Ö.

Anlağan
1996

ME 445
INTEGRATED MANUFACTURING
SYSTEMS

GROUP TECHNOLOGY
 Ö. Anlağan
1996

BATCH MANUFACTURING
IS A DOMINANT MANUFACTURING ACTIVITY IN THE WORLD,
GENERATING A GREAT DEAL OF INDUSTRIAL OUTPUT

IT ACCOUNTS
60% - 80%
OF ALL MANUFACTURING ACTIVITIES

CHARACTERISTICS OF
BATCH MANUFACTURING:

 High level of product variety

 Small manufacturing lot size

Time on Moving and waiting

machine 95%
5%

Cutting Positioning, loading, gauging,

less than idle, etc.
30% 70%

WHAT IS GROUP TECHNOLOGY?

Group technology (GT) is a philosophy that implies the notion of

recognizing and exploiting similarities in three different ways:
1. By performing like activities together
2. By standardizing similar tasks
3. By efficiently storing and retrieving information about recurring
problems
Large manufacturing system can be decomposed into smaller subsystems
of part families based on similarities in design attributes and
manufacturing features.
 Ö. Anlağan
1996
DESIGN ATTRIBUTES:

part configuration (round or prismatic)

dimensional envelope (length to diameter ratio)
surface integrity (surface roughness, dimensional tolerances)
material type
raw material state (casting, forging, bar stock, etc.)

PART MANUFACTURING FEATURES:

operations and operation sequences (turning, milling, etc.)
batch sizes
machine tools
cutting tools
work holding devices
processing times

An essential aspect of the integration of CAD and CAM is the integration

of information used by engineering and manufacturing and all the other
departments in a firm.

Group technology emphasis on part families based on similarities in

design attributes and manufacturing, therefore GT contributes to the
integration of CAD and CAM.

Group technology begun by grouping parts into families based on their

attributes. There are three methods that can be used to form part families:
1. Manuel visual inspection
2. Production flow analysis
3. Classification and coding
Manual visual inspection involves arranging a set of parts into groups
known as part families by visually inspecting the physical characteristics
of the parts.

Manual visual inspection

incorrect results
human error
different judgment by different people
inexpensive
least sophisticated
good for small companies having smaller number of parts
Ö. Anlağan
 1996
Production flow analysis: Parts that go through common operations are
grouped into part families.

The machines used to perform these common operations may be grouped

as a cell, consequently this technique can be used in facility layout
(factory layout)

Coding methods: are employed in classifying parts into part families

Coding refers to the process of assigning symbols to the parts

The symbols represent design attributes of parts or manufacturing

features of part families

The variations in codes resulting from the way the symbols are assigned
can be grouped into three distinct type of codes:

1. Monocode or hierarchical code

2. Polycode or attribute
3. Hybrid or mixed code

MONOCODE (HIERARCHICAL CODE):

This coding system was originally developed for biological classification

in 18th century.

The structure of monocode is like a tree in which each symbol amplifies

the information provided in the previous digit.

A monocode (hierarchical code) provides a large amount of information

in a relatively small number of digits

useful for storage and retrieval of design-related information

such as part geometry, material, size, etc.
it is difficult to capture information on manufacturing sequences
in hierarchical manner, so applicability of this code in
manufacturing is rather limited
 Ö. Anlağan
1996

POLYCODE (ATTRIBUTE CODE):

The code symbols are independent of each other

Each digit in specific location of the code describes a unique property of

the workpiece

it is easy to learn and useful in manufacturing situations where

the manufacturing process have to be described
the length of a polycode may become excessive because of its
unlimited combinational features

Differences in information storage capacity between monocode and

polycode:

Assume that a code consists of five symbols and that in each of the five
code fields the digits 0 to 9 are used. Determine how many mutually
exclusive characteristics can potentially be stored in the monocode and
the polycode.

Number of characteristics may be stored in a monocode:

101 + 102 + 103 + 104 + 105 =111110

Number of characteristics may be stored in a polycode:

10 + 10 + 10 + 10 + 10 = 50

MIXED CODE (HYBRID CODE):

It is the mixture of both monocode and polycode systems. Mixed code

retains the advantages of both systems. Most coding systems use this
code structure.

The first digit for example, might be used to denote the type of part, such
as gear. The next five position might be reserved for a short attribute code
that would describe the attribute of the gear. The next digit (7th digit)
might be used to designate another subgroup, such as material, followed
by another attribute code that would describe the attributes.

Ö. Anlağan
 1996
A code created by this manner would be relatively more compact than a
pure attribute code while retaining the ability to easily identify parts with
specific characteristics.

The OPITZ classification system:

 it is a mixed (hybrid) coding system

 developed by Opitz, Technical University of Aachen, 1970
 it is widely used in industry
 it provides a basic framework for understanding the
classification and coding process
 it can be applied to machined parts, non-machined parts (both
formed and cast) and purchased parts
 it considers both design and manufacturing information

The Opitz coding system consists of three groups of digits:

Form Supplementary Secondary

code code code
12345 6789 ABCD
part geometry information Production
and features relevant to processes and
relevant to part manufacturing production
design (polycode) sequences

PART FAMILY FORMATION:

One of the primary uses of coding systems is to develop part families.

Example: Consider the family of ferrous parts formed by first three digits
of Opitz form code; 132.

This implies that the attributes associated with the family members are
length/diameter ratio in the range 0.5 to 3.0, all parts stepped to one end
and internal shape elements with threads.

A number of mathematical approaches have also been developed to form

part families using classification and coding system.
 Ö. Anlağan
1996
SELECTION OF CLASSIFICATION AND CODING SYSTEMS

For the purpose of selecting or developing your own code, it is important

to understand the attributes of classification and coding systems. Some of
the important classification and coding system attributes include:

1. Flexibility for various applications such as part family formation,

process planning, costing, and purchasing
2. Accuracy, to provide correct information on parts
3. Expandability, to accommodate information on more part attributes
deemed important later on
4. Ease of learning
5. Ease of retrieval
6. Reliability and availability of software
7. Suitability for specific applications

Matching these attributes with the objectives of an organization would be

helpful in selecting or developing a coding system to meet
organizational needs.

BENEFITS OF GROUP TECHNOLOGY

Group technology is a management strategy to help eliminate waste

caused by duplication of effort.

It affects all areas of a company, including:

 engineering
 equipment specification
 facilities planning
 process planning
 production control
 quality control
 tool design
 purchasing
 service
 Ö. Anlağan
1996

Some of the well-known tangible and intangible benefits of implementing

GT :

1. Engineering design
 Reduction in new parts design
 Reduction in the number of drawings through standardization
 Reduction of drafting effort in new shop drawings
 Reduction of number of similar parts, easy retrieval of similar
functional parts, and identification of substitute parts

2. Layout planning
 Reduction in production floor space required
 Reduced material-handling effort

3. Specification of equipment, tools, jigs, and fixtures

 Standardization of equipment
 Implementation of cellular manufacturing systems
 Significant reduction in up-front costs incurred in the release of
new parts for manufacture

4. Manufacturing: process planning

 Reduction in setup time and production time
 Alternative routing leading to improved part routing
 Reduction in number of machining operations and numerical
control (NC) programming time

5. Manufacturing: production control

 Reduced work-in-process inventory
 Easy identification of bottlenecks
 Improved material flow and reduced warehousing costs
 Faster response to schedule changes
 Improved usage of jigs, fixtures, pallets, tools, material handling,
and manufacturing equipment
 Ö. Anlağan
1996

6. Manufacturing: quality control

 Reduction in number of defects leading to reduced inspection
effort
 Reduced scrap generation
 Better output quality
 Increased accountability of operators and supervisors responsible
for quality production, making it easier to implement total quality
control concepts.

7. Purchasing
 Coding of purchased part leading to standardized rules for
purchasing
 Economies in purchasing possible because of accurate knowledge
of raw material requirements
 Reduced number of part and raw materials
 Simplified vendor evaluation procedures leading to just-in-time
purchasing

8. Customer service
 Accurate and faster cost estimates
 Efficient spare parts management, leading to better customer
service

CELLULAR MANUFACTURING

Cellular manufacturing is an application of group technology in

manufacturing in which all or a portion of a firm’s manufacturing
system has been converted into cells.

A manufacturing cell is a cluster of machines or processes located in

close proximity and dedicated to the manufacture of a family of parts.
The parts are similar in their processing requirements, such as
operations, tolerances, and machine tool capacities
 Ö. Anlağan
1996

The primary objectives in implementing a cellular manufacturing

system are to reduce:
 setup times (by using part family tooling and sequencing)
 flow times (by reducing setup and move times and wait time for
moves and using smaller batch sizes)
 reduce inventories
 market response times

In addition, cells represent sociological units that have more tendency to

teamwork. This means that motivation for process improvements often
arises naturally in manufacturing cells.

Manufacturing cells are natural candidates for just-in-time (JIT)

implementation.

Cell Design
Design of cellular manufacturing system is a complex exercise with
broad implications for an organization.

The cell design process involves issues related to both system

structure and system operation

Structural issues include:

1. Selection of part families and grouping of parts into
families
2. Selection of machine and process populations and grouping of
these into cells
3. Selection of tools, fixtures, and pallets
4. Selection of material-handling equipment
5. Choice of equipment layout
 Ö. Anlağan
1996

Issues related to procedures include:

1. Detailed design of jobs

2. Organization of supervisory and support personnel around the
cellular structure
3. Formulation of maintenance and inspection policies
4. Design of procedures for production planning, scheduling,
control, and acquisition of related software and hardware
5. Modification of cost control and reward systems
6. Outline of procedures for interfacing with the remaining
manufacturing system (in terms of work flow and information,
whether computer controlled or not)

Evaluation of Cell Design Decisions

The evaluation of design decisions can be categorized as related to

either the system structure or system operation. Typical considerations
related to system structure include:

1. Equipment and tooling investment (low)

2. Equipment relocation cost (low)
3. Material-handling costs (low)
4. Floor space requirements (low)
5. Extent to which parts are completed in a cell (high)
6. Flexibility (high)

Evaluations of cell system design are incomplete unless they relate

to the operation of the system.

A few typical performance variables related to system operation

are:

1. Equipment utilization (high)

2. Work-in-process inventory (low)
3. Queue lengths at each workstation (short)
4. Job throughput time (short)
5. Job lateness (low)
 Ö. Anlağan
1996
A major problem throughout the cell design process is the necessity of
trading off against each other objectives related to structural parameters
and performance variables.

For example, higher machine utilization can be achieved if several cells

route their parts through the same machine. The drawbacks are increased
queuing and control problems.

System cost and performance are affected by every decision related

to system structure and system operation.

It is necessary to evaluate each important design parameter and relate its

performance to pre-established criteria. For example, structural variables
such as number of machines must be balanced against operational
variables such as machine utilization and throughput time using analytical
and simulation approaches.

CELL FORMATION APPROACHES

Machine - Component Group Analysis

Machine - Component Group Analysis is based on production production
flow analysis

Production flow analysis involves four stages:

Stage 1: Machine classification. Machines are classified on the
basis of operations that can be performed on them. A
machine type number is assigned to machines capable of
performing similar operations.

Stage 2: Checking parts list and production route information.

For each part, information on the operations to be
undertaken and the machines required to perform each of
these operations is checked thoroughly.

Stage 3: Factory flow analysis.

This involves a micro-level examination of flow of
components through machines. This, in turn, allows the
problem to be decomposed into a number of machine-
component groups.
 Stage 4: Machine-component group analysis.
An intuitive manual method is suggested to manipulate the
matrix to form cells. However, as the problem size
becomes large, the manual approach does not work.
Therefore, there is a need to develop analytical approaches
to handle large problems systematically.

EXAMPLE:

Consider a problem of 4 machines and 6 parts. Try to group them.

Components
Machines 1 2 3 4 5 6
M1 1 1 1
M2 1 1 1
M3 1 1 1
M4 1 1 1

Components
Machines 2 4 6 1 3 5
M1 1 1 1
M2 1 1 1
M3 1 1 1
M4 1 1 1

Rank Order Clustering Algorithm:

Rank Order Clustering Algorithm is a simple algorithm used to form
machine-part groups.

Step 1: Assign binary weight and calculate a decimal weight for

each row and column using the following formulas:
 Ö. Anlağan
1996

b 2
m
Decimal weight for row i = ip
m-p

p=1

b 2
n
Decimal weight for column j = pj
n p

p=1

Step 2: Rank the rows in order of decreasing decimal weight values.

Step 3: Repeat steps 1 and 2 for each column.

Step 4: Continue preceding steps until there is no change in the position

of each element in the row and the column.

EXAMPLE:
Consider a problem of 5 machines and 10 parts. Try to group them by
using Rank Order Clustering Algorithm.

Components
Machines 1 2 3 4 5 6 7 8 9 10
M1 1 1 1 1 1 1 1 1 1
M2 1 1 1 1 1
M3 1 1 1 1
M4 1 1 1 1 1 1
M5 1 1 1 1 1 1 1 1
Table 1
 Ö. Anlağan
1996

Binary weight
9 8 7
2 2 2 2 6 25 2 4 23 2 2 21 20

Components
Decimal
Machines 1 2 3 4 5 6 7 8 9 10 equivalent
M1 1 1 1 1 1 1 1 1 1 1007
M2 1 1 1 1 1 451
M3 1 1 1 1 568
M4 1 1 1 1 1 1 455
M5 1 1 1 1 1 1 1 1 1020
Table 2

Binary weight
9
2 28 27 26 25 24 23 22 21 20

Components
Binary Machines
weight
1 2 3 4 5 6 7 8 9 10
24 M5 1 1 1 1 1 1 1 1
23 M1 1 1 1 1 1 1 1 1 1
22 M3 1 1 1 1
21 M4 1 1 1 1 1 1
20 M2 1 1 1 1 1
Decimal
equivalen
28 27 27 27 28 20 28 26 11 11
t
Table 3
 Ö. Anlağan
1996

Binary weight
9
2 28 27 26 25 24 23 22 21 20

Components
Binary Machines Decimal
weight
1 5 7 2 3 4 8 6 9 10 equivalent
24 M5 1 1 1 1 1 1 1 1 1020
23 M1 1 1 1 1 1 1 1 1 1 1019
22 M3 1 1 1 1 900
21 M4 1 1 1 1 1 1 123
20 M2 1 1 1 1 1 115
Decimal
equivalen
28 28 28 27 27 27 26 20 11 11
t
Table 4

Similarity Coefficient-Based Approaches

In similarity coefficient methods, the basis is to define a measure of

similarity between machines, tools, design features, and so forth and then
use it to form part families and machine groups.

Single-Linkage Cluster Analysis (SLCA):

It is a hierarchical machine grouping method known as single-linkage

cluster analysis using similarity coefficients between machines.

The procedure is to construct a tree called a dendrogram.

The similarity coefficient between two machines is defined as the ratio of

the number of parts visiting both machines and the number of parts
visiting one of the two machines:

X
N

ijk

Sij = k=1

(Y
N

ik + Z jk - Xijk )
k 1
 Ö. Anlağan
1996
where: Xijk = operation on part k
performed both on machine i and j,
Yik = operation on part k performed on machine i,
Zjk = operation on part k performed on machine j.

SLCA ALGORITHMS

It helps in constructing dendrograms.

A dendrogram is a pictorial representation of bonds of similarity

between machines as measured by the similarity coefficients.

The steps of algorithm are as follows:

Step 1: Compute similarity coefficients for all possible pairs of machines,

Step 2: Select the two most similar machines to form the first machine
cell,
Step 3: Lower the similarity level (threshold) and form new machine cells
by including all the machines with similarity coefficients not less than the
threshold value,
Step 4: Continue step 3 until all machines are grouped into a single cell.

EXAMPLE:
Consider the matrix of 5 machines and 10 components given below.

Components
Machines 1 2 3 4 5 6 7 8 9 10
M1 1 1 1 1 1 1 1 1 1
M2 1 1 1 1 1
M3 1 1 1 1
M4 1 1 1 1 1 1
M5 1 1 1 1 1 1 1 1
Develop a dendogram and discuss the resulting cell structures.

Step 1: Determine similarity coefficients between all pairs of machines.

 Ö. Anlağan
1996

SC12 = 5 = 0.556
9 +5-5
Similarity coefficients of machine pairs
Machine M1 M1 M1 M1 M2 M2 M2 M3 M3 M4
pairs M2 M3 M4 M5 M3 M4 M5 M4 M5 M5
SC 0.55 0.30 0.67 0.70 0.00 0.83 0.30 0.00 0.50 0.40

Step 2: Select machines M2 and M4, having the highest similarity

coefficients of 0.83, to form the first cell.

Step 3: The next lower coefficient of similarity is between machines M1

and M5. Use these machines to form the second cell.

Step 4: The next lower coefficient of similarity is now 0.67 between

machines M1 and M4. At this threshold value machines M1, M2, M4,
and M5 will form one machine group. The other possible groups will be
evaluated by the same way.

M4 M2 M5 M1 M3
0.83
0.70
0.67
0.50

0.00
Dendrogram
 Ö. Anlağan
1996
EXCEPTIONAL PARTS & BOTTLNECK MACHINES:

One of the important goal in cell design is to create mutually independent

machine cells. However, it may not always be economical or practical to
achieve this goal.

In practice, therefore, some parts need to be processed in more than one

cell. These are known as exceptional parts and the machines processing
them are known as bottleneck machines.

The problem of exceptional elements can possibly be eliminated by:

 Generating alternative process plans

 Duplication of machines
 Subcontracting these operations
Ö. Anlağan
1996
EVALUATION OF CELL DESIGN:

In design of cells, there will be more than one alternative solution. The
objective is to find the best alternative.

Assume we have the following alternative cell configuration:

Similarity Number of Cell configuration

coefficient cells formed
1.00 5 (M1), (M2), (M3), (M4), (M5)
0.83 4 (M2, M4), (M5), (M1), (M3)
0.70 3 (M2, M4), (M1, M5), (M3)
0.67 2 (M1, M2, M4, M5), (M3)
0.50 1 (M1, M2, M3, M4, M5)

The criteria is to minimize the distance that the parts should travel during
the processes; in other words, to minimize the material handling costs of
intercell (between cells) and intracell (within cell) movements of the
parts.
 Ö. Anlağan
1996
The following factors affect the cost of intercell and intracell movements
of parts.
1. The layout of machines in a group
2. The layout of machine groups
3. The sequences of parts through machines and machine groups

The total distances moved by a component visiting a number of machines

in a cell has to be determined.

Assumptions:
1. In the absence of the real data on the sequences in which the
components visit the machines, it is assumed that the machines
are laid out in a random manner.
2. There is one unit distance between each machine in a group of
N machines.
3. A part has to visit two machines in a group of N machines.

Expected distance for a straight-line layout:

N +1
3
Expected distance for a rectangle layout of M rows of L machines:

M+L
2

Expected distance for a square layout:

2 N
3
The total distance moved in jth cell for the ith configuration:

d k
m

ij ij
j
 Ö. Anlağan
1996
where:
dij = expected distance moved between two machines for ith
configuration in jth cell
kij = number of moves between two machines by all the parts for ith
configuration in jth cell

The total cost of intercellular and intracellular movements (TCi) for the
ith configuration:

d k
m
TCi = C1Ni + C2 ij ij
j

where:
C1 = cost of an intercell movement
C2 = cost per unit distance of an intracell movement
Ni = number of intercell movements for ith configuration
EXAMPLE:
Consider the following cell configuration.
Components
Machines 1 5 2 3 4 7 8 9 10 6
M1 1 1 1 1 1 1 1 1 1
M5 1 1 1 1 1 1 1 1
M2 1 1 1 1 1
M4 1 1 1 1 1 1
M3 1 1 1 1
Consider 3-cell case:

Expected movement distance,

in cell (M1, M5) =  23 1  1
in cell (M2, M4) = 
3 
21  1

in cell (M3) = 0

The number of moves passing through two machines by all the parts,
in cell (M1, M5) = 7
in cell (M2, M4) = 5
in cell (M3) = 0
 Ö. Anlağan
1996
The total distance for all intercell moves for 3-cell configuration:

1 x 7 + 1x 5 + 0 = 12

The number of intercell moves in 3-cell configuration is 10.

Assume:
C1 = $2.00 (cost of intercell unit movement)
C2 = $1.00 (cost of intracell unit movement)

The total cost of intercell and intracell movements in 3-cell configuration:

2.00 $ x 10 + 1.00 $ x 12 = 32.00 $

The summary of cost calculation for all possible cell configuration is given in the following table:
Cell Number Total Total cost of
configuration of distance of intercell and
intercell intracell intracell
moves moves moves
5-cells (M1), (M2), 22 0 2 x 22 +
(M3), (M4), 1 x 0 = 44
(M5)
4-cells (M2, M4), 18 5 2 x 18 +
(M5), (M1), 1 x 5 = 41
(M3)
3-cells (M2, M4), 10 12 2 x 10 +
(M1, M5), 1 x 12 = 32
(M3)
2-cells (M1, M2, 4 30 2x4+
M4, M5), 1 x 30 = 38
(M3)
1-cells (M1. M2, 0 44 2x0+
M3, M4, 1x 44 = 44
M5)
 Ö. Anlağan
1996
In 1988, a survey of 53 show that the use of GT and cellular
manufacturing in US industries has met with success. The benefits
reported from these studies include:

 Reduction in throughput time by 46%

 Reduction in work-in-process inventory by 41%
 Reduction in material handling by 39%
 Reduction in setup time by 32%
 Improvement in quality by 29.6%