Coefficients
37
(No Transcript)
38
Clustering

• We start with 6 clusters, one for each machine.
• With a threshold of T 1 machines A and B can be
  grouped. Likewise E and F.
• There are several methods for updating similarity
  coefficients between newly formed clusters and
  existing clusters.
• The single linkage approach uses the maximum Sij
  for any machine i in the first cluster and any
  machine j in the second cluster. Therefore any
  single pair of machines can cause groups to be
  combined

39
Updating Similarity Coefficients (Using Single
Linkage)
Next consider the highest value of T possible.
This gives the cluster CD at T 0.75. The
coefficients then need to be updated again.
40
Dendogram
41
(No Transcript)
42
Variety Reduction

• Basic principle always use common designs and
  components wherever possible.
• Modular design.
• Standardisation.
• Redundant features.
• Can base upon geometric series.
• Imperial / metric series.
• Reduced estimated work planning.
• Simplified stock control.
• Less problems with spares.

43
Variety Reduction

• May use slightly more expensive parts than
  necessary.
• Increases the volume of production of items.
• Reduced planning / jigs and fixtures etc.
• Reduced lead times.

44
Product Family Analysis

• There are a number of different ways of
  identifying part families. The following factors
  should always be considered
• How wide is the range of components?
• How static is workload?
• What changes are anticipated?
• Is Group Technology aimed purely at manufacturing
  or is standardisation and modularisation of
  design a major issue?

45
Manufacturing Layout

• Concerned with the relative location of major
  physical manufacturing resources.
• A resource may be a machine, department, assembly
  line etc.
• A block plan can be produced that shows the
  relative positioning of resources.
• Evaluation criteria are required such as
  minimising transport costs, distance travelled
  etc.

46
Approaches

• Many methods are based upon a static
  deterministic modelling approach.
• Dynamic effects may be guessed by trying out a
  variety of scenarios.
• Dynamic and stochastic effects may be evaluated
  by simulation.
• A premium may be placed upon favourable
  attributes
• Flexibility dealing with changes in design,
  demand etc.
• Modularity the ability to change the system by
  adding or removing component parts to meet major
  changes in demand.
• Reliability
• Maintainability

47
Line Layout
48
Spine Layout

• Spine is central core for traffic.
• Secondary aisles for traffic into departments
• Each department has input /output storage areas
  along the spine.
• Point of use storage reduces material flow.

49
Loop Structure
Circular Structure
50
Layout Configurations

• Eli Goldratt I, V, U, W
• I layouts have linear flow with no direction
  changes, empty pallets may go in reverse
  direction.
• V and U lines have more direction changes but may
  help with empty pallets.
• Rectilinear layouts may restrict operators from
  working multiple machines.
• Circular layouts may enable operators to work
  multiple machines.

51
Analysing Flow

• Sting diagrams provide a very quick way to
  identify the pattern of flow.
• Look at performance measures
• Distance travelled per component
• Material handling costs
• Material handling time
• Load / unload times
• Number of direction changes
• Number of moves per day
• Many, many more.
• Looking at performance measures enables
  alternative layouts to be evaluated.

52
Measures of Performance

• Resource Measures
• Resource utilisation
• Productivity.
• Inventory
• work in progress
• queues.
• Product
• lead times
• delivery performance
• Quality.
• Financial, overhead recovery v.s. ABC costing.

53
Creating Layouts

• If there is a dominant flow, such as all parts
  going from department 1-gt 2 -gt 3 then the layout
  should reflect this.
• At the other extreme, if the flow between
  departments is uniformly distributed, then any
  arrangement may be equally good.
• However, most problems will lie between the
  extremes of dominant and equal flow.

54
Systematic Layout Planning

• 1. Data collection.
• 2. Flow analysis.
• 3 Qualitative considerations.
• 4. Relationship diagram.
• 5. Space requirements.
• 6. Space availability.
• 7. Space relationship diagram.
• 8. Modifying considerations and limitations.
• 9. Evaluation.

(Muther 1973)
55
1. Data Collection

• Products to be produced volumes.
• Routing, Bill of Materials, parts lists.
• Resources for production, layout geometrical
  information.
• Timing information - set-up, processing
  transfer durations.
• Data determines loads resource utilisation.
• Quantity variety determine appropriate layout
  type.
• A Product-Quantity chart, which is a Pareto
  analysis of product importance can be used to
  determine items that justify their own lines or
  families of parts that justify a cell.

56
Product-Quantity Chart
57
2. Flow Analysis

• Operation process charts determine movement
  showing major operations, inspections, moves and
  storage.
• Process charts, similar to operation process
  charts, but more detail.
• Flow diagrams.
• Flow data can be summarised in From-To charts
  (like mileage charts in maps)
• Volumes
• Distance travelled
• Costs
• String diagrams.

58
Operation Process Chart
59
Process Chart Symbols
60
(No Transcript)
61
Flow Diagram
Material flow
Process chart symbols
Facility layout
62
3. Qualitative Considerations
REL Chart
63
4. Relationship Diagram

• Combines quantitative and qualitative
  relationship data.
• Provides a mechanism for visualising
  relationships.

64
5. Space Requirements

• Departmental space requirements need to be
  estimated.
• May have standards that define space requirement
  for each machine type.
• Can work from current space needs.
• Can determine space requirement by considering
  tasks performed, tooling, access, flow of
  materials etc.

65
6. Space Availability

• Need to accommodate machines, material handling
  equipment, people, energy transmission, drainage,
  air lines, communications etc.
• If an existing facility is to be used, the
  available space and constraints need to be
  accurately defined.
• In the case of new facilities there are financial
  and often planning constraints.Need to consider
  possibility of future changes in demand or use.

66
7. Space Relationship Diagram

• Represents departments with templates that are
  proportional in size to space requirements.
• Templates can be rearranged to find improved
  solutions.

67
(a)
(b)
a) Relationship Diagram b) Space Relationship
Diagram
68
8. Modifying Considerations and Limitations

• Steps 1-7 have not taken into account
  implementation details.
• Site specific or operations specific conditions
  may require adjustments to the layout.
• Need to consider
• Utilities, power, heating, light, drainage
  compressed air etc.
• Structural limitations, load-bearing capacity of
  floors, ceiling heights, columns.
• Location of external connections e.g. roads.

69
9. Evaluation

• Several alternatives should be considered.
• Drawings, flow diagrams etc form the basis of
  assessment of advantages and disadvantages of
  each.
• Costs / benefits can be attributed to each
  alternative.
• Quality of flow can be evaluated.
• Flexibility, maintainability, expandability
  safety and ease of operations should be reviewed.

70
Computerised Layout Planning

• Improvement algorithms are based upon an initial
  layout. They generate improvements by
  rearrangement. Suitable for brown field sites.
  Examples CRAFT (Armour Buffa 1963)
• Construction algorithms start with a blank shop
  floor and add machines to it. Suitable for green
  field sites. Example ALDEP (Seehof Evans
  1967), CORELP (Parsaei et al 1987), SHAPE (Hassan
  et al 1986).
• Hybrid algorithms include both construction and
  improvement algorithms.

71
Computerised Relative Allocation of Facilities
(CRAFT)

• Creates layouts by exchanging machine pairs and
  then evaluating the layout.
• When all pairs of exchanges have been completed,
  the exchange with the best evaluation is chosen
  and a new layout in generated.
• This process is repeated until no improvement can
  be made through exchanges.

72
Automated Layout Design Program (ALDEP)

• A machine is randomly selected and added to the
  layout.
• The closeness of all the remaining machines to it
  is calculated. The closest machine is added.
  This is repeated until all machines have been
  placed.
• Once a machine has been placed, it is fixed. This
  makes it difficult to find good solutions.
• Often use an improvement algorithm to improve
  layout produced.

73
Construction Algorithm Differences

• Method for election of next machine and its
  placement.
• Evaluation of the relationship between machines
  already located and the selected machine (e.g. by
  using different definitions of similarity
  coefficient).
• How the layout is represented.

74
Synthetic Machine Concept

• A group of machines form a synthetic machine.
• Resource hierarchy flattened.
• Framework to assist delegated responsibility.
• Local planning, control and work organisation.
• Concerned only with cell inputs and outputs.

75
Types of Cell

• Highly automated - conveyers, robot handling,
  Flexible Manufacturing Systems (FMS).
• Semi-automated - some automated material
  handling.
• Simple cells without automated material handling.
• Work grouped on a single machine using a
  multi-functional machine tool.
• NOTE Need to find an appropriate mix for given
  production volumes. Increasing automation
  normally increases overheads and reduces
  flexibility.

76
Supporting Techniques

• Statistical process control.
• Quality Circles.
• Team working.
• Empowerment.
• Visible performance measures.
• Total preventative maintenance.
• Single minute exchange of dies.
• Simple machine concept.

77
Case Study 1

• World class automotive components supplier.
• Adopted lean manufacturing practices yet
  productivity still 50 of Japanese sister plant.
• WHY?

78
Findings

• Layout - rectilinear v.s clusters.
• Supervision of resources.
• Smallest machine concept.
• Flexible resource variable.
• Cost of capital and accounting philosophies.

79
Case Study 2

• SME supplier of orthotics (surgical appliances).
• Very long delivery.
• High work in progress.
• How can situation be improved?

80
Solution

• Business process analysis
• Non physical processes
• Physical processes .
• Target queuing by streamlining processes or
  increasing capacity.
• Result
• Lead time 14 weeks to 4
• Cash flow improved by 300k on 2M turnover.

81
Other Key Issues

• Batch sizes
• Set-up
• Machining
• Transfer
• Effect on other measures of performance.

82
(No Transcript)
83
(No Transcript)
84
(No Transcript)
85
Hints

• Look at the material flow
• Try to simplify
• Think about removing in-process inventory
• Think about the operators
• Consider other layout constraints

86
(No Transcript)
87
(No Transcript)
88
(No Transcript)
89
Inventory Management
90
Inventory

• Money invested in materials
• 3 Types of inventory
• Raw materials
• Work in progress
• Finished goods

91
Advantages of Inventory

• Raw materials offset lead time
• Work in progress offsets disturbances in the
  production system and may help keep resource
  utilisation high
• Finished goods stocks enable fast delivery
• Economic order quantity methods claim to claim
• People feel busy
• Process decoupling

92
Inventory Disadvantages

• Expensive to keep
• Interest on capital
• Storage costs
• Adverse effect on cash flow and liquidity
• Risk of obsolescence
• Lack of flexibility
• Masks problems with manufacturing system
• Difficult to control

93
Types of Demand

• Independent
• Demand for an item is independent of the demand
  for another item
• Dependant
• Demand for an item is linked to the demand for
  another item
• Product structure defines dependencies

94
Inventory Control

• The activities and techniques of maintaining
  stock items at desired levels, whether they are
  raw materials, work in progress or finished
  products

95
Inventory Control Decisions

• How many?
• (lot size or order quantity
• When
• timing or order point

96
Independent Demand

• Fixed order quantity (FOQ) systems order a
  predetermined quantity of items when stock levels
  drop below a predetermined level e.g. 2 bin
  system
• Economic order quantity systems aim to minimise
  the combination of ordering and carrying costs.
  They make a number of assumptions
• annual demand can be estimated
• demand is uniform
• no quantity discounts
• Ignores the costs associated with stock outs

97
Dependent Demand

• Demand for one item linked to the demand for
  another
• Producing an assembly causes dependent demand for
  all the components that go into the assembly
• Assembling a car requires one windscreen, 5
  wheels, one engine etc.
• One engine requires one crankshaft, one cylinder
  head etc.
• One cylinder head requires .

98
Product Structure
99
Product Structure
A
B
D
c
E
F
G
H
100
Material Requirements Planning

• Method for planning dependent demand
• Requirement for subassemblies and components
  based upon requirements for end items and product
  structure
• Takes into account current stocks of each item to
  calculate net requirements

101
Material Requirements Planning
End Item Requirements
MRP
Stocks
Product Structure
Net Requirements
102
ABC Classification

• Break items into 3 groups
• A - the items that represent 75 of value and 20
  volume
• B - the items that represent 20 value and 30
  volume
• C - the items that represent 5 value and 50
  volume
• This approach is based upon
• Parieto analysis

103
Just-in-Time
104
Just-in-Time

• Approach to achieve excellence in manufacturing
• Minimise waste anything that adds cost but not
  value
• Just the correct quantity
• at just the right quality
• at just the right time
• in the right place

105
Push Scheduling
Manufacturing Systems
Inventory
PUSH
106
Kanban

• Japanese word for card
• One card
• Two card

107
One Card Kanban
Item Kanban
Stock Area
MC1
MC2
Machine 1 operates at a constant rate
Kanban
108
Two Card Kanban
Item Kanban
Item
Stock Area
MC1
MC2
P Kanban
C Kanban
P Production Kanban C Conveyance Kanban
109
Pull scheduling
5
2
1
3
4
Manufacturing System
PULL