1

Chapter 18 Lean Manufacturing

Objective:
In this chapter, we introduce the fundamentals of Lean Manufacturing. Concepts
of waste elimination are discussed. Components for Lean including: Waste
identification and elimination, set-up reduction, part families, cell formation, cell
design, batches of one and pull systems are also discussed.

“Perfection is not attainable. But if we chase perfection, we can catch

excellence.” Vince Lombardi

18.1 Introduction to Lean Manufacturing

Lean manufacturing or lean production are reasonably new terms that can be
traced to Jim Womack, Daniel Jones and Daniel Roos’ book, The Machine that
changed the world [1991]. In the book, the authors examined the
manufacturing activities exemplified by the Toyota Production System. Lean
manufacturing is the systematic elimination of waste. As the name implies, lean
is focused at cutting “fat” from production activities. It has also been successfully
applied to administrative and engineering activities as well. Although lean
manufacturing is a relatively new term, many of the tools used in lean can be
traced back to Fredrick Taylor and the Gilbreaths at the turn of the 20th century.
What Lean has done is to package some well-respected industrial/manufacturing
engineering practices into a system that can work in virtually any environment.

Figure 18.1 provides a definition of lean as a function of the outcomes that one
realizes. The definition comes from Womack and it identifies the results rather
than the method of lean. In the following sections, the procedures and specifics
of lean will be introducted.

18.1.1 The 3 M’s of Lean

Lean manufacturing is a Japanese method focused on 3M’s. These Ms are:

muda, the Japanese word for waste , mura, the Japanese word for
inconsistency, and muri, the Japanese word for unreasonableness. Muda
specifically focuses on activities to be eliminated. Within manufacturing, there
are categories of waste. Waste is broadly defined as anything that adds cost to
the product without adding value to it. Generally, muda (or waste) can be
grouped into the following categories:

1.Excess production and early production

2.Delays
3.Movement and transport
 2

Definition of “Lean”
• Half the hours of human effort in the factory
• Half the defects in the finished product
• One-third the hours of engineering effort
• Half the factory space for the same output
• A tenth or less of in-process inventories

Source: The Machine that Changed the World

Womack, Jones, Roos 1990

Figure 18.1 An early definition of Lean.

4.Poor process design

5.Inventory
6.Inefficient performance of a process
7.Making defective items

These wastes are illustrated in Figure 18.2

Excess production results in waste because it captures resources too early and
retains the value that is added until the product can be used (sold). In today’s
highly changing society, many items produced before they can are sold to a
specific customer often go obsolete before demand is realized. This means that
a perfectly good product is often scrapped because it is obsolete. Producing a
product simply to keep a production resource busy (either machine, operator or
both) is a practice that should be avoided.

Delays, such as waiting for raw material, also result in the poor use of capacity
and increased delivery time. Raw materials and component parts should be
completed at approximately the time that they will be required by downstream
resources. Too early is not good, but late is even worse.

Movement and transportation should always be kept to a minimum. Material

handling is a non-value added process that can result in three outcomes: 1) the
product ends up at the right place at the right time and in good condition, 2) the
 3

7 Forms of Waste
CORRECTION

Repair or MOTION
WAITING
Rework Any wasted motion
Any non-work time to pick up parts or
waiting for tools, stack parts. Also
supplies, parts, etc.. wasted walking

Types
PROCESSING of OVERPRODUCTION
Producing more
Doing more work than Waste than is needed
is necessary
before it is needed

INVENTORY
Maintaining excess CONVEYANCE
inventory of raw mat’ls, Wasted effort to transport
parts in process, or materials, parts, or
finished goods. finished goods into or
out of storage, or
between
processes.

Figure 18.2 The seven Forms of waste.

part ends up in the wrong place, and 3) the part is damaged in transit and
requires rework or scrap. Two of the three outcomes are no desirable, which
further leads to minimizing handling. Because material handling occurs between
all operations, when possible, the handling should be integrated into the process,
and the transport distances minimized.

A poorly designed process results in overuse of manufacturing resources (men

and machines). There are no perfect processes in manufacturing. Generally,
process improvements are made regularly with new efficiencies embedded within
the process. Continuous process improvement is a critical part of Lean
Manufacturing.

Excess inventory reduces profitability. Today, it is not uncommon for a

manufacturer to store a supplier’s product at the production site. The supplier,
right up until the time that they are drawn from inventory, owns the materials. In
many ways this is advantageous to both the user and supplier. The supplier
warehouses his material offsite, and the user does not need to commit capital to
a large “safety stock” of material.
 4

Insufficient (or poor) process performance always results in the over utilization of
manufacturing resources and a more costly product. There is no optimal process
in that improvements can always be made; however, many processes operate far
below the desired efficiency. Continuous process improvement is necessary for
a manufacturing firm to remain competitive. Excess movement or unnecessary
part handling should be the first targets of waste elimination.

Poor quality (making defects) is never desirable. Labor and material waste
results from producing any defect. Furthermore, the cost of mitigating poor
quality (rework) can often exceed the price of the product. A critical balance
between processing speed and quality exists. A process should be run as fast
as possible without sacrificing acceptable quality.

From the above discussion, it should be obvious that waste is a constant enemy
of manufacturing. Waste elimination should be an on-going process that focuses
on improving a process regularly. Regular reviews and worker input should be
conducted as often as allowable.

The second “M” is for mura, or inconsistency. Inconsistency is a problem that

increases the variability of manufacturing. Mura is evidenced in all
manufacturing activities ranging from processing to material handling to
engineering to management. Figures 18.3 and 18.4 illustrate a characterization
of mura as Henry Ford wrote in 1926.

Quality Processes Yield Quality Results

Inconsistent Inconsistent
Process Results

Traditional = People doing whatever they can to get results

Consistent Desired
Process Results

Lean = People using standard process to get results

Figure 18.3 Inconsistency is a problem in manufacturing.

(See Figure 18.4), the best target method should be initially identified. This
method should become the “standard practice”, and the target for future process
improvements.
 5

Henry Ford - Standards

“To standardize a method is to choose out of the many
methods the best one, and use it. Standardization means
nothing unless it means standardizing upward.
Today’s standardization, instead of being a barricade against
improvement, is the necessary foundation on which
tomorrow’s improvement will be based.
If you think of “standardization” as the best that you know
today, but which is to be improved tomorrow - you get
somewhere. But if you think of standards as confining, then
progress stops.”
Henry Ford, 1926
Today & Tomorrow

Figure 18.4 Henry Ford on standards (or against inconsistency).

The final “M” is for muri or unreasonableness. Muri applies to a variety of

manufacturing and management activities. For instance, Figure 18.5 shows an
example of being unreasonable by blaming someone for problems rather than
looking at resolution of problems. It is unreasonable to blame rather than
mitigate issues. This is true for all manufacturing activities -- do what is
reasonable. Don’t be emotional!

New Paradigm: Non-Blaming Culture

Management creates a culture where:

• Problems are recognized as opportunities

• It’s okay to make legitimate mistakes
• Problems are exposed because
of increased trust
• People are not problems -
they are problem solvers PROBLEMS SOLUTIONS

• Emphasis is placed on finding solutions instead of

“who did it”

Figure 18.5 Be reasonable -- muri.

 6

18.1.1 Active Exercises

Waste elimination and process improvement can be applied to virtually

any process. For example, the local Department of Motor Vehicles (DMV)
provides a variety of services. Examples of those include: 1) Administering a
driving test, 2) Testing vision, 3) Testing applicant’s knowledge of driving
practice, 4) Registering vehicles and much more. The past ten years has seen
many improvements for many of these activities. For instance, and interactive
computer system is used to test a person’s understanding of driving practice. A
computer system is also used to issue the license. These are typical process
improvements. Changes at the DMV were result of waste identification and
process improvement.

For this exercise, you should identify a process at work or school that you
feel lends itself to waste elimination. You should first characterize the process,
noting the value added and non-value added activities. The non-value added
activities should be eliminated (if possible). The value added activities should be
examined and analyzed for potential improvements. Describe the system as it
exists and as you think it should be. Describe how you went about identifying
and eliminating waste.

18.2 Laying out a Lean Production Facility

Another critical aspect of Lean is the organization of the production facility.

Since one of the keys to lean is waste elimination, the layout of any system
should be arranged in such manners that waste of motion (material handling and
material transport) and elimination of inve ntory is part of the object for the layout.
You may recall that there are two traditional forms of layout in manufacturing:
process and product. In a process layout (or job shop as it is informally
called), machines are organized and clustered by type, where typically all mills
are in one department, all lathes in another, etc. In a product layout (or flow
shop), machines are located so that sequential operations are performed at
adjacent machines. These types of layout are illustrated in Figures 18.6 and
18.7 respectively.
 7

M M D L D D

D D
M M L L

M G G
M L L

G G
M M A A

G G
A A

Figure 18.6 A typical process or job shop layout.

 8

Part #1 L L M D G

A A
Receiving L M G G

Part #2

L M D Shipping
Part #3

Figure 18.7 A typical product or flow shop layout.

Process layout is typically employed for a large variety of products that are
made in very small batches (ones or twos). The advantages of Process Layout
are: 1) the flexibility of the system to product almost any part that fits within the
volumemetric boundaries of the machines, 2) an i n depth understanding of a
specific process can be obtained, and 3) some tooling and fixtures can be
shared. The disadvantages of process layout are: 1) the spaghetti flow is
difficult to manage and control, 2) there is usually a lot of inventory in front of
each machine, 3) set up is usually expensive, 4) material handling times are
large, and 5) it is difficult to automate these types of systems.

Product layout systems are used effectively for the economic production of
high volume goods. The advantages of these systems are: 1) large batches can
be produced inexpensively, 2) material handling is minimal, 3) in-process
materials are minimized, 4) it is easy to control these systems, and 5) automation
is more achievable and justifiable. The disadvantages of these systems are: 1)
they are inflexible, in that only one or very few products can be produced on
them, 2) set up time for these systems is very large, and 3) duplicate tooling is
required to replace worn tooling so that maintenance can be minimized.

Process systems work effectively on “one of a kind” type of production. As

batches get larger, these systems fail to produce the required “economies of
scale”, and that production time and cost remains relatively constant. Product
systems work very effectively on single item production. For instance, high
volume products like soda, beer, canned foods, cigarettes, and many automotive
components are effectively produced on these flow systems. The reason that
these items are so inexpensive is in part because of the way they are produced.
Unfortunately, the high capital cost and long set-up for these systems mandates
 9

large volumes to offset these initial costs and then the changeover costs
associated with these systems.

In general, very low volume items should be produced on process type

systems, and very high volume items should be produced on product type
systems. A problem facing most manufacturers is that the general trend today is
for medium volume batches that change regularly. This means tha t process and
product layout fails to meet the requirements for much of what is demanded
today. The result is that a hybrid of the two systems has been developed. It is
called a manufacturing cell. See Figure 18.8 and 18.9. Cells are used to make
families of parts, rather than just one-of-a-kinds or high volume items. Cells are
logical clusters of machines organized to produce a variety of parts requiring the
same equipment type, tooling and fixtures. Cells are intended to provide as
many of the benefits of process and product layouts as possible.

Figure 18.8 A manufacturing cellular layout.

Figure 18.9 A U - shaped manufacturing cell.

In many ways, cellular layout separates and groups products within a

manufacturing system into smaller units. The strategy here is to identify parts
that belong to the same “product families”. A product family is a group of
products that normally look similar and require the same (or similar) processing
 10

steps to produce. Traditionally, designers have formed product families by

grouping products that provide similar functions into a common product family.
Examples might be: springs, clips, brackets, etc. The problem however becomes
one where functional names may create very large families (for instance,
brackets typically form a large portion of products the automotive industries
make) or function may require different processes as the size scales. Group
technology (GT) was introduced as a method to characterize products into
code-able families. A descriptive code was used to characterize the product
geometry, function and/or method used to make the part. As database systems
have become more powerful, the code has been replaced with descriptive fields
in a database. Today, more formal techniques are used to identify cells. In the
following section, a methodology to organize factories, machines, tooling will be
presented.

18.3 Production Flow Analysis

Production flow analysis (PFA) is a technique that is used to organize

machines into a logical grouping. PFA was first introduced by J. L. Burbidge
[Burbidge, 1971] to solve product family formation problems. When applied to a
single factory, the classic framework for manual implementation of PFA consists
of four stages, each stage achieving material flow reduction for a progressively
reducing portion of the factory: Factory Flow Analysis (FFA), Group Analysis
(GA), Line Analysis (LA) and Tooling Analysis (TA).

In FFA (Figure 18.10), dominant material flows between shops (or buildings)
are identified. In addition, if parts are observed to backtrack between any of the
shops, these flows are eliminated by a minor redeployment of equipment. FFA
may often be redundant for a factory that essentially consists of a single machine
or fabrication shop.
 11

MATERIAL

28

16 2
15
1

27 3
53 1 3
1
1
1 24
3 1
126

84 151 3
4

8
1
1 12 26
3 DEPARTMENTS
9 27 1 = BLANKS
5 45 1 2 = SHEET METAL WORK
3 3 = FORGE
1 4 = WELDING DEPT
2 5 = MACHINE SHOP
6 = ASSEMBLY
9 = OUTSIDE FIRMS

FINISHED
6
PRODUCT

MATERIALS 1 MATERIALS 1

5 4 3 2 5 3&4 2

6 6

FINISHED FINISHED
PRODUCT PRODUCT

Figure 18.10 A typical Factory Flow Analysis. (Source: Burbidge, J. L.1971,

April/May. Production Flow Analysis. The Production Engineer, 139-152.)

In GA (Figure 18.11), the flows in each of the shops identified by FFA are
analyzed. GA analyzes operation sequences of the parts being produced in a
particular shop to identify manufacturing cells. Loads are calculated for each part
family to obtain the equipment requirements for each cell. Each cell usually
contains all the equipment necessary to satisfy the complete manufacturing
requirements of its part family. Due to sharing and non-availability of equipment,
some inter-cell material flows and flows to/from vendors may arise.
 12

PART/PRODUCT
K L L M M L M M E E E E E K E E K E M K M M M K E E E K K M E
4 4 4 4 4 4 4 4 4 4 3 1 1 4 4 4 3 3 4 4 4 4 4 3 7 4 3 4 4 6 1
8 8 8 4 7 8 8 4 1 8 4 2 2 7 7 8 4 3 8 4 5 5 8 4 3 6 3 5 3 1 8
2 3 2 2 6 3 1 2 7 5 2 2 2 6 7 5 5 4 2 2 6 6 3 0 9 3 2 1 5 5 6
5 8 6 7 9 8 9 7 9 9 6 0 8 9 8 8 9 9 6 7 9 9 8 9 2 6 9 9 9 9 9
1 8 7 6 3 8 5 6 5 6 7 4 8 7 2 6 6 4 5 6 1 1 6 8 4 5 9 0 2 4
A B E F M C D D C D B H A

M DMT(3) X X X X X X X X X X
A DM(3) X X X X X X X X X X X X X
C
H PG X X X X X X
I DXY(3) X X X X X X X X
N P&GR X
E
/ PGR X X
W PGH
O
R
K PGG X X X
S P&G X X X X X X X X X X X X X
T
A RP X
T PGB X X X X X X
I
W&P X X X X X
O
N WG3 X
COMPONENT – MACHINE CHART. INITIAL RECORD. FORGE.

PART/PRODUCT
L K M E K L E K K K M M M E E E E E M K L M M M K E E E E E M
4 3 4 3 4 4 7 3 4 4 6 4 4 3 1 1 4 4 4 4 4 4 4 4 4 4 4 1 3 4 4
8 4 8 3 4 8 3 4 5 3 1 8 4 4 2 8 1 8 8 8 8 5 5 4 7 7 6 2 3 8 7
2 5 2 4 2 3 9 0 1 5 5 1 2 2 2 6 7 5 3 2 3 6 6 2 6 7 3 2 2 5 6
6 9 6 9 7 8 2 9 9 9 9 9 7 6 0 9 9 9 8 5 8 9 9 7 9 8 6 8 9 8 9
7 6 5 4 6 8 8 9 0 2 5 6 7 4 4 5 6 6 1 8 1 1 6 7 2 4 8 5 6 3
B D C M A C D H A D B E F

PG X X X X X X
GROUP-1

DM 3/1 X X X X
M DXY 3/1 X X
A RP X
C FAMILY - 1
H
I P&G X X X X X X X X X X X X ONE “EXCEPTION” X
N DMT 3/2 X X X X X X
E DM 3/2 X X X X X X
/
GROUP-2

W DXY 3/2 X X X X X X
O W&P X X X X X
R WG3 X
K
S FAMILY - 2
T PGG X X X
GROUP-3

A PGB X X X X X X
T
I PGR X X
O DMT 3/3 X X X X
N DM 3/3 X X X
P&GR X
FAMILY –3
COMPONENT – MACHINE CHART. AFTER FINDING FAMILIES AND GROUPS

Figure 18.12 Group Analysis (Source: Burbidge, J. L.1971, April/May.

Production Flow Analysis. The Production Engineer, 139-152.)
 13

In Line Analysis (LA) (Figure 18.12), a linear or U-layout is designed for

the machines assigned to each cell. The routings of each part assigned to the
cell and the frequency of use of each routing are used to develop a cell for
efficient transport as well as minimum material handling and travel by operators.

MATERIAL

65 7
1

1 2 3 2
HS4 MO HS
1
2 6
11
3 3 2 4

5 7 6 4
MV DH MH DS

1 1
1

8
SA

41 2 5 4 4 16 2

GROUP FLOW NETWORK DIAGRAM - GROUP 2

MATERIALS

72

1
HS4

17 6 4 3

4 2 6 1 7 5
42 1
DH MH DS MV
1 2
8
15
5
8
SA

72

SIMPLIFIED GROUP FLOW NETWORK - GROUP 2

Figure 18.12 An example of Line Analysis. (Source: Burbidge, J. L.1971,

April/May. Production Flow Analysis. The Production Engineer, 139-152.)

In Tooling Analysis (TA) (Table 18.1), the principles of GA and LA are

integrated with data on the shape, size, material, tooling, fixturing, etc. (attributes
of the parts). TA helps to schedule the cell by identifying families of parts with
similar operation sequences, tooling and setups. It seeks to sequence parts on
each machine and to schedule all the machines in the cell to reduce setup times
and batch sizes. This increases available machine capacity on bottleneck work
centers in the cell.
 14

Table 18.1 An example of Tooling Analysis. (Source: Gallagher, C. C. & Knight,

W. A. (1973), Group Technology, London, UK: Butterworths.)

Digit 1 Digit 2 Digit 3 Digit 4 Digit 5 Digit 6 Digit 7 Digit 8

Dimension Matching with
Method of 3 Jaw chuck Surface
Special Quadruple single point Material
holding Bore Over Dw L Boring tool carrier accuracy
attachments tool holder
dia. φ all
3 Jaw rough
0 chuck < 40 L/Dw<0.1 w/o w/o w/o GG-formed 0
turned ∇
outer
3 Jaw Boring, counter- fine turned
41…… Uniform cutting, w/o
1 chuck 42 φ 160 L/Dw<0.5 Axial copying sinking, reaming, ST-formed 1
100 accuracy. ∇∇
inner tapping.
L/Dw up to Uniform cut, or staggered
4 Jaw 101…
2 60 φ 250 limit of Face copying Only outer turning. cut, with accuracy, NE-formed outer fit 2
chuck 200
chuck simple boring up to 48 φ.
Outer shaping,
Spring 301… inner fit (+
3 80 φ 315 Shafts<500 2 Axis copying 1 with 2 chamfering, inserting with GG-cut off 3
collet 400 outer)
form tool, not copying.
Conical Surface Shaping, etc. with
Mandrel or 401… Shafts positional
4 80 φ 400 form tool; with 3; not 3 with 4 ST-cut off 4
arbor 500 500…1000 tapering±12° accuracy
copying.
Inner shaping Shaping, inserting
Jig or 501… Shafts
5 125 φ 500 Steep cone inserting chamfering; chamfering with form NE-cut off polishing 5
fixture 1000 1m…2m
with 3; copying. tool; copying.
Between Shafts Short thread Inner & outer at the knurling,
6 > 1000 5 with 2 & 1 or 3 GG-bar 6
centers 2m…5m milling same time etc.
Chuck- Shafts > Threading with
7 6 with back tool holder ST-bar 7
center 5m lead screw
Thread with
8 Steadies NE-bar 8
copying
Eccentric
Unround Automatic cycle with 4th &
9 (face non-metal 9
copying 5th digits
plate)

Several techniques are available to form clusters for FFA, GA, LA and TA.
These techniques use routing or process planning information in order to identify
families. In the following sections, some of these techniques will be described
and illustrated.

18.3.1 Rank –Order Cluster Algorithm

King (1979) presented a rank -order cluster algorithm that is quite simple. We
use his method to show how component families can be determined in our shop.
The data from Table 18.2 will be used to illustrate this procedure. King’s
algorithm can be stated as follows:

Step 1. For ∀j , calculate the total weight of column w j:

w j = ∑ 2 i M ij
∀j
 15

Figure 18.2 The initial PFA matrix for our example.

Step 2. If wj , is in ascending order, go to step 3. Otherwise, rearrange

the columns to make wj , fall in ascending order.
Step 3. For ∀ j calculate the total weight of row wi :

wj = ∑ 2 j M ij
∀j

Step 4. If wj , is in ascending order, stop. Otherwise, rearrange the rows

to make wi fall in ascending order. Go to step 1.

The rank -order clustering algorithm sorts the matrix into a diagonal block
structure. The diagonal blocks are not always mutually exclusive. Final
judgment has to me made by the user.
One of the major drawbacks of applying this algorithm is the need of
storing the binary word. In order to implement this procedure on a computer, the
word length is max (n, m), where n is the number of machines, and m is the
number of components. For a moderate problem with 50 machines and 2000
components, it is impossible to calculate the weights before sorting. For the
 16

machine weights, 50 words of 2000 bits are needed! A word of 2000 bits
requires 250 bytes. To overcome this problem, direct comparison of elements,
either row or column, can be used. A digit-by-digit comparison is performed,
beginning from the most significant digit. Each row or column of the matrix is
treated as a binary number; no weight is ever calculated. Unfortunately, this
procedure has a computational complexity of a cubic order, namely, O[ij(i + j)]
(King and Nakornchai, 1982), where i and j are the number of rows, and
columns, respectively.
Table 18.3 shows the procedure of rearranging the PFA matrix in from
Table 18.2.

Figure 18.3 Rank-order cluster algorithm tables.

 17

Table 18.3 (Continued)

After we obtain the final matrix, we can determine (although somewhat

arbitrarily) that components A123, A120, A131, A432, A451, and A112 form a
family that needs SAW01, LATHE01, LATHE02, and GRIND05, A115, A212,
A230, and A510 form the second family.

18.3.2 The Direct Clustering Technique

An improvement over the rank -order clustering algorithm was proposed

independently by Chan and Milner (1982) and King and Nakornchai (1982). The
algorithms proposed in both papers are almost identical. The technique is called
direct clustering.
The technique is based on the idea of using blocks and rods, and by
changing the sequence in which components and machines are listed in the
matrix. For convenience, we denote those cells requiring a particular operation
(/) as positive cells, and those with blank entries as negative cells. The direct
cluster algorithm consists of going through the matrix sequentially, and moving
the columns with the topmost negative cells to the left and the rows with the
leftmost positive cells to the top of the matrix sequentially, and moving the
columns with the topmost negative cells to the left and the rows with the leftmost
 18

positive cells to the top of the matrix. In repeated trips, the positive cells move
toward the diagonal of the matrix in a clustered pattern.
The basic rule is that each component or machine must be moved
together with its respective row or column entries during matrix transformation,
as if the cells or the blocks were linked together by imaginary rods. The
algorithmic procedure is as follows.

Step 1. For ∀ i , calculate the total number of positive cells in row wi :

wi = ∑ Mi j
∀j

Sort columns in decending order.

Step 2. For ∀ j, calculate the total number of positive cells in column wj:

wj = ∑ Mij
∀j
Sort columns in ascending order.
Step 3. For i = 1 to n, move all columns j, where M ij = 1, to the right,
maintaining the order of the previous rows.
Step 4. For j = m to 1, move all rows i, where M ij = 1, to the top
maintaining the order of the previous columns.
Step 5. If the current matrix is the same as the previous matrix, stop, or
else go to step 3.

The same PFA matrix in Table 18.2 is rearranged using the direct clustering
algorithm. The procedure and result are shown in Table 18.4. As can be seen
the result in this figure is the same as that obtained by the rank-order clustering
algorithm. However, the direct clustering algorithm does not always work. When
one or several of the machines are bottleneck machines, the iteration stops very
quickly. A bottleneck machine is a machine that is used by a large number of
components and prevents the further rearrangement of columns. The rest of the
matrix is not rearranged because bottleneck machines block them. It happens
especially in large matrices. To solve the bottleneck problem, human
intervention is needed. When the iteration stops prematurely, identify the
bottleneck machines and continue the iteration, disregarding the order of those
rows.
We also may have a final matrix with blocks not mutually exclusive. When
closely investigating the matrix, we may find that only a few cells caused this
problem. Those cells can be considered as exception cells. Mark the exception
cells with an asterisk (*), treat them as negative cells and then reapply the
algorithm. Another technique increases the number of machines of a specific
type and merges machines of the same type. This can be done by adding new
rows and merging two rows respectively. More elaborate examples of the direct
clustering algorithm can be found in Chan and Milner (1982) and King and
Nakornchai (1982).
 19

So far, we have illustrated two algorithms for part-family formation. Many

other methods have been developed, such as the similarity coefficient method,
the set-theoretic method, and the evaluative methods. None of these methods is
as simple as the ones introduced here.

Table 18.4 An example using direct clustering.

The results also vary. Family formation is not always objective; many other
factors such as the cost of implementing the machine cell and material-handling
costs have to be considered. The problem should not be considered as just
clustering cells in an incidence matrix. The methods discussed here provide a
starting point for further improvement.

18.3.3 Mathematical Programming in Group Technology

In solving the group-technology problem, there are two approaches that have
been used extensively: heuristic algorithms and mathematical programming
models. Mathematical programming deals with the optimization of a function
 20

consisting of several variables. In addition, the variables must satisfy a set of

constraints that are inequalities and/or equalities (Zoutendijk, 1976).

Table 18.4 (Continued)

When applying mathematical programming models to group-technology

problems, most models consider a measure of the distance, d i j , between part i
and part j. Some of the measures used most often are the Minkowski distance
measure, the weighted Monkowski distance measure, and the hamming distance
(Kusiak, 1990).
Given the distance between parts, the objective function can be defined as
consisting of a set of variables satisfying the constraints that minimizes the total
sum of distances between any two parts i and j. Two models that have been
applied to the group-technology problem are the p-median model and quadratic
programming (Kusiak, Boe, and Cheng, 1993). Both models are utilized to group
n parts into p families. Unlike the p-median model, however, the quadratic
programming model requires that both the number of part families and the
number of parts in a family be specified beforehand.
There are two general categories of mathematical programming models:
ones that are prescriptive and ones that are descriptive. Linear programming
models are prescriptive because the objective function is predefined and once
the model is “turned on,” the decision variables that optimize the objective
function are determined. On the other and, descriptive models start with the
 21

definition of the decision variables, and during simulation, for example an

estimate of the overall system performance is provided (Askin and Standridge,
1991).
Askin and Standridge (1991) developed a mathematical statement for
assigning part operations and machines to group utilizing binary ordering of the
machine-part matrix to determine independent groups. Boctor (1991) proposed a
mixed-integer linear program to minimize the number of exceptional elements in
the machine-part group formulation problem. Another mathematical approach in
solving the problem of exceptional elements in cellular manufacturing was
presented (Shafer, Kern, and Wei, 1992). A The model dealt with the
minimization of three costs: (1) intercellular transfer, (2) machine duplication, and
(3) subcontracting. Srinivansan, Narendran, and Mahadevan (1990) developed
an improvement over the p-median model in the problem of creating part families
in group technology. The assignment method was found to be superior in the
quality of the solution, as well as computational time. The problem of grouping
parts and tools was solved using 0-1 linear integer problem and the Lagrangian
dual problem (Ventura, Chen, and Wu, 1990).

18.3.4 Production Flow Analysis And Simplification Toolkit

Production flow analysis and simplification toolkit (PFAST) is a technique

introduced in Irani, et al [2001] is presented. Much of the material comes directly
from their paper.

The evaluation and simplification of the material flow network in a facility

prior to undertaking the design of the layout can now be achieved using the
Production Flow Analysis and Simplification Toolkit (PFAST). PFAST is an
experimental library of programs for machine grouping, part family formation, cell
layout and shop layout design developed at The Ohio State University (Daita,
Irani & Kotamraju; Irani & Zhou). It extends the classical framework of Production
Flow Analysis to the analysis of flows at any level of resolution in a facility.
Examples of levels of resolution are: Between buildings, between departments or
shops in the same building, between cells in a shop, between machines in a
manufacturing cell, between locations around a single machining center, etc.

In a typical facility layout study, it is assumed that the layout design will
eliminate inefficient material flow in the facility. Therefore, no effort is made to
simplify the material flow network that is input to the block layout algorithm that
generates the optimal layout. The symptoms shown in Table 18.5 usually
indicate the inefficiency of material flows in a facility that is due to one or more of
the causes listed in Table 18.6.
 22

Table 18.5 Symptoms of Inefficient Material Flows in a Facility

• Large travel distances in the material flow diagram
• Perceived shortage of floorspace for facility expansion
• High WIP inventory levels and inter-machine transfer
batch sizes
• High levels of finished product inventories
• Bottleneck resources with large queues
• Large cycle times for order fulfillment
• Significant queuing and material handling delays
• Significant forklift truck activity
• Poor order tracking capability
• Highly unpatterned material flow network
• Absence of a network of material handling aisles
• Inefficient communications between workcenters
• Low machine and labor utilization
• Absence of scheduling based on resource constraints
• Inflexibility to handle demand and/or part mix changes

Table 18.6 Causes of Inefficient Material Flows in a Facility

• Building architecture
• Locations of manufacturing departments
• Locations of support services and utilities
• Design of the network of material handling aisles
• Locations of input/output points of access between
departments
• Shortage of space for facility expansion
• Process plans for making parts and products
• Variety of routings in the part mix
• Current manufacturing technology
• Current material handling equipment
• Current parts and WIP storage systems
• Current material handling scheduling policies
• Choice of subcontracted operations and/or parts

In addition to layout design, PFAST outputs could help to evaluate other

strategies for material flow simplification, as shown in Table 18.7.
 23

Table 18.7 Utility of Algorithms in PFAST

• Consolidation of buildings and departments
• Strategic duplication of equipment among departments
• Formation of manufacturing cells and focused factories
• Design of a network of material handling aisles
• Modification of process plans and product designs
• Systematic investment in Flexible Manufacturing Cells
• Choice of subcontracted operations and /or parts
• Reduction of variety of routings in the part mix
• Elimination of parts that complicate the flow network
• Enhancement of flexibility in routing products

Since PFAST algorithms use three types of input data – operation sequences,
machine-part matrices and From-To charts – they could be used for a variety of
applications, as listed in Tables 18.8 (a)-(c).

Table 18.8(a) Using PFAST for Material Flow Analysis

• Descriptive statistics for routing data
• Descriptive statistics for material flow network in facility
• Grouping of similar routings
• Detection of redundant variety in routings
• Identification of “misfit” (or outlier) routings
• Analysis of in- house vs. subcontracted material flows
• Elimination or reduction of poorly utilized material flow
paths
• Detection of flow backtracking in routings
• Detection of flow backtracking in material flow network
• Detection of cross flows among aisles in the facility
• Detection of recurrent combinations (or sequences) of
operations in routings
• Evaluation of current vs. desired flexibility of existing
manufacturing equipment
• Creation of alternative routings for key products
 24

Table 18.8 (b) Using PFAST to plan for Cellular Manufacturing

• Rapid assessment of feasibility of conversion to manufacturing cells
• Range for the number of cells that could be introduced
• Composition (machine group and part family) and size of
each cell
• Equipment variety in each cell
• Part mix variety in each cell
• Analysis of homogeneity of part mix and routings
→ Parts that do not belong to any cell
→ Parts whose routings span more than one cell
→ Parts with “exception” operations in their routings
→ Parts that could be produced in more than one cell
• Duplication of equipment required in two or more cells
• Alternative solutions for the cell composition problem
→ For a fixed number of cells
→ For a variable number of cells
• Analysis of the stability of cell compositions
• Giving focus to Integrated Product and Process Design
(IPPD) activities in order to eliminate intercell flows

Table 18.8 (c) Using PFAST to support Facility Layout

• Pareto Analysis of parts using multi-criterion sampling
• Sorting of parts to identify those with identical routings
• Design of a block layout for a factory site, building,
department or shop
• Design of a flowline or U- layout for a cell
• Design of non-traditional layouts
→ Hybrid Cellular Layouts
→ Cascading Cells
→ Modular Layouts
→ Virtual Cellular Layouts
• Strategic duplication of equipment in several shops or
departments
• Strategic consolidation of shops or departments
• Design of a flexible layout using multiple samples of
routings
• Design of a network of material handling aisles

18.3.4.1 Input Data for PFAST

In order to illustrate the capabilities of the current algorithms in PFAST, a 12-

machine, 19-part example from the literature was used, as shown in Table 18.9.
Table 18.10 shows the area of each machine. Based on this data, we can obtain
 25

its machine-part matrix (Table 18.11), asymmetric From-To Chart (Table 18.12)
and symmetric From-To Chart (Table 18.13) for the machines.

Table 18.9 Essential Data for PFAST

PART # OPERATION BATCH
SEQUENCE QUANTITY
1 1,4,8,9 2
2 1,4,7,4,8,7 3
3 1,2,4,7,8,9 1
4 1,4,7,9 3
5 1,6,10,7,9 2
6 6,10,7,8,9 1
7 6,4,8,9 2
8 3,5,2,6,4,8,9 1
9 3,5,6,4,8,9 1
10 4,7,4,8 2
11 6 3
12 11,7,12 1
13 11,12 1
14 11,7,10 3
15 1,7,11,10,11,12 1
16 1,7,11,10,11,12 2
17 11,7,12 1
18 6,7,10 3
19 12 2

Table 18.10 Area of each Machine in the Facility

Machine Area of the Machine
(expressed in # of grid squares)
1 1
2 1
3 1
4 2
5 1
6 2
7 3
8 1
9 1
10 3
11 3
12 1
 26

Table 18.11 Machine-Part Matrix for the Routings in Table 18.9

MACHINE
1 2 3 4 5 6 7 8 9 10 11 12
1 1 1 1 1
2 1 1 1 1
3 1 1 1 1 1 1
4 1 1 1 1
5 1 1 1 1 1
6 1 1 1 1 1
7 1 1 1 1
P 8 1 1 1 1 1 1 1
A 9 1 1 1 1 1 1
10 1 1 1
R 11 1
T 12 1 1 1
13 1 1
14 1 1 1
15 1 1 1 1 1
16 1 1 1 1 1
17 1 1 1
18 1 1 1
19 1

Table 18.12 Asymmetric From-To Chart for the Routings in Table 18.9
Machine 1 2 3 4 5 6 7 8 9 10 11 12
1 1 8 2 3
2 1 1
3 2
4 9 11
5 1 1
6 4 3 3
7 5 2 5 6 3 2
8 3 8
9
10 3 3
11 5 3 4
12
 27

Table 18.13 Symmetric From-To Chart for the Routings in Table 18.9
TO
1 2 3 4 5 6 7 8 9 10 11 12
1 1 8 2 3
F 2 1 1 1
R 3 2
O 4 4 14 11
M 5 1
6 3 3
7 5 5 9 8 2
8 8
9
10 6
11 4
12

18.3.3.2 Using PFAST for Material Flow Simplification and Facility Layout

The algorithms in PFAST use three types of input data  operation sequences,
machine-part matrix and From-To Chart  instead of only From-To charts. This
makes it possible to use the outputs from two or more programs in a synergistic
fashion. In this section, some examples of these synergistic uses are presented
to illustrate the possibility of simplifying the material flow network in a facility prior
to designing a physical layout for it.

Prevention of redundant machine duplication in a cellular layout: Tables

18.14 and 18.15 suggest that, to implement a cellular layout with two cells,
machine 1 would have to be duplicated in each of them. However, the design
skeleton in Figure 14(a), in conjunction with the routings in Table 18.9 and Table
18.11, indicate that machine 1 is always the first machine in all the operation
sequences. Therefore, it may be possible to design a hybrid layout for the facility
where all machines of type 1 are retained in a functional group.
 28

Table 18.14 Final Machine-Part Matrix obtained using Algorithm PG-TSP

MACHINE
5 2 3 6 9 8 4 1 7 10 11 12
11 1
8 1 1 1 1 1 1 1
9 1 1 1 1 1 1
7 1 1 1 1
1 1 1 1 1
3 1 1 1 1 1 1
P 10 1 1 1
A 2 1 1 1 1
R 4 1 1 1 1
T
5 1 1 1 1 1
6 1 1 1 1 1
18 1 1 1
14 1 1 1
16 1 1 1 1 1
15 1 1 1 1 1
12 1 1 1
17 1 1 1
13 1 1
19 1

Table 18.15 Final Machine-Part Matrix obtained using Algorithm PG-QAP

MACHINE
5 3 2 4 8 9 6 1 7 10 11 12
11 1
8 1 1 1 1 1 1 1
9 1 1 1 1 1 1
7 1 1 1 1
1 1 1 1 1
3 1 1 1 1 1 1
10 1 1 1
P 2 1 1 1 1
A 4 1 1 1 1
6 1 1 1 1 1
R 5 1 1 1 1 1
T 18 1 1 1
14 1 1 1
16 1 1 1 1 1
15 1 1 1 1 1
12 1 1 1
17 1 1 1
13 1 1
19 1

Purchase of multi-function flexible equipment to combine steps in the

operation sequences of a family of parts: Table 18.16 shows that each of the
substrings of operations  4→8, 8→9 and 4→8→9  has a high frequency of
occurrence in the origina l routings (Table 18.9). Subject to equipment design
constraints, this string matching result could be used to replace the three single-
function machines  4, 8 and 9  by a multi-function machining center. This
 29

planned introduction of flexible automation would eliminate inter-operation

material handling, loading/unloading and setups for the following parts: 1, 2, 3, 6,
7, 8, 9 and 10.

Table 18.16 Distinct Common_Substrings in the Routings in Table 18.9

No. Common_Substring Frequency of % Occurrence
Occurrence
S1 1→4 3 15.79
S2 1→4→7 2 10.53
S3 1 → 7 → 11 → 10 → 11 2 10.53
→ 12
S4 3→5 2 10.53
S5 4→7 4 21.05
S6 4→7→4→8 2 10.53
S7 4→8 6 31.58
S8 4→8→9 4 21.05
S9 6→4→8→9 3 15.79
S10 6 → 10 → 7 2 10.53
S11 7→8→9 2 10.53
S12 7→ 9 2 10.53
S13 7 → 10 2 10.53
S14 8→9 6 31.58
S15 11 → 7 3 15.79
S16 11 → 7 →12 2 10.53
S17 11 → 12 3 15.79

Design of a Next Generation Layout (NGL): Traditionally, the Cluster Analyses

of Figures 18.10-18.13, in conjunction with Tables 18.10 and 18.11, would be
used to implement a 2-cell or 3-cell layout. Instead, using the results in Tables
18.12 and 18.13, the Next Generation Layout (NGL) in Figure 18.14 could be
implemented. This layout is a combination of a cell (M2), three layout modules
(or multi-process departments)  (M1, M3, M4)  and a single machine (#2)
arranged in a network layout.
 30

Distance
0.90
Threshold for
Cluster Formation

0.60

0.30

0.00
2 10 1 3 4 7 8 9 5 6 18 11 12 17 13 14 15 16 19
Parts
Exceptional parts

Figure 18.13 Formation of Part Families using Algorithm HC-SC

Figure 18.14 Formation of Part Families using Algorithm NHC-KM

 31

Distance
0.76 Threshold for
Cluster Formation

0.51

0.25

0.00
1 7 4 3 5 6 2 10 8 9 12 17 14 18 11 13 19 15 16
Parts

Figure 18.15 Formation of Part Families using Algorithm HC-LD

Table 18.17 Layout Modules obtained by Cluster Analysis of the

Common_Substrings in Table 18.16
Module # Cluster of Layout Module created
Common_Substrings from the Cluster
M1 S1, S2, S5, S6, S7, S8, 1
S9, S11, S14
6 4 8 9

7
M2 S3, S15, S16, S17 1 7 11 10

12
M3 S4 3→ 5
M4 S10, S12, S13 6 10 7 9
 32

M4 2 M1 1 M3

3 1 M2

Figure 18.16 Next Generation Layout (NGL) for the Facility described in
Table 18.9

Synthesis of multiple design skeletons and flow directions during the

design of a block layout for a facility: The design skeletons in Figures
18.17(a) and 18.18(a) would suggest to the facility planner to locate machines 10
and 11 across from each other in the spine layouts of Figures 18.18(b) and
18.18(b). However, Figure 18.17 suggests that machines 7, 10 and 11 are best
located in a cell on one side of the aisle. Or, those machines could be replaced
by a single multi-function machining center to eliminate the delays that occur
when consecutive operations are done on separate machines, or in separate
departments.

R
3

1 3 5 2

5
4 6 2
8
4

7 7 9
8

Aisle

11 10
9 11 10

12
12

Figure 18.17(a) Design Skeleton obtained by Algorithm MSA Figure 18.17(b) Final Layout based on Design Skeleton in Figure
18.17(a)
 33

Aisle

4 8 9 10

11

7
6
7 12

4 9
1 8

11 6
10

5 3

3 2
12
2

Figure 18.18(a) Design Skeleton obtained by Algorithm MST Figure 18.18(b) Final Layout based on
Design Skeleton in Figure 18.18(a)

4 10

8 11

Figure 18.19 Machine Group identified by Algorithm SC-ATC

Incorporating part family formation based on similarity of routings into the

classical P-Q Analysis: Figure 18.20 shows the typical sorting result that is
obtained when a facility planner uses P-Q Analysis to select a sample of parts on
which to base the design of a facility layout. As shown in Table 18.118, this
sampling approach would have excluded all parts after the Cumulative Quantity
reached 28. However, if the clustering results in Figure 18.15 are analyzed, this
sampling approach based on demand volume alone is seen to exclude highly
similar parts  (8,9), (12,17) and (15,16). These are parts whose similar
operation sequences could have been the basis for creation of flowline cells as a
portion of the overall layout.
 34

35

30

Cumulative Production Quantity

25

20

15

10

0
2 4 11 14 18 1 5 7 10 16 19 3 6 8 9 12 13 15 17

Routing #
Figure 18.20 Sorting of Routings using Algorithm SORT-PQ

Table 18.18 Product Routings in Table 18.12 sorted by Production Quantity

Routing # Production Cumulative
Quantity Production Quantity
2 3 3
4 3 6
11 3 9
14 3 12
18 3 15
1 2 17
5 2 19
7 2 21
10 2 23
16 2 25
19 2 27
3 1 28 r
6 1 29
8 1 30
9 1 31
12 1 32
13 1 33
15 1 34
17 1 35

Controlling the complexity of the product mix in custom manufacturing

facilities: In custom manufacturing and Make-To-Order (MTO) facilities, it is
important to base the design of the facility layout on a sample of parts (or
products) tha t excludes “outlier” routings i.e. routings that are highly dissimilar
from the other routings. If the cluster agglomeration levels in Figure 18.15 are
studied, then the pairs of parts  (8,9) and (15,16)  are seen to merge with the
two dominant clusters at low values of similarity. The dissimilarities between
these routings and those of the parts already in the clusters is confirmed by
Table 18.9. The routings of parts 8 and 9 are the only routings to use machines 3
 35

and 5. And, in the cluster of parts 12~17, parts 15 and 16 are the only ones that
require machine 1 for their first operations, whereas, machine 1 is used by all the
parts (1~5) in the other cluster.
From the above example, the reader should see that PFA is a highly
interactive process even using analytical and computer tools. PFA can be
effectively used to form part families and manufacturing calls; however, the
process is not automatic.

18. 4 Controlling a Lean Production Facility

The foundation of Lean manufacturing is to create efficient methods for

producing goods. Eliminating waste, minimizing inconsistencies and taking a
reasonable approach to producing products and managing facilities is the key to
lean manufacturing. The largest portion of time a part spends in a manufacturing
facility is spent waiting rather than having value added. The result is that Lean
Manufacturing has become more than just eliminating waste. A new key has
become reducing lot sizes and applying better production control methods. In
this section, we will introduce several production control techniques that have
been used as part of lean developments.

18.4.1 An analogy: Marching soldiers

In Eliyahu Goldratt's book ``The Goal''[Goldratt, 1996], the importance of a

bottleneck in a factory is described through an analogy to a troop of boy scouts
out for a march. One of the scouts, who is carrying an extra-heavy backpack,
walks more slowly than the rest, so a gap keeps opening between him and the
scouts in front. This is then connected to how inventory masses up in front of a
slow machine in the factory.

But this is less than half the story. In a column of marching soldiers, the problem
is not a slow marcher falling behind. Each soldier carries the same weight, so the
line is balanced, and there is no pronounced bottleneck. The problem is
variability amplification: If the
first soldier for some reason speeds up a little bit, the second soldier will see a
gap open in front of him, and take this as a signal to speed up, as well. But he
will have to speed up more than the first soldier did, in order to catch up with him.
When he has caught up, he
then needs to slow down again to avoid bumping into the one in front.

Now the third soldier sees a gap opening up even faster than the second one did,
so he has to speed up by even more, and has to slow down more abruptly when
he has closed the gap. This way, the small change in speed amplifies down the
line like a whiplash, and the poor guy at the end of the line will alternate between
running flat out and marching in place.
 36

This is what occurs in a kanban line. The last machine in the line tries to track the
demand process, but adds some noise to it due to process variability. The
second last machine tries to track the input process of the last machine, but adds
some more noise. This amplifies the noise upstream, so the first machine in the
line will alternate between working at capacity and waiting for something to be
taken out of its output buffer. To get rid of the problem, one has to eliminate all
process variability, such as machine failures and operation time variability. This
can be time-consuming and expensive.

How do soldiers counteract this age-old problem? If the soldiers are recruits,
they get the attention of a very loud drill sergeant that yells out the cadence.
More seasoned soldiers will be singing a marching song as they go along, and
any infantry outfit has a large supply of these songs. Both of these techniques
have the effect of distributing the proper cadence to every soldier in the line,
simultaneously.

This is what the CONWIP control does. It passes the demand information,
without any noise, to the first machine in the line. All downstream machines know
that any part arriving in their input buffer can be worked on, so they hear the
signal, too.

But marching soldiers do not close their eyes and march blindly. Even if they
receive the proper cadence, they will still be watching the distance to the marcher
in front. If the gap widens, they will take longer strides, and if it narrows, they will
shorten their steps. This
way, the marchers act on two types of information at once: The global
information flow that determines the overall speed, and the local information that
is used for minor adjustments.

This is also the way our hybrid policy works: The CONWIP control gives a global
information flow (like the drill sergeant), and the kanban control gives a local flow
of information (like watching the distance to the guy in front). In our hybrid policy,
the global information flow from the demand process is supplemented by the
local information from the buffer levels. This attains the advantages of CONWIP
control, while using the strengths of kanban control to cancel its disadvantages.

18.5 The Five Steps of Lean Implementation

The process used to implement lean manufacturing is a straightforward one.

However it is critical that lean is implemented in a logical manner. The steps
associated in implementing lean follow:

Step 1: Specify Value

 37

Define value from the perspective of the final customer. Express

value in terms of a specific product, which meets the customer's needs at a
specific price and at a specific time.

Step 2: Map
Identify the value stream, the set of all specific actions required to
bring a specific product through the three critical management tasks of any
business: the problem-solving task, the information management task, and the
physical transformation task. Create a map of the Current State and the Future
State of the value stream. Identify and categorize waste in the Current State, and
eliminate it!

Step 3: Flow
Make the remaining steps in the value stream flow. Eliminate
functional barriers and develop a product-focused organization that dramatically
improves lead-time.

Step 4: Pull
Let the customer pull products as needed, eliminating the need for
a sales forecast.

Step 5: Perfection
There is no end to the process of reducing effort, time, space,
cost, and mistakes. Return to the first step and begin the next lean
transformation, offering a product that is ever more nearly what the customer
wants.

18.6 References

Arvindh, B. and Irani, S. A. (1994). Cell formation: The need for an integrated
solution of the subproblems. International Journal of Production Research, 32(5),
1197-1218.

Chen, C.Y. and Irani, S.A. (1993). Cluster first-sequence last heuristics for
generating block diagonal forms for a machine-part matrix. International Journal
of Production Research, 31(11), 2623-2647.

Burbidge, J. L. (1971), “Production Flow Analysis”, Production Engineer, April

1971, 31-40

Burbidge, J. L. (1975), The Introduction of Group Technology, John Wiley,

New York.
 38

Daita, S.T., Irani, S.A. and Kotamraju, S. Algorithms for production flow analysis.
To appear in International Journal of Production Research.

Gallagher, C. C. & Knight, W. A. (1973), Group Technology, London, UK: Butterworths.

Irani, S.A., Huang. H. and Udai, T. K. Custom design of facility layouts for multi-
product facilities using layout modules. Submitted to IEEE Transaction on
Robotics and Automation: Special Issue on Automation of Manufacturing
Systems Design.

Irani, S.A. and Udai, T. K. Facility layout using directed flow paths. Submitted to
IIE Transactions.

Irani, S.A. and Zhou, J. Production Flow Analysis. To appear in Industrial

Engineering Applications and Practice: A Users’ Encyclopaedia, Editors: A. K.
Mital and J.G. Chen, Published by the International Journal of Industrial
Engineering: Theory, Applications and Practice, University of Cincinnati,
Cincinnati, OH.

Irani, S.A. and Huang, H. (1998). Layout Modules: A novel extension of hybrid
cellular layouts. Proceedings of the 1998 International Mechanical Engineering
Congress and Exposition, Winter Annual Meeting of the ASME, November 15-20,
Anaheim, CA.

Irani, S.A. and Ramakrishnan, R. (1995). Production Flow Analysis using

STORM. In Planning, Design and Analysis of Cellular Manufacturing Systems.
Editors: A.K. Kamrani, H.R. Parsaei and D.H. Liles, Elsevier Science BV: New
York, NY.

Irani, S.A., Cavalier, T.M. and Cohen P. H. (1993). Virtual manufacturing cells:
Exploiting layout design and intercell flows for the machine sharing problem,
International Journal of Production Research, 31(4), 791-810.

Irani, S.A., Cohen P. H. and Cavalier, T.M. (1992). Design of cellular

manufacturing systems. Journal of Engineering for Industry, 114(3), 352-361.

King, J.R. (1980). Machine-component group formation in production flow

analysis: An approach using a rank order clustering algorithm. International
Journal of Production Research, 18(2), 213-232.

Minitab (Rel. 11) Reference Manual (1996, June). State College, PA: Minitab Inc.
 39

Glossary of terms

Andon lights/boards - a system of flashing lights used to indicate

production status in one or more work centers; the number of lights and their
possible colors can vary, even by work center within a plant; however, the
traditional colors and their meanings are:

green - no problems
yellow - situation requires attention
red - production stopped; attention urgently needed

autonomation - in Toyota parlance, automation with a human touch.

Autonomation normally referes to semi-automatic processes where a machine
and human work as a well planned system. Literally, the English translation of
jidoka.

baka-yoke - a manufacturing technique of preventing mistakes (error-

proofing) by designing the manufacturing process, equipment, and tools so that
an operation literally cannot be performed incorrectly; an attempt to perform
incorrectly, as well as being prevented, is usually met with a warning signal of
some sort; the term poka-yoke is sometimes referred to as a system where only
a warning is provided.

Balanced production – a system where the operations for various

machines are approximately the same. A well-balanced system has a takt time
only slightly larger than the operation time.

cellular manufacturing - an approach in which manufacturing work

centers [cells] have the total capabilities needed to produce an item or group of
similar items; contrasts to setting up work centers on the basis of similar
equipment or capabilities, in which case items must move among multiple work
centers before they are completed; the term group technology is sometimes used
to distinguish cells that produce a relatively large family [group] of similar items.

cycle time - the normal time to complete an operation on a product. This

in NOT the same as takt time, which is the allowable time to produce one product
at the rate customers are demanding it.

error-proofing - a manufacturing technique of preventing mistakes (baka-

yoke) by designing the manufacturing process, equipment, and tools so that an
operation literally cannot be performed incorrectly; an attempt to perform
incorrectly, as well as being prevented, is usually met with a warning signal of
some sort; the term poka-yoke is sometimes referred to as a system where only
a warning is provided.
 40

flexible manufacturing system - an integrated manufacturing capability

to produce small numbers of a great variety of items at low unit cost; an FMS is
also characterized by low changeover time and rapid response time.

flow manufacturing – a manufacturing methodology that pulls items from

suppliers through a synchronized manufacturing process to the end product. The
principle goal is a faster response to customer demand.

heijunka - A production scheduling/leveling tool, essentially to distribute

kanban cards in an efficient manner.

Hoshin Kanri – a strategic planning approach that integrates the

practices of leadership with those of management.

jidoka - a Japanese word which translates as autonomation; a form of

automation in which machinery automatically inspects each item after producing
it, ceasing production and notifying humans if a defect is detected; Toyota
expands the meaning of jidoka to include the responsibility of all workers to
function similarly, i.e. to check every item produced and to make no more if a
defect is detected, until the cause of the defect has been identified and corrected.

jishu kanri - self-management, or voluntary participation.

just-in-time - a production scheduling concept that calls for any item

needed at a production operation - whether raw material, finished item, or
anything in between, to be produced and available precisely when needed,
neither a moment earlier nor a moment later.

jutsu - to talk, or ‘the art of’ (i.e., 'leanjutsu: the art of lean production').

kaikaku - A rapid and radical change process, sometimes used as a

precursor to kaizen activities.

kaizen - the philosophy of continual improvement, that every process

can and should be continually evaluated and improved in terms of time required,
resources used, resultant quality, and other aspects relevant to the process.

kanban - a card or sheet used to authorize production or movement of

an item; when fully implemented, kanban (the plural is the same as the singular)
operates according to the following rules:

1.All production and movement of parts and material take place only as
required by a downstream operation, i.e. all manufacturing and
procurement are ultimately driven by the requirements of final
assembly or the equivalent.
 41

2.The specific tool which authorizes production or movement is called

a
kanban. The word literally means card or sign, but it can legitimately
refer to a container or other authorizing device. Kanban have various
formats and content as appropriate for their usage; for example, a
kanban for a vendor is different than a kanban for an internal
machining operation.

3.The quantity authorized per individual kanban is minimal, ideally one.

The number of circulating or available kanban for an item is
determined by the demand rate for the item and the time required to
produce or acquire more. This number generally is established and
remains unchanged unless demand or other circumstances are
altered dramatically; in this way inventory is kept under control while
production is forced to keep pace with shipment volume. A routine
exception to this rule is that managers and workers are continually
exhorted to improve their processes and thereby reduce the number
of kanban required.

karoshi - death from overwork.

lean manufacturing or lean production - the philosophy of continually

reducing waste in all areas and in all forms; an English phrase coined to
summarize Japanese manufacturing techniques (specifically, the Toyota
Production System).

line balancing - equalizing cycle times [productive capacity, assuming

100% capacity utilization] for relatively small units of the manufacturing process,
through proper assignment of workers and machines; ensures smooth production
flow.

mixed-model production - capability to produce a variety of models,

that in fact differ in labor and material content, on the same production line;
allows for efficient utilization of resources while providing rapid response to
marketplace demands.

mizusumashi - the classic 'water spider', who performs a wide range of

tasks which allow workers to perform 'value -added' tasks.

mokeru - the Japanese term for the industrial engineering, more

properly translated as ‘profit-making I.E.’.

muda (waste) - activities and results to be eliminated; within

manufacturing, categories of waste, according to Shigeo Shingo, include:
 42

1.Excess production and early production

2.Delays
3.Movement and transport
4.Poor process design
5.Inventory
6.Inefficient performance of a process
7.Making defective items

mura - inconsistency

muri - unreasonablness

nagara - smooth production flow, ideally one piece at a time,

characterized by synchronization [balancing] of production processes and
maximum utilization of available time, including overlapping of operations where
practical.

ninjutsu - the art of invisibility (applies to management)

one piece flow – producing one unit at a time, as opposed to producing

large batches.

poka-yoke - a means of providing a visual or other signal as to the

location or condition of a part characteristic. Often referred to as 'error-proofing',
poke-yoke is actually the first step in truly error-proofing a system (see baka-
yoke).

pull system - a manufacturing planning system based on communication

of actual real-time needs from downstream operations ultimately final assembly
or the equivalent - as opposed to a push system which schedules upstream
operations according to theoretical downstream results based on a plan which
may not be current.

5S’s - refers to the five Japanese words seiri, seiton, seison, seiketsu,
shitsuke. These words are shorthand expressions for principles of maintaining
an effective, efficient workplace.

seiri - eliminating everything not required for the work being

performed
seiton - efficient placement and arrangement of equipment and
material
seison - tidiness and cleanliness
seiketsu - ongoing, standardized, continually improving seiri,
seiton, seison
shitsuke - discipline with leadership
 43

6 Sigma – a structured process improvement program for achieving

virtually zero defects in manufacturing and business.

seiban - Seiban is the name of a Japanese management practice taken

from the Japanese words "sei", which means manufacturing, and "ban", which
means number. A Seiban number is assigned to all parts, materials, and
purchase orders associated with a particular customer job, or with a project, or
anything else. This enables a manufacturer to track everything related with a
particular product, project, or customer. It also facilitates setting aside inventory
for specific projects or priorities. That makes it great for project and build-to-
order manufacturing.

sensei - one who provides information; a teacher, instructor, or rabbi.

setup time - work required to change over a machine or process from one
item or operation to the next item or operation; can be divided into two types:

1.internal: setup work that can be done only when the machine or
process is not actively engaged in production; OR
2.external: setup work that can be done concurrently with the
machine or process performing production duties.

shojinka - continually optimizing the number of workers in a work center

to meet the type and volume of demand imposed on the work center; shojinka
requires workers trained in multiple disciplines; work center layout, such as U-
shaped or circular, that supports a variable number of workers performing the
tasks in the layout; the capability to vary the manufacturing process as
appropriate to fit the demand profile.

SMED - abbreviation for Single Minute Exchange of Die; literally, changing

a die on a forming or stamping machine in a minute or less; broadly, the ability to
perform any setup activity in a minute or less of machine or process downtime;
the key to doing this is frequently the capability to convert internal setup time to
external setup time; variations on SMED include:

1.Single-digit setup: performing a setup activity in a single-digit

number of minutes, i.e. fewer than ten.
2.OTED: One touch exchange of die; literally, changing a die with one
physical motion such as pushing a button; broadly, an extremely
simple procedure for performing a setup activity.

standard operations – clearly defined operations and standardized steps

for both workers and machines.

takt time - takt, is a German term for rhythm or pace. Takt time is the
allowable time to produce one product at the rate a customer demands it. This is
 44

NOT the same as cycle time, which is the normal time to complete an operation
on a product (which should be less than or equal to takt time).

teian - a proposal, proposition, or suggestion. A teian system can be

likened to a system that allows and encourages workers to actively propose
process and product improvements.

Toyota - changed from the true form, Toyoda, meaning abundant rice
field, by the Toyota marketing department. Toyoda is the family name of the
founders of the Toyota Motor Company.

Value Stream Mapping (VSM) – a process to determine the value added

to a product as it goes through a manufacturing system.

water spider - one who performs a wide range of tasks which allow
workers to perform 'value -added' tasks.

WCM - world class manufacturing is the philosophy of being the best, the
fastest, and the lowest cost producer of a product or service. It implies the
constant improvement of products , processes, and services to remain an
industry leader and provide the best choice for customers, regardless of where
they are in the process.

Outline:
§ Introduction to Lean Manufacturing
§ Set-up reduction
§ Group technology
§ Control of lean systems
§ Product examples and illustrations
§ Chapter summary

Review Questions
Review Problems
References/Bibliography