CELLULAR MANUFACTURING

Dr. Mohamed A.A. Mansour

Page 1 of 59
Group Technology (GT)
y Group technology (GT) is a manufacturing philosophy to increase production
efficiency by grouping a variety of parts having similarities of shape, dimension
and/or process route.

y GT may be defined as a manufacturing philosophy that justifies batch production

by capitalizing on design and/or manufacturing similarities among component
parts. In batch production, the products are made in small batches and in large
variety.

y Every batch contains identical items but every batch is different from the others.

y For example, a plant producing many parts (say 5000 different parts) may be
grouped into several distinct families (say 20 to 25 part families). Each family
possesses similar design and manufacturing characteristics.

y This grouping philosophy results in increased manufacturing efficiencies.

Page 2 of 59
Group Technology (GT)

y Efficiencies are due to reduced setup times, lower in-process inventories, better
scheduling, streamlined material flow, improved quality, improved tool control
and the use of standardized process plans.

y In many plants where GT has been implemented, the production equipment is

arranged into ‘machine groups’ (also known as ‘cells’) to facilitate work flow and
parts handling.GT is felt advantageous in the product design stage also. GT is a
prerequisite for computer integrated manufacturing.

y GT is not an automation strategy associated with either the design or the

production engineering area, Implementation of GT is a critical first step for
computer-aided process planning (CAPP) and many of the production
engineering activities.

Page 3 of 59
Group Technology (GT)

BENEFITS OF GT

y Group technology, when successfully implemented, offers many benefits to

industries.

y GT benefits can be realized in a manufacturing organization in the following

areas;

1. Product design
2. Tooling and setups
3. Materials handling
4. Production and inventory control
5. Process planning
6. Management and employees.

Page 4 of 59
Group Technology (GT)

1. Product design

y Importance of GT for product design come in cost and time savings.

y Design engineers can quickly and easily search the database for parts that either
presently exist or can be used with slight modifications, rather than issuing new
part numbers.

y Similar cost savings can be realized in the elimination of two or more identical
parts with different part numbers.

y Advantage is the standardization of designs.

y Design features such as comer radii, tolerances, counter bores, and surface
finishes can be standardized with GT.

Page 5 of 59
Group Technology (GT)
2. Tooling and Setups

y In the area of tooling, group jigs and fixtures are designed to accommodate
every member of a part family.

y Also work holding devices are designed to use special adapters in such a way
that this general fixture can accept each part family member.

y Since setup times are very short between different parts in a family, a group
layout can also result in dramatic reductions in setup times.

3. Materials Handling

y GT facilitates a group layout of the shop.

y Since machines are arranged as cells, in a group layout, the materials handling
cost can be reduced by reducing travel and facilitating increased automation.

Page 6 of 59
Group Technology (GT)
4. Production and Inventory Control

y GT simplifies production and planning control.

y Complexity of the problem has been reduced from a large portion of the shop to
smaller groups of machines.

y Production scheduling is simplified to a small number of parts through the

machines in that cell.

y In addition, reduced setup times and effective materials handling result in

shorter manufacturing lead times and smaller work-in-process inventories.

Page 7 of 59
Group Technology (GT)
5. Process Planning

y Concept of group technology parts, classification and coding lead to an

automated process planning system.

y Grouping parts allows an examination of the various planning/route sheets for

all members of a particular family.

y Once this has been accomplished, the same basic plans can be applied to other
members, there by optimizing the shop flow for the group.

Page 8 of 59
Advantages of Group Technology (GT)
y GT facilitates (a) efficient retrieval of similar parts, (b) development of a database containing effective
product design data and (c) avoidance of design duplication.

y GT encourages standardization of designs, tooling, fixing and setups.

y GT facilitates (a) development of a computer-aided process planning (CAPP) system, (b) retrieval of
process plans for part families and (c) development of standard routings for part families.

y Times and costs for material handling and waiting between stages are reduced.

y Production planning and control is simplified.

y Setup time and setup cost for each job are reduced, because several jobs are grouped and processed
in sequence.

y Machining cells can reduce work-in-process inventory, resulting in shorter queues and shorter
manufacturing throughput times.

y Part and product quality are improved.

y GT facilitates better employee involvement and increases workers satisfaction.

Page 9 of 59
Limitations of Group Technology (GT)

y Implementing GT is expensive.

y Large costs may be incurred in rearranging the plant into machine cells or
groups.

y Installing a coding and classification system is very time-consuming.

y As there is no common implementation approach, the implementation of GT is

often difficult.

Page 10 of 59
Part Families

y Part family is a collection of parts which are similar either because of geometric
shape and size or because similar processing steps are required in their
manufacture.

y Parts which are similar in their design characteristics (i.e. shape and geometry)
are grouped in a family referred to as a design part family.

y Parts which are similar in their manufacturing characteristics are grouped in a

family referred to as a manufacturing part family.

y Characteristics used in classifying parts are referred to as “attributes”.

y The two parts are placed in the same family based on design characteristics.

y They have exactly the same shape and size.

y They differ in terms of manufacturing requirements such as tolerances,

production quantities and material.
Page 11 of 59
Part Families
y Design part family

y Manufacturing part family

y A family of parts with similar

manufacturing process
requirements but different
design attributes

Page 12 of 59
Part Families
Methods for Part Family Formation

y The three general methods for grouping parts into families are:

1. Visual inspection

2. Parts classification and coding system

3. Production flow analysis.

Page 13 of 59
Visual Inspection Method

y Visual inspection method is the simplest and least expensive method.

y It involves looking at parts, photos of parts or drawings of parts and arranging

them into similar groups.

Page 14 of 59
Parts Classification and Coding

y Coding is a systematic process of establishing an alphanumeric value for parts

based on selected part features. Classification is the grouping of parts based on
code values.

y It is the most sophisticated, most difficult, most time-consuming and widely

used of the three methods.

y Here the various design and/or manufacturing attributes of a part are identified,
listed and assigned a code number.

y Though several classification and coding systems have been developed, no

system has been universally adopted. one of the reasons for this is that the
information that is to be represented in the classification and coding system will
vary from one company to another company.

Page 15 of 59
Parts Classification and Coding

Design and Manufacturing Attributes

y Any parts classification systems fall into one of the following three categories:

1. Systems based on part design attributes.

2. Systems based on part manufacturing attributes.

3. Systems based on both design and manufacturing attributes.

Page 16 of 59
Parts Classification and Coding

y Parts classified by design attributes can be coded from information on the

engineering drawing. This first category systems are useful for design retrieval
and to promote design standardization.

y In grouping of manufacturing attributes, in addition to drawing information,

other information such as operation sequence, lot size, machines used,
production processes, surface finish, etc. are also considered.

y Systems in the second category are used for computer-aided process planning,
tool design and other production related functions.

y The third category represents an attempt to combine the functions and

advantages of the other two systems into a single classification scheme..

Page 17 of 59
Parts Classification and Coding

Coding System Structure

y A GT code is a string of characters capturing information about an item.

y A coding scheme is a vehicle for the efficient recording, sorting and retrieval of
relevant information about objects.

y A part coding scheme consists of a sequence of symbols that identify the part’s
design and/or manufacturing attributes.

y The symbols in the code can be all numeric, all alphabetic or a combination of
both types.

Page 18 of 59
Parts Classification and Coding
1. Hierarchical Code (or Mono code)

y Interpretation of each successive symbol depends on the value of the preceding

symbols.

y Each symbol amplifies the information contained in the preceding digit, so a

digit in the code cannot be interpreted alone. Structure of these codes is like a
tree in which each symbol amplifies the information provided in the previous
digit.

y Hierarchical coding system can be depicted using a tree structure as shown in

Figure.

Page 19 of 59
Parts Classification and Coding

Merits and demerits of mono code system:

y Provides a large amount of information in a relatively small number of digits.

y This tree structure works well for designing an existing ordered structure but is
more difficult to use in classifying things that have no apparent order.

y Defining the meanings for each digit in a hierarchical system (and hence the
construction) is difficult.

y Frequently used in design departments for part retrieval.

y Their utility is limited in manufacturing departments, because it is difficult to

retrieve and analyze process-related information when it is in a hierarchical
structure.

Page 20 of 59
Parts Classification and Coding

2. Attribute Code (or Poly code)

y In this structure, the interpretation of each symbol in the sequence does not
depend on the value of preceding symbols.

y Each digit in this code represents information in its own right and does not
directly qualify the information provided by the other digits.

y Attribute code is also known by other names ‘poly code’, ‘chain code’, ‘discrete
code’ and ‘fixed-digit code’.

Page 21 of 59
Parts Classification and Coding

y Illustration: shows an example for attribute code.

y For the spur gear shown in Figure. using code, we can obtain the poly code as
“22213”.

Page 22 of 59
Parts Classification and Coding
Merits and demerits of poly codes:

y The major advantages of poly codes are that they are compact and easy to use
and develop.

y It is popular with manufacturing departments because it makes it easy to

identify parts that have similar features that require similar processing.

y Because a poly code represents a class of items as a string of features, it is also

particularly suitable for computer analysis.

y The primary disadvantage is that, for comparable code size, a poly code lacks
the detail presence in a mono code structure. also poly codes tend to be longer
than mono codes.

Page 23 of 59
Parts Classification and Coding
3. Decision-Tree (or Hybrid) Code

y A hybrid code captures the best features of the hierarchical and poly code
structures.

y This system is also known as decision-tree coding and it combines both design
and manufacturing attributes.

y In practice, most coding systems use a hybrid construction to combine the best.

y To reduce the length of a strict poly code, the first digit of such a system may
split the population into appropriate subgroups, as in a mono code structure.
Then each subgroup can have its own poly code structure.

y For example, the first digit might be used to denote the type of part, such as
gear.

y The next four positions might be reserved for a short attribute that would
describe the attributes of the gear.
Page 24 of 59
Parts Classification and Coding

y The next digit position 6, might be used to designate another subgroup, such as
material, followed by another attribute code that would describe the attributes.
Thus, a hybrid code can be generated.

y Hybrid code is relatively more compact than a pure attribute code while
retaining the ability to easily identify parts with specific characteristics.

Page 25 of 59
Production Flow Analysis

y Developed by Burbridge in 1971, Is a method for identifying part families and

associated machine groupings that uses the information contained on production
route sheets rather on part drawings.

y Work parts with identical or similar routings are classified into part families.
y PFA neither uses a classification and coding system nor part drawings to identify
families.

y It uses the information such as part number, operation sequence, lot size, etc.,
contained on the route sheet.

y This method is based on the route sheet information and sometimes referred as
the route sheet inspection method.

Page 26 of 59
Production Flow Analysis

Steps Involved in PFA

y The following four steps are followed to carryout PFA:

i. Data collection

ii. Sortation of process routings

iii. Preparation of PFA chart

iv. Cluster analysis.

Page 27 of 59
Production Flow Analysis
Step 1: Data collection

y The step in the PFA procedure is to collect the necessary data.

y Route sheets of all the components to be manufactured in the shop are
prepared.
y Route sheet should contain the part number and operation sequence.
y Other data that can be collected/obtained from route sheet/operation sheet
include lot size, time standards and annual demand.

Step 2: Sortation of process routes

y The second step in the PFA is to arrange the parts into groups according to the
similarity of their process routings.
y A typical card format is required for organizing the data such as the part
number, sequence of code and lot size. A sortation procedure is used to arrange
the parts into ‘packs’.
y Pack is nothing but a group of parts with identical process routings. Some pack
may even contain only one part number. A pack identification or letter is
provided for each pack.

Page 28 of 59
Production Flow Analysis

Step 3: PFA chart

y A PFA chart is a graphical representation of the process used for each pack.

y It is a tabulation of the process or machine code numbers for all of the part
packs. Also known as ‘part-machine incidence matrix’ or ‘component-machine
incidence matrix’.

y The table below Illustrates a typical PFA chart having 7 machines (M1 to M7)
and 9 parts (P1 to P9).

Page 29 of 59
Production Flow Analysis
y In this matrix, the entries have a value xij = 1 or 0:

y A value of xij = 1 indicates that the corresponding part i requires processing on

machine j

y xij = 0 indicates that no processing of component i is accomplished on machine

j.

y However, in Table , the 0’s are indicated as blank (entry) entries for better
clarity of the matrix.

Page 30 of 59
Production Flow Analysis
Step 4: Cluster analysis

y From the PFA chart, related grouping are identified and rearranged into a new
pattern that brings together packs with similar machine sequences.

y Table shows one possible rearrangement of the original PFA chart.

y It is clear that for the PFA chart considered we have three part families and
three machine cells, as shown below.

Page 31 of 59
Production Flow Analysis

Table : Rearranged PFA chart, indicating possible machine grouping

Part Families: Cell groups

Page 32 of 59
Production Flow Analysis
Advantages of PFA

y Parts classification and coding uses design data and the PFA uses manufacturing
data (i.e., route sheet) to identify part families.

y Due to this fact, as pointed out by Groover, PFA can overcome two possible
anomalies that can occur in parts classification and coding.

y First, parts whose basic geometries are quite different may nevertheless require
similar or identical process routings.

y Second, parts whose geometries are similar may nevertheless require process
routings that are quite different.

y Also PFA requires less time than a complete parts classification and coding
procedure.

Page 33 of 59
Production Flow Analysis
Disadvantages of PFA

y PFA does not provide any mechanism for rationalizing the manufacturing
routings.

y No consideration being given to routing sheet whether the routings are optimal
or consistent or logical.

y Process sequences from route sheets are prepared by different process

planners, hence the routings may contain processing steps that are non-
optimal, illogical and unnecessary.

Page 34 of 59
Cellular Manufacturing
y It is an application of group technology in which dissimilar machines have been
aggregated into cells, each of which is dedicated to the production of a part
family.

y Primary advantage of CM implementation is that a large manufacturing system

can be decomposed into smaller subsystems of machines called cells. Cells are
dedicated to process part families based on similarities in manufacturing
requirements. Parts having similar manufacturing requirements can be
processed entirely in that cell.

y In addition, cells represent sociological units conducive to team work which lead
to higher levels of motivation for process improvements.

y Benefits associated with the application of CM include improved market

response, more reliable delivery promises, reduced tooling and fixtures and
simplified scheduling.

y Literature surveys confirm substantial benefits from implementing cellular

manufacturing in manufacturing industries.
Page 35 of 59
Cellular Manufacturing

Page 36 of 59
Cellular Manufacturing
y Design Considerations Guiding the Cell Formation

y We know that cell formation is the early activity in the cell design process where
part families and associated machine groups are identified. Cell formation is
influenced by a variety of objectives and concerns.

y Lists the important design considerations that should be taken into account
during cell formation.

Page 37 of 59
Composite Part Concept
y Mitrofanov (1959) and Edwards (1970) have proposed composite part approach
to implement the concept of cellular manufacturing.

y A composite part is formed by merging the primitives of all the parts of a part
family.

y Composite is a single hypothetical part that can be completely processed in a

manufacturing cell/group.

y If a new part is loaded in a machine group, the degree of dissimilarity of the

part of its related part family or the hypothetical composite should have
minimum deviation and desired to be zero.

y The manufacturing facility could be planned on the basis of composite part to

facilitate economical production.

Page 38 of 59
Composite Part Concept
y The primitives of three parts shown are merged into composite part by
incorporating all the primitives of the three parts

Page 39 of 59
Composite Part Concept
y It may not be judicious to merge all the primitives of parts due to various
production considerations,

y In such situation the shop will converge back to a large job shop and all the
benefits of CMS will be lost. Size of the manufacturing group depends on initial
capital investment capacity, machines available and outsourcing facilities.

y Individual parts features (in terms of primitives) could be merged in the

composite part based on their repetitions in the parts.

y Primitives having more repetitions will be more eligible candidates for merging in
the composite part.

y Various techniques could be used for selection of optimum primitives for

merging in composite parts.

y Genetic algorithm is proved to be one of the effective techniques.

Page 40 of 59
Machine Cell Design and Layout
y Machine layout aims at determining the best arrangement of machines in each
product cell.

y Minimization of material handling cost is an often used objective in determining

the layout of machines in a cell.

y Constraints related to the availability of space, material handling system type

and so on are considered.

y Type of operations and parts are not the only factors that impact the layout of
machines.

y Type of material handling system to be used also needs to be considered;

Page 41 of 59
Machine Cell Design and Layout
y Example, the articulated robot (R) in figure(a) implies a circular arrangement of
machines.

y If an AGV had been selected to tend the same machines, it would have been
necessary to use the layout in figure(b).

y Two step design of system

Page 42 of 59
Machine Cell Design and Layout
y Goal of machine cell layout is to arrange the product or functional cells formed
on the factory floor.

y Determining the layout of machine cells involves locating the cells in order to
minimize the total material handling cost subject to some constraints (e.g. shape
of the facility).

y If all cells were square in shape and of the same size, then the cell layout could
be modelled as the quadratic assignment problem (QAP).

y Cell layout problem can be viewed as a machine layout problem, where each
machine represents a cell.

y Though cellular manufacturing offers numerous benefits, it is not always

implemented due to the following:

1. Parts and machines may not form mutually exclusive clusters.

2. The data required from the formation of cells might not be available.
Page 43 of 59
Quantitative Analysis in Cellular Manufacturing
Rank Order Clustering Method

y It also known as binary ordering algorithm (BOA), It is a simple algorithm used

to form machine-part groups. it was Developed by J. R. King (1980).

y It considers two data:

1. Number of components and Component sequences. Based on the

component sequences, a machine-part incidence matrix is developed.

2. Rows of the machine component incidence matrix represent the

machines which are required to process the components. Columns of the
matrix represent the component numbers.

Page 44 of 59
Quantitative Analysis in Cellular Manufacturing
Concept:

y Each row and each column of the matrix are considered as binary words.
y Example, in a row if we have numbers (1 0 1 0 1), then the decimal equivalent
is computed as follows:

Page 45 of 59
Quantitative Analysis in Cellular Manufacturing
Steps in ROC Algorithm

The steps in ROC algorithm are given below:

Step 0: Input: Total number of components, component sequences.

Step 1: Form the machine-component incidence matrix using the component sequences.

Step 2: Compute binary equivalent of each row.

Step 3: Rearrange the rows of the matrix in rank wise (high to low from top to bottom).

Step 4: Compute binary equivalent of each column and check whether the columns of the matrix are
arranged in rank wise (high to low from left to right). If not, go to Step 7.

Step 5: Rearrange the columns of the matrix rank wise and compute the binary equivalent of each row.

Step 6: Check whether the rows of the matrix are arranged rank wise, If not, go to Step 3; otherwise, go
to Step 7.

Step 7: Print the final machine-component incidence matrix.

Page 46 of 59
Quantitative Analysis in Cellular Manufacturing
Arranging Machines in a GT cell

There are three basic ways to arrange machines in a GT cell are:

1. Line (or product) layout.

2. Functional (or process) layout.

3. Group (or combination) layout.

Page 47 of 59
Quantitative Analysis in Cellular Manufacturing
1) Line (or Product) Layout

y Here the machines are arranged in the sequence as required by the product.

y If volume of production of one or more products is large, the facilities can be

arranged to achieve efficient flow of materials and lower cost per unit.

Suitability:

y Suitable for the continuous mass production of goods as it makes it possible for
the raw material to be fed into the plant and take out finished product on the
other end.

Page 48 of 59
Quantitative Analysis in Cellular Manufacturing

Page 49 of 59
Quantitative Analysis in Cellular Manufacturing
2) Functional (or Process) Layout

y Characterized by keeping similar machines, operations at one location, i.e. all

lathes at one place, all milling machines at another place.

y In process layout, machines are arranged according to their functions.

Suitability: Suitable for job order/non-repetitive type production.

Page 50 of 59
Quantitative Analysis in Cellular Manufacturing

Page 51 of 59
Quantitative Analysis in Cellular Manufacturing
3) Group (or Combination) Layout

y It is a combination of the product layout and process layout.

y This layout Combines the advantages of both layout systems.

y Here machines are arranged into cells , each cell is capable of performing
manufacturing operations on one or more families of part.

y If there are m machines and n components, in a group layout, the m-machines

and n - components will be divided into distinct number of machines-component
cells (groups) such that all the components assigned to a cell are almost
processed within that cell itself.

y Objective is to minimize the inter-cell movements.

Suitability: Preferred for batch type production, where the products are made in
small batches and in large variety.

Page 52 of 59
Quantitative Analysis in Cellular Manufacturing
3) Group (or Combination) Layout

y If there are m machines and n components, in a group layout, the m-machines

and n - components will be divided into distinct number of machines-component
cells (groups) such that all the components assigned to a cell are almost
processed within that cell itself.

y Objective is to minimize the inter-cell movements.

Suitability: Preferred for batch type production, where the products are made in
small batches and in large variety.

Page 53 of 59
Quantitative Analysis in Cellular Manufacturing
3) Group (or Combination) Layout

Page 54 of 59
Hollier Method-Simple Problems

Hollier Method 1:
The first method uses the sums of flow "From" and "To" each machine in the cell. The
method can be outlined as follows:

1. Develop the From - To chart from part routing data. The data
contained in the chart indicates number of part moves between the machines
for workstations).
2. Determine the "From” and "To" sums for each machine. This is
accomplished by summing all of the "From" trips and "To" trips for each
machine (or operation).The "From" sum for a machine is determined by
adding the entries in the corresponding row and the "To" sum is found by
adding the entries in the corresponding column.
3. Assign machines to the cell based on minimum "From" or To sums.
The machine having the smallest sum is selected. If the minimum value is a
"To" sum, then the machine is placed at the beginning of the sequence. If the
minimum value is a “From” sum, then the machine is placed at the end of the
sequence. Tie breaker rules:

Page 55 of 59
Hollier Method-Simple Problems
Hollier Method 1:

Tie breaker rules:

y If a tie occurs between minimum. "To" sums or minimum "From"

sums, then the machine with the minimum “From/To” ratio is
selected.

y If both "To" and "From" sums are equal for a selected machine, it
is passed over and the machine with the next lowest sum is
selected.

y If a minimum "To" sum is equal to a minimum "From" sum, then

both machines are selected and placed at the beginning and the
end of the sequence respectively.

Page 56 of 59
Hollier Method-Simple Problems
Hollier Method 1:

1. (a) If a tie occurs between minimum. "To" sums or minimum "From" sums,
then the machine with the minimum “From/To” ratio is selected.

2. (b) If both "To" and "From" sums are equal for a selected machine, it is
passed over and the machine with the next lowest sum is selected.

3. (c) If a minimum "To" sum is equal to a minimum "From" sum, then both
machines are selected and placed at the beginning and the end of the
sequence respectively.

y Reformat the From-To chart After each machine has been selected, restructure
the From-To chart by eliminating the row and column corresponding to the
selected machine and recalculate the "From" and "To" sums. Repeat steps 3 and
4 until all machines have been assigned.

Page 57 of 59
Hollier Method-Simple Problems
Hollier Method 2:
y This approach is based on the use of From/To ratios formed by summing the total flow from and to
each machine in the cell. The method can be reduced to three steps:

1. Develop the From—To chart. This is the same step as in Hollier Method 1.

1. Determine the From/To ratio for each machine. This is accomplished by summing up all of
the "From" trips and "To" trips for each machine (or operation). The "From" sum for a machine is
determined by adding the entries in the corresponding row and the "To" sum is determined by
adding the entries in the corresponding column. For each machine, the From/To ratio is calculated
by taking the "From" sum for each machine and dividing by the respective "To" sum.

1. Arrange machines in order of decreasing From/To ratio. Machines with a high From/To
ratio distribute work to many machines in the cell but receive work from few machines. Conversely
machines with a low From/To ratio receive more work than they distribute. Therefore, machines
are arranged in order of descending From/Ip ratio. That is, machines with high ratios are placed at
the beginning of the work flow and machines with low ratios are placed at the end of the work
flow. In case of a tie, the machine with the higher "From" value is placed ahead of the machine
with a lower value.

Page 58 of 59
Hollier Method-Simple Problems
Percentage of in-sequence moves

y Percentage of backtracking moves.

y The percentage of in sequence moves is computed by adding all the values

representing in sequence moves and dividing by the total number of moves.

y The percentage of back tracking moves is determined by summing all of the

values representing back tracking moves and dividing by the total number of
moves.

Page 59 of 59