Final Report

Upload
chandrashekhar 
Category
Documents

view
50 
download
0
Embed Size (px)
Transcript of Final Report
MINOR PROJECT
A Study onSix Sigma Techniques
AndIts application in reduction of seat rejection
At BOSCH LTD.
Submitted bySuyog Gholap(107269)
R.Rahul(107254)Chandra shekhar.L(107266)
Sudip Pal(107237)K.Seshi Kiran Reddy(107249)
1
Introduction to Six Sigma:
Sigma (σ) is a letter in the Greek alphabet that has become the statistical symbol and metric of
process variation. The sigma scale of measure is perfectly correlated to such characteristics as
defectsperunit, partspermillion defectives, and the probability of a failure. Six is the number of
sigma measured in a process, when the variation around the target is such that only 3.4 outputs out of
one million are defects under the assumption that the process average may drift over the long term by
as much as 1.5 standard deviations. Six sigma may be defined in several ways. Tomkins defines Six
Sigma to be “a program aimed at the nearelimination of defects from every product, process and
transaction.” Harry (1998) defines Six Sigma to be “a strategic initiative to boost profitability, increase
market share and improve customer satisfaction through statistical tools that can lead to breakthrough
quantum gains in quality.”
Six sigma was launched by Motorola in 1987. It was the result of a series of changes in the quality
area starting in the late 1970s, with ambitious tenfold improvement drives. The toplevel management
along with CEO Robert Galvin developed a concept called Six Sigma. After some internal pilot
implementations, Galvin, in 1987, formulated the goal of “achieving SixSigma capability by 1992” in a
memo to all Motorola employees. The results in terms of reduction in process variation were ontrack
and cost savings totaled US$13 billion and improvement in labor productivity achieved 204% increase
over the period 1987–1997.In the wake of successes at Motorola, some leading electronic companies
such as IBM, DEC, and Texas Instruments launched Six Sigma initiatives in early 1990s. However, it
was not until 1995 when GE and Allied Signal launched Six Sigma as strategic initiatives that a rapid
dissemination took place in nonelectronic industries all over the world. In early 1997, the Samsung
and LG Groups in Korea began to introduce Six Sigma within their companies. The results were
amazingly good in those companies. For instance, Samsung SDI, which is a company under the
Samsung Group, reported that the cost savings by Six Sigma projects totaled US$150 million. At the
present time, the number of large companies applying Six Sigma in Korea is growing exponentially,
with a strong vertical deployment into many small and mediumsize enterprises as well. Six sigma
tells us how good our products, services and processes really are through statistical measurement of
quality level. It is a new management strategy under leadership of toplevel management to create
quality innovation and total customer satisfaction. It is also a quality culture. It provides a means of
doing things right the first time and to work smarter by using data information. It also provides an
atmosphere for solving many CTQ (criticaltoquality) problems through team efforts. CTQ could be a
critical process/product result characteristic to quality, or a critical reason to quality characteristic.
Defect rate, PPM and DPMO:
The defect rate, denoted by p, is the ratio of the number of defective items which are out of
specification to the total number of items processed (or inspected). Defect rate or fraction of defective
items has been used in industry for a long time. The number of defective items out of one million
inspected items is called the ppm (partspermillion) defect rate. Sometimes a ppm defect rate cannot
be properly used, in particular, in the cases of service work. In this case, a DPMO (defects per million
2
opportunities) is often used. DPMO is the number of defective opportunities which do not meet the
required specification out of one million possible opportunities.
Standard Deviation:
In probability theory and statistics, standard deviation is a measure of the variability or dispersion of
a population, a data set, or a probability distribution. A low standard deviation indicates that the data
points tend to be very close to the same value (the mean), while high standard deviation indicates that
the data are “spread out” over a large range of values.
For example, the average height for adult men in the United States is about 70 inches, with a
standard deviation of around 3 inches. This means that most men (about 68%, assuming a normal
distribution) have a height within 3 inches of the mean (67 inches – 73 inches), while almost all men
(about 95%) have a height within 6 inches of the mean (64 inches – 76 inches). If the standard
deviation were zero, then all men would be exactly 70 inches high. If the standard deviation were 20
inches, then men would have much more variable heights, with a typical range of about 50 to 90
inches.
Fig: A data set with a mean of 50 (shown in blue) and a standard deviation (σ) of 20.
Fig: A plot of a normal distribution (or bell curve).
Each colored band has a width of one standard deviation.
In addition to expressing the variability of a population, standard deviation is commonly used to
measure confidence in statistical conclusions. For example, the margin of error in polling data is
3
determined by calculating the expected standard deviation in the results if the same poll were to be
conducted multiple times. (Typically the reported margin of error is about twice the standard deviation,
the radius of a 95% confidence interval.) In science, researchers commonly report the standard
deviation of experimental data, and only effects that fall far outside the range of standard deviation are
considered statistically significant. Standard deviation is also important in finance, where the standard
deviation on the rate of return on an investment is a measure of the risk.
Consider a population consisting of the following values
There are eight data points in total, with a mean (or average) value of 5:
To calculate the standard deviation, we compute the difference of each data point from the mean,
and square the result:
Next we average these values and take the square root, which gives the standard deviation:
Therefore, the population above has a standard deviation of 2.
Note that we are assuming that we are dealing with a complete population. If our 8 values are
obtained by random sampling from some parent population, we might prefer to compute the sample
standard deviation using a denominator of 7 instead of 8.
The standard deviation of a discrete random variable is the rootmeansquare (RMS) deviation of its
values from the mean.
If the random variable X takes on N values (which are real numbers) with equal
probability, then its standard deviation σ can be calculated as follows:
1. Find the mean, , of the values.
2. For each value xi calculate its deviation ( ) from the mean.
3. Calculate the squares of these deviations.
4. Find the mean of the squared deviations. This quantity is the variance σ2.
5. Take the square root of the variance.
This calculation is described by the following formula:
4
Where is the arithmetic mean of the values xi, defined as:
If not all values have equal probability, but the probability of value xi equals pi, the standard deviation
can be computed by:
and
Where
And N' is the number of nonzero weight elements.
For example
Suppose we wished to find the standard deviation of the data set consisting of the values 3, 7, 7, and
19.
Step 1: find the arithmetic mean (average) of 3, 7, 7, and 19,
Step 2: find the deviation of each number from the mean,
Step 3: square each of the deviations, which amplifies large deviations and makes negative values
positive,
5
Step 4: find the mean of those squared deviations,
Step 5: take the nonnegative square root of the quotient (converting squared units back to regular
units),
So, the standard deviation of the set is 6. This example also shows that, in general, the standard
deviation is different from the mean absolute deviation (which is 5 in this example).Note that if the
above data set represented only a sample from a greater population, a modified standard deviation
would be calculated to estimate the population standard deviation, which would give 6.93 for this
example.
Rules for normally distributed data
Dark blue is less than one standard deviation from the mean. For the normal distribution, this
accounts for 68.27 % of the set; while two standard deviations from the mean (medium and dark blue)
account for 95.45%; three standard deviations (light, medium, and dark blue) account for 99.73%; and
four standard deviations account for 99.994%. The two points of the curve which are one standard
deviation from the mean are also the inflection points.
The central limit theorem says that the distribution of a sum of many independent, identically
distributed random variables tends towards the normal distribution. If a data distribution is
approximately normal then about 68% of the values are within 1 standard deviation of the mean
(mathematically, μ ± σ, where μ is the arithmetic mean), about 95% of the values are within two
standard deviations (μ ± 2σ), and about 99.7% lie within 3 standard deviations (μ ± 3σ). This is known
as the 689599.7 rule, or the empirical rule.
For various values of z, the percentage of values expected to lie in the symmetric confidence
interval (−zσ,zσ) are as follows:
zσ percentage
1σ 68.2689492%
1.645σ 90%
6
1.960σ 95%
2σ 95.4499736%
2.576σ 99%
3σ 99.7300204%
3.2906σ 99.9%
4σ 99.993666%
5σ 99.9999426697%
6σ 99.9999998027%
7σ 99.9999999997440%
In Fig: Dark blue is less than one standard deviation from the mean. For the normal distribution, this
accounts for 68.27 % of the set; while two standard deviations from the mean (medium and dark blue)
account for 95.45%; three standard deviations (light, medium, and dark blue) account for 99.73%; and
four standard deviations account for 99.994%. The two points of the curve which are one standard
deviation from the mean are also the inflection points.
The standard deviation of a data set is the same as that of a discrete random variable that can
assume precisely the values from the data set, where the point mass for each value is proportional to
its multiplicity in the data set.
The term "standard deviation" was first used in writing by Karl Pearson in 1894 following use by him in
lectures. This was as a replacement for earlier alternative names for the same idea: for
example Gauss used "mean error". A useful property of standard deviation is that, unlike variance, it
is expressed in the same units as the data.
Sigma quality level
Specification limits are the tolerances or performance ranges that customer's demand of the products
or processes they are purchasing. Figure 1.8 illustrates specification limits as the two major vertical
7
lines in the figure. In the figure, LSL means the lower specification limit, USL means the upper
specification limit and T means the target value. The sigma quality level (in short, sigma level) is the
distance from the process mean (μ) to the closer specification limit. In practice, we desire that the
process mean to be kept at the target value. However, the process mean during one time period is
usually different from that of another time period for various reasons. This means that the process
mean constantly shifts around the target value. To address typical maximum shifts of the process
mean, Motorola added the shift value ±1.5 s to the process mean. This shift of the mean is used when
computing a process sigma level. From this figure, we note that a 6 sigma quality level corresponds to
a 3.4ppm rate.
Fig: Sigma quality levels of 6σ and 3σSigma level for discrete data:
Suppose two products out of 100 products have a quality characteristic which is outside of
specification limits. Then in one million parts 20,000 parts will be defects so, sigma level will be
between 3 & 4.Preciously it will come as 3.51σ. The broad classification of sigma level is shown
below
PPM Defectives Sigma level
6,91,000 1
8
3,09,000 2
67,000 3
6,200 4
230 5
3.4 6
Product Definition:
Fig: DSLA Nozzle Assembly
Fig: Injector Assembly
Fig: Body of DSLA type nozzle
Step Turning
Guide Bore Drilling
Seat Profile Grinding
Inlet hole Drilling
Dowel hole drilling
Shoulder Turning
Pressure Chamber machining
9
Sack Hole
Seat Surface
Seat seen under Microscope
DMAIC Process in Six Sigma methodology: The most important methodology in Six Sigma management is perhaps the formalized improvement methodology characterized by DMAIC (definemeasureanalyzeimprove control) process. This DMAIC process works well as a breakthrough strategy. Six Sigma companies everywhere apply this methodology as it enables real improvements and real results.
Fig: Flow diagram of DMAIC methodology adopted
10
Literature Survey
Case study of manufacturing Industry
Identification of problem Industry
Create solution statement
Create improvement Ideas
Implement improvement solutionsImprove
Data Collection
Define Define customer Requirements
Identify Specific problem
Set Goals
SIPOC diagram
Data Collection Plan
Measurement System Analysis
Identify variation due to measurement system
SIPOC diagram
Measure
Analyze Process Capability Analysis
Draw conclusion from data verification
Determine root causes
Map cause & effect diagram
Make needed adjustments
Monitor Improvement progress
Establish standard measures to maintain performance
Control
Scope of future work
Improvement Results
Conclusions
DEFINE PHASE:
1. Why the project? (The Business case) DSLA nozzle parts are hardened at UDA (Hardening
process) and after subsequent chamfer grinding they come at UVA (High precision internal grinding)
machines for Guide bore and Seat grinding. The seat and guide bore surface grinding is done on UVA
and then they are sent to inspection for seat visual checking. At seat visual checking section the no. of
parts getting rejected are quite high. From Jan08 to July08 average 22600 ppm (parts per million)
were rejected due to Bad seat problem (Rejection due to other reasons are not included in the scope
of the project).
Due to these rejections the first pass yield and type wise fulfillment of parts decreases. Also Due to
added seat repair operation at UVA the m/c utilization decreases and at the same time it increases
the defect cost associated with it. By successfully implementing the project we can save up to 1, 50
TINR.per month.
2. Initial project charter
Due to this current project scenario an 8 member team was formed for this project. Source of the
project is Seat repair monthly data available with inspection dept. Enclosures include Rejection
analysis of W174.
Characteristics of the project is: Seat Repair
Measures : Monthly PPM
Defect Definition : Seat visually not OK.
3. SIPOC (SupplierInputProcessOutputCustomer):
SIPOC is a six sigma tool. The acronym SIPOC stands for Suppliers, inputs, process, outputs, and
customers. A SIPOC is completed most easily by starting from the right ("Customers") and working
towards the left.
Suppliers to UVA process are Company, TEF1, TEF2, PLP, and MSEB.
Inputs to UVA process are Man, Machine, Electricity, Drawings, and H.T. over parts, Gauges, Tooling
Compressed air, JML, Cutting oil, Check list , Instruction charts, Program etc.
Process taking place at UVA process is Internal grinding of seat surface.
Output of the UVA process are Seat Grinding over parts, Worn out tooling, Grinding muck, PMI chart,
Rerelease chart.
Customers of the UVA process are Inspection, Repair process, Stores, Scrap yard, Etamic check,
Honing, Profile Grinding.
11
Using this data a SIPOC diagram is created.
Fig: SIPOC for UVA (Internal grinding) process.
4. CTQ (Critical to Quality) Identification:
A CTQ tree (Criticaltoquality tree) is used to decompose broad customer requirements into more
easily quantified requirements. CTQ Tree is often used in the Six Sigma methodology.
CTQs are derived from customer needs. Customer delight may be an addon while deriving Critical to
Quality parameters. For cost considerations one may remain focused to customer needs at the initial
stage. CTQs (Critical to Quality) are the key measurable characteristics of a product or process
whose performance standards or specification limits must be met in order to satisfy the customer.
They align improvement or design efforts with customer requirements. CTQs represent the product or
service characteristics that are defined by the customer (internal or external). They may include the
upper and lower specification limits or any other factors related to the product or service. A CTQ
usually must be interpreted from a qualitative customer statement to an actionable, quantitative
business specification. To put it in layman's terms, CTQs are what the customer expects of a
product... the spoken needs of the customer. The customer may often express this in plain English,
but it is up to us to convert them to measurable terms using tools such as DFMEA, etc.
The requirements of the output of the process and measures of Critical Process Issues are collected
as CTQ’s. They have to be derived from customer/business requirements, risks, economics, and
regulations. CTQ’s could be the combination of CTB’s or CTC’s, where CTB means Critical to
Business and CTC means Critical to Customer.
The CTQ tree is generated because it
SUPPLIER INPUT PROCESS OUTPUT CUSTOMER
CompanyElectricity
MaintenanceTEF1
Purchase
ManMachineElectricityDrawingsH.T. over partsGauges, ToolingCompressed airJML ,Cutting oilCheck list Instruction chartsProgram
UVA process
High Precision InternalGrindingProcess
Seat Grinding over partsWorn out toolingGrinding muckPMI chartRerelease chart
InspectionRepair process
Stores Scrap yard
Etamic check,Honing
Profile Grinding
12
Soft Stage Operations
Hardening UVA process(High Precision
Internal Grinding)
Profile Grinding
Seat Visual Inspection
– Translates broad customer/business requirements into specific criticaltoquality (CTQ)
requirements.
– Helps the team to move from highlevel (Big Y) to Specific Measurable CTC/CTB (Small “y”).
– Ensures that all aspects of the need are addressed.
CTQ tree is generated when there are Unspecific customer/business requirements or complex, broad
needs from the customer.
Steps to create CTQ diagram?
1. List customer/business needs.
2. Identify the major drivers for these needs (major means those which will ensure that the need
is addressed).
3. Break each driver into greater detail.
3. Stop the breakdown of each level when you have reached sufficient detailed information that
enables you to measure whether you meet the customer/business need or not.
Fig: CTQ tree for UVA process.
By the reference of CTQ tree there are 5 elements in UVA process seat repair. To select the right
CTQ for the project Pareto Analysis was performed on the data gathered from Jan’08 to July’08.
Pareto Analysis:
The Pareto chart was introduced in the 1940s by Joseph M. Juran, who named it after the Italian
economist and statistician Vilfredo Pareto, 1848–1923. It is applied to distinguish the “vital few from
the trivial many” as Juran formulated the purpose of the Pareto chart. It is closely related to the so
called 80/20 rule – “80% of the problems stem from 20% of the causes,” or in Six Sigma terms “80%
of the poor values in Y stem from 20% of the Xs.”
In the Six Sigma improvement methodology, the Pareto chart has two primary applications. One is for
selecting appropriate improvement projects in the define phase. Here it offers a very objective basis
for selection, based on, for example, frequency of occurrence, cost saving and improvement potential
in process performance. The other primary application is in the analyze phase for identifying the vital
few causes (Xs) that will constitute the greatest improvement in Y if appropriate measures are taken.
A procedure to construct a Pareto chart is as follows:
1) Define the problem and process characteristics to use in the diagram.
Seat repair
Guide bore repair
Taper repair
Repair
Scrap
Seat scrap
Guide bore scrap
To reduce UVA
processRepair
13
2) Define the period of time for the diagram – for example, weekly, daily, or shift. Quality
improvements over time can later be made from the information determined within this step.
3) Obtain the total number of times each characteristic occurred.
4) Rank the characteristics according to the totals from step 3.
5) Plot the number of occurrences of each characteristic in descending order in a bar graph along with
a cumulative percentage overlay.
6) Trivial columns can be lumped under one column designation; however, care must be exercised
not to omit small but important items.
From this Analysis we clearly see that Seat repair is the most critical of all rejections.
Kano model of Quality:
The Kano model is a theory of product development and customer satisfaction developed in the 80's
by Professor Noriaki Kano which classifies customer preferences into five categories:
Attractive
OneDimensional
MustBe
Indifferent
Less the better
The Kano model offers some insight into the product attributes which are perceived to be important to
customers. The purpose of the tool is to support product specification and discussion through better
development team understanding. Kano's model focuses on differentiating product features, as
opposed to focusing initially on customer needs. Kano also produced a methodology for mapping
consumer responses to questionnaires onto his model.
14
As per Kano model of Quality A CTQ specification table is generated for giving the specifications of
rejections.
CTQ MEASURE SPECIFICATION DEFECT DEFINITION KANO STATUS
G.B. Repair Monthly PPM G.B. size out of
specificationMust Be
Seat Repair Monthly PPMSeat Damage/
Finish BadSeat visually not O.K. Less the Better
Taper bad
RepairMonthly PPM 
Taper out of
specificationLess the Better
G.B. Scrap Monthly PPM G.B. size out of
specificationLess the Better
Seat Scrap Monthly PPM Seat Damage Seat visually not O.K. Less the Better
Fig: CTQ table
15
5. FINAL PROJECT CHARTER:
BOSCH Six sigma project charterProject: To reduce UVA process (High pressure Internal grinding process) seat repair.1. Background & reasons for selecting the project :Average seat repair from Jan.08 to May.08 is 22,600 ppm.2. Aim of the project :To reduce defect cost.To increase first pass yield.To increase m/c utilization.To increase type wise fulfillment.To reduce process seat repair from 22,600 to 11,300 ppm.3. Sponsor: Mr. Klaassen Enno. 4.Team Leader : Kulkarni Rahul G.5.Team members: 6.Mr.Khiratkar Ajay 1. Mr.Ganesh B. 7. Mr.Vidwans Devendra2. Mr.Mishra Vipin 8.W174 Associates3. Mr. Deshak rahul 9.W174 Foremen4. Mr.Singhal K.P 10.W172 Associates5. Mr.Jadhav Kailash 11.W172 Foremen
6.Characteristics of output of product / process & it’s measures:
Characteristics Measures Defect Identification
Seat Repair Monthly PPM Seat visually not OK.
7. Source of the project:Monthly repair data available with inspection.
8. Benefits / Cost impact:Process seat repair reduction from 22,600 ppm. To 11,300 ppm. (Expected saving of 150 TINR per month
9. Meeting frequency:Every Thursday at 03.30pm.
10. Enclosures:Rejection analysis of W174.
Dept: MFN2 Page: 1 Revision : 0 Date: 05/08/2008
16
MEASURE PHASE:
Fig: Approach to measure phase.
Creating a data collection plan: As per the approach specified a plan for collecting the base line
data is created. It is given below.
Data Collection Plan Action: Data collection from Seat Rejection
What question do you want to answer? Body seat visually OK?
Data Operational definition and procedures
WhatMeasure type/
data type
How
measured
Related
conditions to
record
Sampling
notes
How/where
recorded
(Attached form)
Seat defects Discrete data visually lot wise 100% 
Fig: Data collection plan
During weekly project meeting it was decided to change the format for recording of parts checked at
seat visual section as it was outdated. So with the help of line foremen new format was developed by
brainstorming. It is as follows:
New format developed for Seat visual section:
Collect baseline data on defects & possible causes
Develop a sampling strategy
Validate your measurement system using Gauge R & R.
Analyze patterns in data
Determine process capability
17
Ungroundseat
No sack hole
Rubbing at sack hole
PatchesRingsBad
FinishTypeLot No.Scrap
Seat DefectsItem No.
QtyRejected
Qty.OK
Qty.Inspected
Token No: Name_________________________ShiftDate
BOSCH Nashik plant
Ungroundseat
No sack hole
Rubbing at sack hole
PatchesRingsBad
FinishTypeLot No.Scrap
Seat DefectsItem No.
QtyRejected
Qty.OK
Qty.Inspected
Token No: Name_________________________ShiftDate
BOSCH Nashik plant
Segregation of defects observed at seat visual section:
Pareto Analysis of Seat rejections:
18
162779219215025665058
00044331782/9/2008Day20001483712991/9/2008Day1901669553216330/08/08Day180000633810128/08/08Day1700018632119270808Day1600016948219226/08/08Day15
00067011318925/08/08Day14200129020430823/08/08Day1310089011521422/08/08Day120100957417020/08/08Day11720056208519/08/08Day1010004324618/08/08Day9
280572939045017/08/08Day81004788016314/08/08Day7
100224204713/08/08Day6
0102516541660712/8/2008Day5
012412440646111410/8/2008Day4
0029631001748/8/2008Day3
11561721823677/8/2008Day2
0110932773726/8/2008Day1
Rubbing at sack hole end
due to burr
No sackhole
Ungroundseat
PatchesRingsBad finish
(rough surface)Total no. of
parts checkedDateDay count
162779219215025665058
00044331782/9/2008Day20001483712991/9/2008Day1901669553216330/08/08Day180000633810128/08/08Day1700018632119270808Day1600016948219226/08/08Day15
00067011318925/08/08Day14200129020430823/08/08Day1310089011521422/08/08Day120100957417020/08/08Day11720056208519/08/08Day1010004324618/08/08Day9
280572939045017/08/08Day81004788016314/08/08Day7
100224204713/08/08Day6
0102516541660712/8/2008Day5
012412440646111410/8/2008Day4
0029631001748/8/2008Day3
11561721823677/8/2008Day2
0110932773726/8/2008Day1
Rubbing at sack hole end
due to burr
No sackhole
Ungroundseat
PatchesRingsBad finish
(rough surface)Total no. of
parts checkedDateDay count
Measurement System Analysis:
A Measurement System Analysis, abbreviated MSA, is a specially designed experiment that seeks
to identify the components of variation in the measurement.
Just as processes that produce a product may vary, the process of obtaining measurements and data
may have variation and produce defects. A Measurement Systems Analysis evaluates the test
method, measuring instruments, and the entire process of obtaining measurements to ensure the
integrity of data used for analysis (usually quality analysis) and to understand the implications of
measurement error for decisions made about a product or process. MSA is an important element
of Six Sigma methodology and of other quality management systems.
MSA analyzes the collection of equipment, operations, procedures, software and personnel that
affects the assignment of a number to a measurement characteristic.
A Measurement Systems Analysis considers the following:
Selecting the correct measurement and approach
Assessing the measuring device
Assessing procedures & operators
Assessing any measurement interactions
Calculating the measurement uncertainty of individual measurement devices and/or
measurement systems
Common tools and techniques of Measurement Systems Analysis include: calibration studies, fixed
effect ANOVA, components of variance, Attribute Gage Study, Gage R&R, ANOVA Gage R&R,
Destructive Testing Analysis and others. The tool selected is usually determined by characteristics of
the measurement system itself.
Accuracy and Precision:
In the fields of engineering, industry and statistics, accuracy is the degree of closeness of
a measured or calculated quantity to its actual (true) value. Accuracy is closely related to precision,
also called reproducibility or repeatability, the degree to which further measurements or calculations
show the same or similar results. The results of calculations or a measurement can be accurate but
not precise, precise but not accurate, neither, or both. A measurement system or computational
method is called valid if it is both accurate and precise. The related terms are bias (nonrandom or
directed effects caused by a factor or factors unrelated by the independent variable)
and error (random variability), respectively.
19
Fig: Accuracy indicates proximity to the true value, precision to the repeatability or reproducibility of
the measurement.
Accuracy versus precision:
Accuracy is the degree of veracity while precision is the degree of reproducibility. The analogy used
here to explain the difference between accuracy and precision is the target comparison. In this
analogy, repeated measurements are compared to arrows that are shot at a target. Accuracy
describes the closeness of arrows to the bull’s eye at the target center. Arrows that strike closer to the
bull’s eye are considered more accurate. The closer a system's measurements to the accepted value,
the more accurate the system is considered to be.
To continue the analogy, if a large number of arrows are shot, precision would be the size of the arrow
cluster. (When only one arrow is shot, precision is the size of the cluster one would expect if this were
repeated many times under the same conditions.) When all arrows are grouped tightly together, the
cluster is considered precise since they all struck close to the same spot, if not necessarily near the
bull’s eye. The measurements are precise, though not necessarily accurate.
Fig: High accuracy, but low precision Fig: High precision, but low accuracy
However, it is not possible to reliably achieve accuracy in individual measurements without precision
if the arrows are not grouped close to one another, they cannot all be close to the bull’s eye.
(Their average position might be an accurate estimation of the bull’s eye, but the individual arrows are
inaccurate.) See also Circular error probable for application of precision to the science of ballistics.
20
Factors affecting MSA include:
Equipment: measuring instrument, calibration, fixturing, etc
People: operators, training, education, skill, care
Process: test method, specification
Samples: materials, items to be tested (sometimes called "parts"), sampling plan, sample
preparation, etc
Environment: temperature, humidity, conditioning, preconditioning,
Management: training programs, metrology system, support of people, support of quality
management system, etc.
ANOVA Gauge Repeatability & Reproducibility: (GRR study)
ANOVA Gauge R&R (or ANOVA Gauge Repeatability & Reproducibility) is a Measurement Systems
Analysis technique which uses Analysis of Variance (ANOVA) model to assess a measurement
system. The evaluation of a measurement system is not limited to gauges (or gages) but to all types
of measuring instruments, test methods, and other measurement systems.
ANOVA Gauge R&R measures the amount of variability induced in measurements that comes from
the measurement system itself and compares it to the total variability observed to determine the
viability of the measurement system. There are several components affecting a measurement system
including:
Measuring instruments, the gauge or instrument itself and all mounting blocks, supports,
fixtures, load cells etc. The machine ease of use, sloppiness among mating parts, "zero" blocks
are examples of sources of variation in the measurement system;
Operators (people), the ability and/or discipline of a person to follow the written or verbal
instructions.
Test methods, how to setup your devices, how to fixture your parts, how to record the data, etc.
Specification, the measurement is reported against a specification or a reference value. The
range or the engineering tolerance does not affect the measurement, but is an important factor
affecting the viability of the measurement system.
Parts (what is being measured), some items are easier to measure than others. A measurement
system may be good for measuring steel block length but not for measuring rubber pieces.
21
There are two important aspects on a Gauge R&R:
1. Repeatability is the variation in measurements taken by a single person or instrument on the
same item and under the same conditions. A measurement may be said to be repeatable when this
variation is smaller than some agreed limit. Repeatability conditions include:
the same measurement procedure
the same observer
the same measuring instrument, used under the same conditions
the same location
Repetition over a short period of time.
The repeatability coefficient is a precision measure which represents the value below which the
absolute difference between two repeated test results may be expected to lie with a probability of
95%. The standard deviation under repeatability conditions is part of precision and accuracy.
2. Reproducibility is the variability induced by the operators. It is the variation induced when
different operators (or different laboratories) measure the same part.
Reproducibility is one of the main principles of the scientific method, and refers to the ability of a test
or experiment to be accurately reproduced, or replicated, by someone else working independently.
The results of an experiment performed by a particular researcher or group of researchers are
generally evaluated by other independent researchers who repeat the same experiment themselves,
based on the original experimental description. Then they see if their experiment gives similar results
to those reported by the original group. The result values are said to be commensurate if they are
obtained (in distinct experimental trials) according to the same reproducible experimental description
and procedure. Reproducibility is different from repeatability, which measures the success rate in
successive experiments, possibly conducted by the same experimenters. Reproducibility relates to
the agreement of test results with different operators, test apparatus, and laboratory locations. It is
often reported as a standard deviation.
How to perform GR & R:
The Gauge R&R (GRR) is performed by measuring parts using the established measurement system.
The goal is to capture as many sources of measurement variation as possible, so they can all be
assessed and addressed. Please note that the purpose is not to "pass". A small variation reported on
a GRR may be because an important source of error was missed during the study.
To capture reproducibility errors, multiple operators are needed. Some (ASTM code) call for at least
ten operators (or laboratories) but others use only 2 or 3 to measure the same parts. To capture
repeatability errors, the same part is usually measured several times per operator. To capture
interactions of operators with parts (e.g. one part may be more difficult to measure than other), usually
between 5 and 10 parts are measured.
22
There is not a universal criteria for minimum requirements for the GRR matrix, being up to the Quality
Engineer to assess risks depending on how critical the measurement is and how costly they are. The
30x2x2 (30 parts, 2 operators, 2 repetitions) is acceptable for some studies, although it has very few
degrees of freedom for the operator component. Several methods of determining the sample size and
degree of replication are available
In this project GRR study I along with a quality over checker took 30 parts and checked its angle
twice. The recorded measurements were fed to standard Minitab software and the results obtained
are as follows:
If GRR <10 Gauge is acceptableIf 10<GRR<30 Gauge is conditionally acceptableIf 30<GRR Gauge is unacceptable & must be replaced/modified.
Misconceptions about GR & R:
Need only one GRR per family of gauges. It is usual to say "There is an acceptable GRR for
this caliper". This statement is false, as a GRR is for the measurement system, which includes the
part, specification, operator and method. As an example, measuring a steel block with a caliper
may be achieved with a good precision, but the same caliper may not be suitable to measure soft
rubber parts that may deform while it is being measured.
The GRR will not pass using parts, so it has to be done with standard weights and blocks.
The GRR done in this way will assess the precision while measuring standard weights. The
device might not be suitable to measure that specific type of parts. If the part "changes" while
being measured, this has to be counted as a measurement system error.
Need to report on PPAP documentation GRR results for everything that is measured. This
is not necessarily a requirement. The Quality Engineer usually makes an educated assessment. If
the characteristic is critical to safety, a valid GRR is required. Instead, if there is enough
understanding that some particular part is easy to measure with acceptable precision, a formal
GRR is not required. Customers may ask for additional GRRs during PPAP reviews. Knowing that
a GRR is not good and still uses the measurement system does not make sense. This is like
using bent calipers to get measurements, you get a number but it does not mean anything.
Performing a GRR is very expensive. To perform a GRR usually a number of parts (sometimes
between 5 to 10) is required to be measured by at least 3 operators (some suggest ten or more) 2
to 3 times. So the measurement costs are the ones associated with those additional
measurements. For simple devices this may not be very costly, and the results is a known
measurement error that can be used to assess all measurements subsequent to that. The costs
can be higher for destructive testing.Measuring Table20249 Measuring Table19389
Gage R & R 18.82 13.23
No. Of Distinct Categories 8 10
23
GRRs must be within 10% to pass. There are AIAG guidelines for GRR errors relative to the
specification, and what to report on a PPAP process. The final call is between the supplier and
customer, and it is a function of the criticality of the characteristic and the assessed measurement
error. GRR is a tool that helps making this assessment, but it does not give you the answer.
Process Capability Analysis
Process capability analysis was performed to find out the actual state of the process.
Minitab was used to draw a process capability analysis curve for Seat Rejections measured over a
month. As the data is discrete the Sigma level what we get is in terms of PPM (Defective Parts per
Million Opportunities)The Minitab output obtained for the Analysis is shown below.
Fig 8: Process Capability analysis of Seat visual process before
Implementing DMAIC methodology
From Results the PPM Def level is 22,624 (i.e.22, 624 Defectives in 1 Million parts.)
The below table shows different Sigma levels for PPM rejections.
PPM Defectives Sigma level
6,91,000 1
3,09,000 2
67,000 3
6,200 4
230 5
3.4 6
Fig: PPM defectives & Sigma level Comparison
By doing interpolation between 3 & 3σ levels the Sigma level of the Seat visual process comes out to
be 3.5 Sigma.
ANALYZE PHASE:
24
To analyze the defects and its generation, the tool of brainstorming is used. The suspected Sources
of variations were identified using tool of tree diagram.
Brainstorming: It is a group creativity technique designed to generate a large number of ideas for
the solution of a problem. The method was first popularized in the late 1930s by Alex Faickney
Osborn in a book called Applied Imagination. Osborn proposed that groups could double their creative
output with brainstorming. Four basic rules were followed in brainstorming. These ware intended to
reduce social inhibitions among group’s members, stimulate idea generation, and increase overall
creativity of the group.
1. To keep Focus on quantity: This rule was a means of enhancing divergent production, aiming
to facilitate problem solving through the maxim, quantity breeds quality. The assumption is that
the greater the number of ideas generated, the greater the chance of producing a radical and
effective solution.
2. To withhold criticism: In brainstorming, criticism of ideas generated was put 'on hold'. Instead,
participants focused on extending or adding to ideas, reserving criticism for a later 'critical stage'
of the process. By suspending judgment, participants feel free to generate unusual ideas.
3. Welcome unusual ideas: To get a good and long list of ideas, unusual ideas were welcomed.
They can be generated by looking from new perspectives and suspending assumptions. These
new ways of thinking may provide better solutions.
4. To combine and improve ideas: Good ideas may be combined to form a single better good
idea, as suggested by the slogan "1+1=3". It is believed to stimulate the building of ideas by a
process of association.
Method: Method followed during brainstorming is as follows
Set the problem: Before a brainstorming session, it is critical to define the problem. The problem
must be clear, not too big, and captured in a specific question. If the problem is too big, the facilitator
should break it into smaller components, each with its own question.
Create a background memo: The background memo is the invitation and informational letter for the
participants, containing the session name, problem, time, date, and place. The problem is described
in the form of a question, and some example ideas are given. The memo is sent to the participants
well in advance, so that they can think about the problem beforehand.
Select participants: The facilitator composes the brainstorming panel, consisting of the participants
and an idea collector. A group of 10 or fewer members is generally more productive. Many variations
are possible but the following composition is suggested.
Several core members of the project who have proved themselves.
Several guests from outside the project, with affinity to the problem.
One idea collector who records the suggested ideas.
25
Create a list of lead questions: During the brainstorm session the creativity may decrease. At this
moment, the facilitator should stimulate creativity by suggesting a lead question to answer, such
as Can we combine these ideas? Or How about looking from another perspective? It is best to
prepare a list of such leads before the session begins.
The process
Participants who have ideas but were unable to present them are encouraged to write down
the ideas and present them later.
The idea collector should number the ideas, so that the chairperson can use the number to
encourage an idea generation goal, for example: We have 44 ideas now, let’s get it to 50!.
The idea collector should repeat the idea in the words he or she has written verbatim, to
confirm that it expresses the meaning intended by the originator.
When more participants are having ideas, the one with the most associated idea should have
priority. This to encourage elaboration on previous ideas.
During a brainstorming session, managers and other superiors may be discouraged from
attending, as it may inhibit and reduce the effect of the four basic rules, especially the generation
of unusual ideas.
Evaluation
Brainstorming is not just about generating ideas for others to evaluate and select. Usually the group
itself will, in its final stage, evaluate the ideas and select one as the solution to the problem proposed
to the group.
The solution should not require resources or skills the members of the group do not have or
cannot acquire.
If acquiring additional resources or skills is necessary, that needs to be the first part of the
solution.
There must be a way to measure progress and success.
The steps to carry out the solution must be clear to all, and amenable to being assigned to the
members so that each will have an important role.
There must be a common decision making process to enable a coordinated effort to proceed,
and to reassign tasks as the project unfolds.
There should be evaluations at milestones to decide whether the group is on track toward a
final solution.
There should be incentives to participation so that participants maintain their efforts.
26
Fig: Tree diagram created from brainstorming session for Input part parameters
27
Chamfer height variation.
Acqueous Cleaning not okJet broken,Pump pressure less
Uneven chamfer band
Guide to shaft TR not okGuide to shaft TR not checked after TBT as per freq.
TR more than 100 microns
Measure by gauge
Vibrations & chatter marks on seat in soft stage
Roundness, Straightness, Guide bore to seat TR
No specification in drawing
UVA PROCESS REPAIR &
SCRAP
Seat repair
Rough finish, Rings, Patches, No sack hole,Rubbing at sack hole, Unground seat
I/P parts
100% sack hole checking poka yoke on all 5 spinner
Possibility of poka yoke failure
Parts without sack hole from soft
stage
Sack hole Drill breakage on Retco
Poka yoke not working properly
Type Mixup ( P type in DSLA & vise versa
Possibility on all operations during lot change, 80% on Benzinger, ECM(10%), Remaining 10%
Manual element
Guide bore to shaft T.R bad
Guide to shaft TR not checked after TBT as per freq.
TR more than 100 microns
Seat TR wrt guide bore
On spinner & retco m/cmore than 70 microns
Seat angle in soft stage
On spinner & retco m/cspecification 58.8° (+/ 0.2°)
More/less than spec.
Chamfer mandrel angle in hard stage
More/less than spec.
Fig: Tree diagram due to machine related parametersFrom two tree diagrams created above it is clear that there are 7 parameters related to input part
parameters & 23 machine related parameters. To know the impact of each parameter on seat
rejections it was necessary to validate each parameter using statistical methods. In Six Sigma method
used for root cause validation is Hypothesis testing.
Statistical hypothesis testing:
A statistical hypothesis test is a method of making statistical decisions using experimental data. It is
sometimes called confirmatory data analysis. In frequency probability, these decisions are almost
always made using nullhypothesis tests; that is, ones that answer the question assuming that the null
hypothesis is true, what is the probability of observing a value for the test statistic that is at least as
extreme as the value that was actually observed? The use of hypothesis testing is deciding whether
experimental results contain enough information to cast doubt on conventional wisdom.
Null hypothesis (H0) formally describes some aspect of the statistical behaviour of a set of data; this
description is treated as valid unless the actual behaviour of the data contradicts this assumption.
Thus, the null hypothesis is contrasted against another hypothesis. Statistical hypothesis testing is
used to make a decision about whether the data contradicts the null hypothesis: this is called
significance testing. A null hypothesis is never proven by such methods, as the absence of evidence
against the null hypothesis does not establish it. In other words, one may either reject, or not
reject the null hypothesis; one cannot accept it. Failing to reject it gives no strong reason to change
28
decisions predicated on its truth, but it also allows for the possibility of obtaining further data and then
reexamining the same hypothesis.
Alternative hypothesis is always set out for a particular significance test in conjunction with a null
hypothesis. Although in some cases it may seem reasonable to consider the alternative hypothesis as
simply the negation of the null hypothesis, this would be misleading. In fact, significance testing and
statements about hypotheses always take place within the context of a set of assumptions (which may
unfortunately be unstated). This provides a way of considering alternative hypotheses which are the
negation of the null hypothesis within the context of the overall assumptions. However not all
alternative hypotheses are of this "negation type": the simplest cases are directional hypotheses. An
important case arises in testing for differences across a number of different groups, where the null
hypothesis may be "no difference across groups" with the alternative hypothesis being that the mean
values for the groups would be in a certain prespecified order. In the theory of statistical hypothesis
testing, the triple of "assumptions", "null hypothesis" and "alternative hypothesis" provides the basis
for choosing an appropriate test statistic.
Example
For example, one may want to compare the test scores of two random samples of men and women,
and ask whether or not one group (population) has a mean score (which really is) different from the
other. A null hypothesis would be that the mean score of the male population was the same as the
mean score of the female population:
H0 : μ1 = μ2
Where:
H0 = the null hypothesis
μ1 = the mean of population 1, and
μ2 = the mean of population 2.
Alternatively, the null hypothesis can postulate (suggest) that the two samples are drawn from the
same population, so that the variance and shape of the distributions are equal, as well as the means.
Formulation of the null hypothesis is a vital step in testing statistical significance. One can then
establish the probability of observing the obtained data (or data more different from the prediction of
the null hypothesis) if the null hypothesis is true. That probability is what is commonly called the
"significance level" of the results.
That is, in scientific experimental design, we may predict that a particular factor will produce an effect
on our dependent variable — this is our alternative hypothesis. We then consider how often we would
expect to observe our experimental results or results even more extreme, if we were to take many
samples from a population where there was no effect (i.e. we test against our null hypothesis). If we
find that this happens rarely (up to, say, 5% of the time), we can conclude that our results support our
experimental prediction — we reject our null hypothesis.
29
Pvalue: In statistical hypothesis testing, the pvalue is the probability of obtaining a result at least as
extreme as the one that was actually observed, assuming that the null hypothesis is true. The fact that
pvalues are based on this assumption is crucial to their correct interpretation.
The lower the pvalue, the less likely the result, assuming the null hypothesis, so
the more "significant" the result, in the sense of statistical significance – one often uses pvalues of
0.05 or 0.01, corresponding to a 5% chance or 1% of an outcome that extreme, given the null
hypothesis. More technically, a pvalue of an experiment is a random variable defined over
the sample space of the experiment such that its distribution under the null hypothesis is uniform on
the interval [0,1]. Many pvalues can be defined for the same experiment.
Generally, one rejects the null hypothesis if the pvalue is smaller than or equal to the significance
level,often represented by the Greek letter α (alpha). If the level is 0.05, then results that are only 5%
likely or less are deemed extraordinary, given that the null hypothesis is true.
In the above example we have:
null hypothesis (H0) — fair coin;
observation (O) — 14 heads out of 20 flips; and
Probability (pvalue) of observation (O) given H0 — p(O  H0) = 0.0577 × 2 (twotailed) =
0.1154 (percentage expressed as 11.54%).
The calculated pvalue exceeds 0.05, so the observation is consistent with the null hypothesis — that
the observed result of 14 heads out of 20 flips can be ascribed to chance alone — as it falls within the
range of what would happen 95% of the time were this in fact the case. In our example, we fail to
reject the null hypothesis at the 5% level. Although the coin did not fall evenly, the deviation from
expected outcome is just small enough to be reported as being "not statistically significant at the 5%
level".
However, had a single extra head been obtained, the resulting pvalue (twotailed) would be 0.0414
(4.14%). This time the null hypothesis  that the observed result of 15 heads out of 20 flips can be
ascribed to chance alone  is rejected. Such a finding would be described as being "statistically
significant at the 5% level".
Some common misunderstandings about pvalues.
1. The pvalue is not the probability that the null hypothesis is true. (This false conclusion is
used to justify the "rule" of considering a result to be significant if its pvalue is very small.)
In fact, frequentist statistics does not, and cannot, attach probabilities to hypotheses.
Comparison of Bayesian and classical approaches shows that a pvalue can be very close to
zero while the posterior probability of the null is very close to unity. This is the Jeffreys
Lindley paradox.
2. The pvalue is not the probability that a finding is "merely a fluke." (Again, this conclusion
arises from the "rule" that small pvalues indicate significant differences.)
As the calculation of a pvalue is based on the assumption that a finding is the product of
chance alone, it patently cannot also be used to gauge the probability of that assumption
30
being true. This is subtly different from the real meaning which is that the pvalue is the
chance that null hypothesis explains the result: the result might not be "merely a
fluke," andbe explicable by the null hypothesis with confidence equal to the pvalue.
3. The pvalue is not the probability of falsely rejecting the null hypothesis. This error is a version
of the socalled prosecutor's fallacy.
4. The pvalue is not the probability that a replicating experiment would not yield the same
conclusion.
5. 1 − (pvalue) is not the probability of the alternative hypothesis being true.
6. The significance level of the test is not determined by the pvalue.
The significance level of a test is a value that should be decided upon by the agent
interpreting the data before the data are viewed, and is compared against the pvalue or any
other statistic calculated after the test has been performed.
7. The pvalue does not indicate the size or importance of the observed effect (compare
with effect size).
31
Validation of all SSVs using Statistical testing: (Input part parameters)
conc
lusi
ons
The
impa
ct o
f aq
ueou
s cl
eani
ng o
n ch
amfe
r he
ight
va
riatio
n is
In
sig
nif
ican
t.
The
impa
ct o
f cha
mfe
r he
ight
var
iatio
n on
se
at r
ejec
tions
is
Insi
gn
ific
ant
The
impa
ct o
f U
neve
n ch
amfe
r ba
nd o
n S
eat
reje
ctio
ns is
S
ign
ific
ant
The
impa
ct o
f dril
l life
on
sea
t rej
ectio
ns is
In
sig
nif
ican
t
The
impa
ct o
f dril
l da
mag
e in
sof
t sta
ge
on S
eat r
ejec
tions
is
Sig
nif
ican
t.
Res
ults
obt
aine
d
0 ba
d pa
rts
in
275
ok p
arts
0 ba
d pa
rts
in 2
5 w
ithou
t cle
anin
g
part
s
All
part
s ca
me
ok
on U
VA
, cha
mfe
r he
ight
var
iatio
n di
d no
t cau
se
any
defe
ct o
n U
VA
.
12 p
arts
bad
in
50 T
R b
ad p
arts
1 ba
d in
50
TR
ok
par
ts
The
sea
t RZ
&
Rm
ax v
alue
s of
al
l par
ts a
re
with
in li
mits
49 b
ad in
50
with
ch
atte
r m
arks
, 1
bad
in 5
0 w
ithou
t ch
atte
r m
arks
Tes
t use
d
2 pr
opor
tions
test 2
prop
ortio
ns te
st 2 pr
opor
tions
te
st 2 pr
opor
tions
te
st
End
date
8N
ov0
8
15N
ov
08
3M
ar0
9
8Ja
n08
8Ja
n08
Sta
rt d
ate
8N
ov0
8
15N
ov
08
3M
ar0
9
16D
ec
08
16D
ec
08
Tria
l tak
en
Tak
e 27
5 pa
rts
with
cle
anin
g &
25
part
s w
ithou
t cle
anin
g &
pro
cess
th
em o
n sa
me
cham
fer
grin
ding
m
/c &
sam
e U
VA
m/c
.
Tak
e 30
par
ts w
ith c
ham
fer
heig
ht
(30
to 
10µ
), 6
0 pa
rts
with
in s
pec
(10
µ to
+10
) &
30
part
s w
ith (
+10
to
+30
µ)
& p
roce
ss th
em o
n U
VA
.
Tak
e 50
par
ts w
ith T
R m
ore
than
85µ
& p
ut th
em o
n U
VA
als
o pr
oces
s 50
nor
mal
par
ts
One
par
t fro
m e
ach
mac
hine
gi
ven
to F
MR
lab,
Life
no.
are
not
ed
50 p
arts
with
cha
tter
mar
ks w
ere
pro
cess
ed o
n U
VA
alo
ng w
ith 5
0 ok
par
ts
Act
ions
take
n
Tak
e a
tria
l whi
ch in
volv
es
proc
essi
ng p
arts
with
out
aque
ous
clea
ning
.
To
take
a tr
ial t
his
invo
lves
ta
king
par
ts w
ith c
ham
fer
heig
ht m
ore,
less
& w
ithin
sp
ecifi
catio
n &
pro
cess
ing
them
on
UV
A.
A tr
ial T
R c
heck
ing
gaug
e is
dev
elop
ed
Tak
e on
e pa
rts
each
from
sp
inne
rs &
Ret
co h
avin
g
diffe
rent
tool
life
& g
ive
them
to
FM
R la
b fo
r se
at fo
rm
chec
king
Whe
n su
ch p
arts
com
e on
U
VA
sort
out
suc
h pa
rts
& p
ut
them
on
UV
A fo
r tr
ial.
Sus
pect
ed s
ourc
es
of v
aria
tions
(SS
V's
)
Sea
t doe
s no
t get
cl
eane
d pr
oper
ly s
o lo
catio
n of
par
t on
cham
fer
grin
ding
m
/c is
out
side
due
to
dirt
pre
sent
. Thi
s ou
tsid
e lo
catio
n re
sults
in s
eat
reje
ctio
ns.
Par
t loc
atio
n in
UV
A
beco
mes
impr
oper
due
to
cham
fer
varia
tion.
Gui
de to
sha
ft T
R is
no
t c
heck
ed in
sof
t st
age
The
dril
l for
m
dete
riora
tes
with
us
age
& th
e pa
rts
at
late
r st
ages
of
tool
life
hav
e
mor
e ro
ughn
ess
Due
to d
rill d
amag
e on
mac
hine
s v
ibra
tions
& d
eep
lines
are
pro
duce
d on
sea
t.
sub
caus
e
Aqu
eous
cl
eani
ng n
ot o
kJe
t bro
ken,
P
ump
pres
sure
le
ss
Cha
mfe
r he
ight
va
riatio
n ca
uses
se
at r
ejec
tions
at
UV
A
Gui
de to
sha
ft
TR
not
ok
Rou
ndne
ss,
Str
aigh
tnes
s G
B
to s
eat T
R n
ot
chec
ked
in s
oft
stag
e
Dril
l dam
age
on
Spi
nner
s &
R
etco
Roo
t ca
use
cham
fer
heig
ht
varia
tions
Une
ven
ch
amfe
r ba
nd
Vib
ratio
ns
&
chat
ter
mar
ks o
n se
at in
so
ft st
age
Sr.
No. 1 2 3
32
(Input part parameters continued..)
Con
clus
ions
The
impa
ct o
f No
sack
hol
e p
arts
on
se
at r
ejec
tions
is
Sig
nif
ican
t
The
impa
ct o
f ty
pe
mix
up
on
Sea
t rej
ectio
ns is
S
ign
ific
ant.
The
impa
ct o
f S
eat a
ngle
mor
e on
sea
t rej
ect
ions
is
Insi
gn
ific
an
t
The
impa
ct o
f ch
amfe
r m
and
rel
angl
e on
se
at
reje
ctio
ns is
In
sig
nif
ica
nt
Res
ults
obt
aine
d
No
sac
k ho
le p
art
br
eaks
the
grin
din
g
wh
eel t
ip &
m/c
ge
ts im
med
iate
ly
sto
pped
, du
ring
redr
essi
ng 5
0 pa
rts
cam
e b
ad.
pty
pe in
DS
LA lo
t br
eaks
the
ad
apto
r& g
rindi
ng
wh
eel,
whi
ch
resu
lts in
50
bad
in
50,w
ith n
orm
al
part
s 0
bad
in 5
0.
3 ba
d in
285
ang
le
mor
e p
arts
,0
bad
in 3
00 a
ngle
ok
par
ts
As
the
re in
no
va
riatio
n in
ou
tput
sta
tistic
al
test
can
not
be
perf
orm
ed
Tes
t use
d
2 pr
opor
tions
test
2pr
opor
tions
test
2pr
opor
tions
test
No
va
riatio
n in
out
put
End
date
13J
an
09
20N
ov
08
28N
ov
08
25D
ec
08
Sta
rt d
ate
13J
an
09
20N
ov
08
21N
ov
08
25N
ov
08
Tria
l tak
en
On
e no
sac
k ho
le
part
wa
s pu
t on
UV
A 2
0315
& it
's
effe
ct o
n
reje
ctio
ns w
as
obse
rved
On
e ty
pe m
ix u
p p
art
wa
s pu
t o
n U
VA
203
15
& it
's
effe
ct o
n s
eat
re
ject
ion
s is
ob
serv
ed
285
part
s w
ith s
eat
angl
e m
ore
we
re p
roce
ssed
up
to
seat
vis
ual
alon
g w
ith 3
00 a
ngle
ok
par
ts
Ch
amfe
r m
andr
el
angl
es c
heck
ed b
y S
ine
bar
met
hod
& M
icro
scop
e m
etho
d
Act
ions
take
n
Co
llect
at l
eas
t 15
No
sac
k ho
le p
art
s pr
efa
rab
ly o
f DS
LA
norm
al S
haft
Co
llect
at l
eas
t 15
mix
up
par
ts
Tria
l is
take
n w
hic
h in
volv
es
seat
an
gle
mor
e
part
s ar
e p
roce
ssed
up
to s
eat v
isua
l for
ch
ecki
ng.
4 m
andr
els
giv
en to
to
ol r
oom
fo
r ch
amfe
r an
gle
verif
icat
ion
Sus
pect
ed s
ourc
es o
f va
riatio
ns(S
SV
's)
Pok
a Y
oke
pu
t off
due
to v
ario
us
reas
ons
80%
on
75%
Ben
zing
er,
10%
on
EC
M.
Man
ual e
lem
ent
may
be
pre
sent
,E
leva
tor
cond
ition
Ang
le n
ot
chec
ked
as
per
freq
uen
cy/D
rill l
ife
over
, Dri
ll re
shar
peni
ng
impr
oper
Ch
amfe
r m
andr
el
angl
e to
be
veri
fied
in to
ol
room
sub
caus
e
Pok
a y
oke
fa
ilure
on
spin
ner
mac
hin
e
Pok
a y
oke
fa
ilure
on
Re
tco
mac
hin
e
Pos
sibi
lity
on
all o
pera
tions
On
spi
nner
&
Ret
co
mac
hin
es
Mor
e o
r le
ss th
an
spec
ifica
tion
Roo
t cau
se
Par
ts
with
out
sack
hol
e fr
om s
oft
sta
ge
Par
t typ
e m
ix u
p
Sea
t ang
le
in s
oft
sta
ge
Ch
amfe
r m
andr
el
angl
e in
sof
t st
age
Sr.
N
o. 4 5 6 7
33
Actions taken for machine related parameters
conc
lusi
ons
The
impa
ct o
f wor
khea
d vi
brat
ion
on s
eat
reje
ctio
ns is
Insi
gn
ific
ant
The
impa
ct o
f Wor
khea
d rp
m
on s
eat r
ejec
tions
is
Insi
gn
ific
ant
The
impa
ct o
f Spi
ndle
he
ight
re
peat
abili
ty o
n S
eat
reje
ctio
ns is
Insi
gn
ific
ant
The
impa
ct o
f fem
ale
cent
er
grin
ding
on
seat
re
ject
ions
is In
sig
nif
ican
t
The
impa
ct o
f Job
cl
ampi
ng p
ress
ure
on
seat
rej
ectio
ns is
In
sig
nif
ican
t.
Res
ults
obt
aine
d
Wor
khea
d vi
brat
ion
valu
es
of a
ll m
achi
nes
are
with
in 3
m
m/s
ec.
At b
oth
rpm
val
ues
all
50 p
arts
cam
e vi
sual
ly o
k
At b
oth
repe
atab
ility
leve
ls a
ll pa
rts
cam
e vi
sual
ly o
k
All
part
s be
fore
doi
ng
fem
ale
cent
er g
rindi
ng
cam
e ok
, als
o al
l par
ts a
fter
doin
g fe
mal
e ce
nter
gr
indi
ng c
ame
ok
At 5
bar
pre
ssur
e 0
bad
in
50,
at 4
bar
pre
ssur
e 29
bad
in
50 p
arts
.
Tes
t use
d
No
varia
tion
outp
ut
obse
rved
No
varia
tion
in
outp
ut
obse
rved
2pr
opor
tions
te
st
2pr
opor
tions
te
st
2 pr
opor
tions
te
st
End
date
16F
eb0
9
16F
eb0
9
12M
ar0
9
12M
ar0
9
30J
an0
9
Sta
rt d
ate
13F
eb0
9
13F
eb0
9
12M
ar0
9
12M
ar0
9
30J
an0
9
Tria
l tak
en
Wor
khea
d vi
brat
ion
valu
es o
f al
l mac
hine
s ar
e ch
ecke
d w
ith
help
of v
ibra
tom
eter
Tak
e 50
par
ts w
ith 2
150
rpm
,ta
ke 5
0 pa
rts
with
175
0 rp
m
50 p
arts
eac
h w
ere
proc
esse
d w
ith r
epea
tabi
lity
of 1
0µ &
at
20µ
.
50 p
arts
wer
e pr
oces
sed
be
fore
doi
ng fe
mal
e ce
nter
gr
indi
ng &
50
part
s w
ere
proc
esse
d af
ter
doin
g fe
mal
e ce
nter
grin
ding
The
job
clam
ping
pre
ssur
e w
as v
arie
d ti
4 ba
r &
5 b
ar &
it'
s im
pact
on
seat
rej
ectio
ns is
ob
serv
ed.
Act
ions
take
n
Che
ck w
orkh
ead
vibr
atio
n v
alue
s of
all
mac
hine
s
Rat
ed R
PM
val
ue is
215
0
RP
M
Che
ck r
epea
tabi
lity<
20µ
, ta
ke tr
ial w
ith p
roce
ssin
g pa
rts
with
diff
eren
t re
peat
abili
ty v
alue
s.
we
chec
ked
part
s be
fore
&
afte
r do
ing
fem
ale
cent
er
grin
ding
for
chec
king
di
ffere
nce
Air
supp
ly to
job
clam
ping
is
var
ied
to d
iffer
ent l
evel
s &
it's
effe
ct w
as o
bser
ved
Sus
pect
ed
sour
ces
of
varia
tions
(SS
V's
)
Ear
lier
not
know
n
valu
e18
00
rpm
Rep
eata
bilit
y be
low
20µ
Grin
ding
fr
eq. n
ot
deci
ded
Chu
ck c
lam
p gr
indi
ng
sub
caus
e
Vib
ratio
n
RP
M
Spi
ndle
he
ight
Fem
ale
ce
nter
Job
clam
ping
pr
essu
re
Roo
t cau
se
Wor
khea
d
Sr.
No. 1 2 3 4 5
34
(Machine related parameters continued…)
conc
lusi
ons
The
impa
ct o
f loa
ding
sp
ring
brok
en o
n se
at
reje
ctio
ns is
Sig
nif
ican
t
The
impa
ct o
f Loa
ding
al
ignm
ent o
f com
pone
nt
on s
eat r
ejec
tions
is
Insi
gn
ific
ant.
The
impa
ct o
f Air
cylin
der
on s
eat r
ejec
tions
is
Insi
gn
ific
ant.
The
impa
ct o
f Ang
le
mas
ter
on s
eat r
ejec
tions
is
Insi
gn
ific
ant.
The
impa
ct o
f sea
t vi
sual
mic
rosc
ope
cond
ition
on
seat
re
ject
ions
is
Sig
nif
ican
t.
The
impa
ct o
f Air
supp
ly
for
part
s cl
eani
ng o
n S
eat r
ejec
tions
is
Sig
nif
ican
t
Res
ults
obt
aine
d
with
ok
sprin
g al
l 50
part
s ca
me
ok, w
ith b
roke
n sp
ring
35 b
ad in
50.
with
& w
ithou
t che
ckin
g lo
adin
g al
ignm
ent a
ll 50
pa
rts
cam
e vi
sual
ly o
k
The
qui
ck h
it ac
hiev
ed
GR
R fo
und
to b
e ok
Whe
n 50
par
ts c
heck
e w
ith fa
ulty
mic
rosc
ope
35
cam
e ba
d, w
hen
they
are
ch
ecke
d w
ith o
k sc
ope
only
50
cam
e ba
d.
With
air
clea
ning
10
par
ts b
ad in
50,
w
ithou
t air
clea
ning
22
par
ts b
ad in
50.
Tes
t use
d
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
No
hypo
thes
is
test
per
form
ed
No
test
per
form
ed
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
End
dat
e
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
18D
ec0
8
30J
an0
9
Sta
rt d
ate
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
18D
ec0
8
30J
an0
9
Tria
l tak
en
Cha
ngin
g fr
eq. o
nce
in tw
o m
onth
s.
Tak
e a
tria
l with
out c
heck
ing
load
ing
alig
nmen
t of
com
pone
nt.
No
prob
lem
of a
ir le
akag
e
take
GR
R o
f sea
t an
gle
mas
ter
Sco
pe c
ondi
tion
stud
y sc
hedu
le to
be
prep
ared
A w
orks
hop
on m
icro
scop
e h
andl
ing
to b
e ar
rang
ed
50 p
arts
take
n w
ith a
ir cl
eani
ng
& 5
0 pa
rts
take
n w
ithou
t air
clea
ning
Act
ions
take
n
Load
ing
sprin
g w
as
chan
ged
with
a b
roke
n on
e &
it's
effe
ct o
n se
at
reje
ctio
ns w
as o
bser
ved
Whi
le s
ettin
g m
achi
ne
chec
k a
lignm
ent f
or o
k / N
ot o
k
Ele
ctric
al s
ervo
mot
or u
sed
Che
ckin
g fr
eq. t
o be
re
duce
d
Che
ck r
equi
rem
ent o
f fr
eque
nt v
erifi
catio
n of
m
icro
scop
e co
nditi
on
Ass
ocia
tes
awar
enes
s ab
out m
icro
scop
e ad
just
men
t to
be d
one.
Par
ts to
be
chec
ked
with
ai
r cl
eane
d &
with
out a
ir cl
eani
ng
Sus
pect
ed
sour
ces
of
varia
tions
(SS
V's
)
Cha
ngin
g fr
eq.
Vis
ual c
heck
Air
leak
age
Mas
ter
show
ing
wro
ng
read
ing
Alig
nmen
t of
bot
h ey
es
not t
here
Fre
quen
t ch
ecki
ng b
y as
soci
ates
No
supp
ly
prov
ided
sub
caus
e
Load
ing
sp
ring
wor
nout
Load
ing
al
ignm
ent o
f co
mpo
nent
Load
ing
cy
linde
r
Ang
le m
aste
r
Vis
ual
insp
ectio
n
mic
rosc
ope
Air
supp
ly
for
part
s cl
eani
ng
Roo
t cau
se
Load
ing
/ U
nloa
ding
Che
ckin
g be
nch
Sr.
No. 6 7 8 9 10 11
35
(Machine related parameters continued…)
conc
lusi
ons
The
impa
ct o
f Spi
ndle
co
olin
g o
n S
eat r
ejec
tion’
s is
In
sig
nif
ican
t
The
impa
ct o
f Spi
ndle
co
olin
g s
yste
m o
n S
eat
reje
ctio
ns is
In
sig
nif
ican
t.
The
impa
ct o
f Ini
tial
setti
ng o
n se
at r
ejec
tions
is
Sig
nif
ican
t
The
impa
ct o
f new
sea
t w
heel
set
ting
on S
eat
reje
ctio
ns is
Sig
nif
ican
t.
The
impa
ct o
f Ada
ptor
T
R o
n S
eat r
ejec
tions
is
Insi
gn
ific
ant.
The
impa
ct o
f Dre
ssin
g rin
g w
orn
out o
n se
at
reje
ctio
ns is
Sig
nif
ican
t.
Res
ults
obt
aine
d
3 pa
rts
in b
ad 1
00 w
ith
60,0
00 r
pm,1
bad
in 1
00 w
ith
50,0
00 r
pm
Spi
ndle
coo
ling
syst
ems
of a
ll m
achi
nes
are
foun
d to
be
wor
king
ok.
Whe
n in
itial
set
ting
ok 0
bad
in
50,
whe
n in
itial
set
ting
di
stur
bed
25 b
ad in
50.
Whe
n ne
w s
eat w
heel
set
ting
ok
0 b
ad in
50,
whe
n in
itial
se
tting
not
ok
30 b
ad in
50.
whe
n T
R<
10µ
0 b
ad in
50,
whe
n T
R>
10µ
0 b
ad in
50
with
wor
nout
spr
ing
45 b
adin
50
, with
ok
ring
2 ba
d in
50.
Tes
t use
d
2 pr
opor
tions
te
st
No
test
per
form
ed
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
End
dat
e
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
Sta
rt d
ate
15J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
Tria
l tak
en
100
part
s pr
oces
sed
with
60
,000
rpm
, 100
par
ts
proc
esse
d w
ith 5
0,00
0 rp
m
All
syst
ems
chek
ced
with
Mai
nten
ance
peo
ple
Initi
al s
ettin
g pa
ram
eter
s w
ere
dist
urbe
d &
tria
l is
take
n.
The
new
sea
t whe
el
heig
ht w
as s
et a
t 3.1
5mm
&
it's
effe
ct o
n se
at
reje
ctio
ns w
as o
bser
ved.
take
50
part
s w
ith a
dapt
or
TR
<10
µ
& a
gain
take
50
part
s w
ith
adap
tor
TR
>10
µ
One
wor
n ou
t rin
g w
as
plac
ed &
whe
el w
as
dres
sed
with
that
rin
g.
Par
ts a
re ta
ken
for
tria
l.
Act
ions
take
n
Tak
e a
tria
l with
di
ffere
nt R
PM
val
ues
Che
ck w
heth
er s
pind
le
cool
ing
sys
tem
s of
all
mac
hine
s ar
e ru
nnin
g ok
Initi
al s
ettin
g w
as
dist
urbe
d &
it's
impa
ct
on s
eat r
ejec
tions
was
ob
serv
ed
New
sea
t whe
el h
eigh
t to
be
set 3
.1 m
m, t
ake
tria
l with
mor
e he
ight
.
Ada
ptor
Tr
chec
ked
ev
ery
time
mac
hine
is
dist
urbe
d &
it's
impa
ct
on s
eat r
ejec
tions
obs
erve
d
Tria
l tak
en w
hich
in
volv
es p
laci
ng a
wor
n ou
t rin
g on
Mac
hine
&
taki
ng p
arts
Sus
pect
ed
sour
ces
of
varia
tions
valu
e to
be
60,0
00 R
PM
To
be a
sked
to
mai
ntai
nanc
e
Whe
el fo
rm w
ear
Ens
ure
posi
tive
cutti
ng a
fter
dres
sing
If T
R o
ut o
f sp
ecifi
catio
n se
at
bad
com
es
If dr
essi
ng r
ing
is
wor
n ou
t, th
e gr
indi
ng w
heel
fo
rm g
ets
dam
aged
. Due
to
whi
ch p
art c
omes
se
at b
ad.
sub
caus
e
RP
M
Spi
ndle
co
olin
g
Initi
al s
ettin
g
New
sea
t w
heel
se
tting
TR
<10
µ
Per
iodi
c re
plac
emen
t &
TR
Roo
t cau
se
Grin
ding
sp
indl
es
Set
ting
para
met
ers
Ada
ptor
s
Dre
ssin
g rin
g
Sr.
No. 12 13 14 15 16 17
(Machine related parameters continued…)
36
conc
lusi
ons
The
impa
ct o
f coo
lant
sy
stem
s on
Sea
t re
ject
ions
is
Insi
gn
ific
ant.
The
impa
ct o
f pok
a yo
ke o
n S
eat
reje
ctio
ns is
S
ign
ific
ant.
The
impa
ct o
f dre
ssin
g de
pth
of c
ut o
n S
eat
rejc
tions
is
Insi
gn
ific
ant.
The
impa
ct o
f dre
ssin
g fr
eq. o
n S
eat r
ejec
tions
is
Insi
gn
ific
ant.
The
impa
ct o
f fee
d ra
te
on S
eat
reje
ctio
ns is
In
sig
nif
ican
t.
The
impa
ct o
f Ope
rato
r eq
ualiz
atio
n on
sea
t re
ject
ions
is
Sig
nif
ican
t.
Res
ults
obt
aine
d
The
coo
lant
sys
tem
pa
ram
eter
s ar
e w
ithin
lim
its
Pok
a yo
ke o
tip
1 ba
d in
50,
whe
n po
ka y
oke
not o
n tip
16
bad
in 5
0.
0 ba
d in
50
with
3µ
de
pth
of c
ut.
0 ba
d in
50
with
2µ
de
pth
of c
ut.
0 ba
d in
50
with
6
part
s fr
eq. 0
bad
in 5
0 w
ith 8
par
ts fr
eq.
with
100
% fe
ed r
ate
all
50 p
arts
okw
ith 5
0 %
fe
ed r
ate
all 5
0 pa
rts
ok a
gain
.
Due
to c
ontin
uous
re
ject
ions
from
as
sem
bly
sect
ion
fear
is
set
in v
isua
l op
erat
ors.
Tes
t use
d
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
2 pr
opor
tions
te
st
End
dat
e
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
Sta
rt d
ate
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
30J
an0
9
Tria
l tak
en
Onl
y ch
ecki
ng is
invo
lved
as
taki
ng a
tria
l is
very
da
nger
ous.
50 p
arts
take
n w
hen
poka
yo
ke o
n tip
, aga
in 5
0 pa
rts
take
n w
ith p
oka
yoke
in
back
swor
d po
sitio
n.
Tak
e pa
rts
with
3µ
dep
th o
f cu
t.T
ake
part
s w
ith 2
µ d
epth
of
cut.
Tak
e 50
par
ts w
ith 8
par
ts
dres
sing
freq
. Aga
in ta
ke 5
0 pa
rts
with
6 p
arts
dre
ssin
g fr
eq.
Tak
e pa
rts
with
50%
feed
ra
te,
Tak
e pa
rts
with
100
% fe
ed
rate
.
50 b
orde
r ca
se p
arts
wer
e sh
own
to o
pera
tors
& th
ey
wer
e sh
own
to a
ssem
bly
oper
ator
s.
Act
ions
take
n
Che
ck p
ress
ure,
te
mpe
ratu
re
of c
oola
nt s
yste
m
Pok
a yo
ke w
as s
hifte
d to
ba
ckw
ard
posi
tion
& it
s ef
fect
on
seat
rej
ectio
ns
was
obs
erve
d.
Dre
ssin
g de
pth
of c
ut is
var
ied
& tr
ial i
s ta
ken
Dre
ssin
g fr
eq. c
hang
ed &
tria
l is
take
n
The
feed
rat
e w
as
chan
ged
man
ually
& it
's
effe
ct o
n se
at r
ejec
tions
is
obse
rved
.
Dai
ly r
ejec
tions
at s
eat
visu
al is
che
cked
for
verif
icat
ions
Sus
pect
ed s
ourc
es
of v
aria
tions
(SS
V's
)
The
dre
ssin
g/
Grin
ding
pr
essu
re v
arie
s
Con
firm
atio
n of
po
ka y
oke
once
in
a s
hift
3 m
icro
ns
6 pa
rts
Man
ual k
nob
pres
ent
Inco
rrec
t dec
isio
n du
e to
fear
of g
ettin
g re
ject
ed
from
ass
embl
y.
sub
caus
e
3.5
to 4
bar
gr
indi
ng/
dres
sing
co
olan
t
Tip
bre
akag
e se
nsin
g po
ka
yoke
Dre
ssin
g de
pth
of c
ut
Dre
ssin
g fr
eq.
Fee
d ra
te
Lack
of
oper
ator
eq
ualiz
atio
n
Roo
t cau
se
Coo
lant
sy
stem
s
Grin
ding
w
heel
Grin
ding
pr
ogra
m
Ope
rato
r
Sr.
No. 18 19 20 21 22 23
Ishikawa Diagram for Major defects:
37
Ishikawa diagrams (also called fishbone diagrams or causeandeffect diagrams) are diagrams that
show the causes of a certain event. Ishikawa diagrams were proposed by Kaoru Ishikawa in the
1960s, who pioneered quality management processes in the Kawasaki shipyards, and in the process
became one of the founding fathers of modern management. It was first used in the 1960s, and is
considered one of the seven basic tools of quality management, along with the histogram, Pareto
chart, check sheet, control chart, flowchart, and scatter diagram. It is known as a fishbone diagram
Causes in the diagram are often based on a certain set of causes, such as the 6 M's, described
below. Causeandeffect diagrams can reveal key relationships among various variables, and the
possible causes provide additional insight into process behavior. Causes in a typical diagram are
normally grouped into categories, the main ones of which are:
The 6 M's
Machine, Method, Materials, Maintenance, Man and Mother Nature (Environment): Note: a more
modern selection of categories is Equipment, Process, People, Materials, Environment, and
Management.
Causes should be derived from brainstorming sessions. Then causes should be sorted through
affinitygrouping to collect similar ideas together. These groups should then be labeled as categories
of the fishbone. They will typically be one of the traditional categories mentioned above but may be
something unique to our application of this tool. Causes should be specific, measurable, and
controllable. Most Ishikawa diagrams have a box at the right hand side, where the effect to be
examined is written. The main body of the diagram is a horizontal lines from which stem the general
causes, represented as "bones". These are drawn towards the lefthand side of the paper and are
each labeled with the causes to be investigated—often brainstormed beforehand—and based on the
major causes listed above.
Off each of the large bones there may be smaller bones highlighting more specific aspects of a certain
cause, and sometimes there may be a third level of bones or more. These can be found using the '5
Whys' technique. When the most probable causes have been identified, they are written in the box
along with the original effect. The more populated bones generally outline more influential factors, with
the opposite applying to bones with fewer "branches". Further analysis of the diagram can be
achieved with a Pareto chart.
38
Fig: Cause & Effect diagram for majority of defects
The Five elements of Fish bone diagram generated during Brainstorming session are:
Man:
Motivation less in workmen due to incentive less.
New operator working in area
Negligence during night shift
Lack of Awareness among operators
Machine:
Frequent Breakdowns, causing increase in vibration level
Detection of Defects is not effective
Coolant pressure varies abruptly
No Poka Yoke present to detect Drill breakage which causes ring formation
Material:
Tool quality not up to the mark, drill life less
Drill breakage due to drill overuse
In coming quality of parts not ok (Part bend which causes drill breakage)
Checking frequency is less
Method:
Gauges are not calibrated on daily basis
Elevator which lifts the part to chuck gets jammed causing part damage
Work instructions are over dated
Program corrections are complex during type change
Environment:
Machine is near to open window which causes dirt accumulation on part which damages
surface during grinding.
Bar chart
39
The ideas generated during Brainstorming session were verified by Process Experts and the causes
having positive impact on rejections were listed out. Bar chart analysis was performed on these
parameters to know the causes which have significant impact on rejections.
Causes & their contribution in Rejections
21
45
158 11
05
101520253035404550
Drill overuse No PokaYoke presentto detect Drill
breakage
Gauges notcalibrated on
time
Coolantpressure
varies
Others
Causes
% R
ejec
tio
ns
% wise causes
Fig 11: Bar Chart for Significant parameters
Chart clearly indicates that some system for early detection of Drill breakage needs to be
developed.
Causes & their contribution in Rejections
21
45
158 11
05
101520253035404550
Drill overuse No PokaYoke presentto detect Drill
breakage
Gauges notcalibrated on
time
Coolantpressure
varies
Others
Causes
% R
ejec
tio
ns
% wise causes
Fig: Bar chart for causes & their contribution
IMPROVE PHASE:
40
A) Detection of drill breakage on machine:
To reduce rejections which were caused by drill breakage, a new Laser sensor was installed on
machine and its feedback was given to PLC logic of machine. When tip of drill is Ok Laser falls on drill
& gets distracted, ensuring the machine to run continuously. This Tip Breakage Sensor (TBS) was
installed such that it overlaps with part loading, so change in cycle time due to Sensor installation is
zero.
Fig 12: Tool breakage sensing Poka Yoke with OK drill mounted on machine
Fig 13: Tool breakage sensing Poka Yoke when tip of drill is broken
After successfully implementing this on one pilot machine, there was horizontal deployment of this
Poka yoke on all 8 machines.
B) Drill overuse by operator:
When 5 why Analysis was done for this problem, it was found that the new drills were issues from
stores on monthly basis, so at the end of every month drill overuse was a common problem. It was
decided to topup drill shortage on every Saturday of week so as to maintain drill float on the line. Line
foremen were given clear instructions about drill records maintenance. Accurate drill
breakage/obsolescence is maintained and this point is added to Surprise audit committee.
C) Gauges & Microscopes are not calibrated on time:
For this cause a team of operators was formed to escalate the matter immediately when gauges are
not calibrated. Also calibration work was equally divided among quality people who calibrate gauges
once in three days.
D) Coolant pressure varies:
For this cause complete hydraulic circuit was checked for leakage. The team found that on Flow
control valve was faulty (worn out). The team insisted to change every valve of the circuit and
41
complete hydraulic circuit connections were changed with new one. Due to this major action the
leakage completely stopped. The coolant pressure variation problem is completely eliminated.
E) Others:
For all other causes following actions are taken
Window responsible for dirt accumulation was permanently closed & one exhaust fan was
installed at that place.
For new operator coming in area training sessions & supervision by skilled operators was
made compulsory.
Warning letters were issued for negligence from operators.
New & updated work instructions were put on machine boards.
CONTROL PHASE:
This phase defines control plans specifying process monitoring and corrective actions. It ensures that
the new process conditions are documented and monitored. All possible causes of specific identified
problems from the analysis phase were tackled in the control phase. Control solutions to identified
problems have been prepared in sequence to the improvements as explained above. This will prevent
the problems from recurring. The proposed control solutions to improve the previous solutions are
listed in sequence as follows.
A) Drill breakage Poka Yoke:
A Poka yoke monitoring sheet is maintained by shop. One shop Forman daily checks that all Poka
Yoke are working correctly & records it on a check sheet. A clear escalation model for problem
reporting is prepared for Poka Yoke failure.
B) Drill overuse by operators:
As weekly drill quantity topup is done, it automatically ensures that every week drill quantity is verified
for shortage. A record sheet is maintained to keep all drills records.
C) Gauges calibration:
This issue was taken seriously by quality department & they have assigned special audit team to
ensure that gauges are calibrated on time.
D) Coolant pressure:
For all hydraulic circuits in shop, one preventive maintenance program is prepared. Operators are
given authorities to stop machine if leakage is found on it.
E) Operator related issues:
All operator related issues were taken to Worker Union and after their consent it is decided to take
strict action against the operator negligence is company.
42
RESULTS:
After completing the DMAIC methodology of Six Sigma, again the process capability Analysis was
done to know the improvement in Sigma level. One month data on Control phase was taken for the
Analysis.
Figure 14: Process capability of seat visual process after applying DMAIC methodology
From the Minitab output it is clear that PPM defect level is reduced from 22,624 ppm to 11,031 ppm.
And Sigma level is Improved from 3.5σ to 3.79 σ.
Rejections in PPM
11039
22,624
0
5,000
10,000
15,000
20,000
25,000
PPM rejections before Project PPM rejections after DMAICProject
PP
M l
evel
Rejections in PPM
Fig 15: Results showing improvement in Sigma level of the process
A few more agreed recommendations are still to be implemented during plant shut down. The
estimated savings from the project after the implementation of all recommendations are expected to
be 1, 50,000Rs per Annum.
43
Sigma level3.5σ
Sigma level improved to 3.79σ
CONCLUSIONS:
The immediate goal of Six Sigma is defect reduction. Reduced defects lead to yield improvement;
higher yields improve customer satisfaction. The ultimate goal is enhanced net income. The money
saved is often the attention getter for senior executives. It has a process focus and aims to highlight
process improvement opportunities through systematic measurement. Six Sigma defect reduction is
intended to lead to cost reduction. Six sigma is a toolset, not a management system and can be used
in conjunction with other comprehensive quality standards present in the industry. The application of
Six Sigma technique for this project shows that company has taken a small step towards Six Sigma
Implementation on Company wide basis. Once Six Sigma finds its rightful place in the minds of higher
management, enormous gains can always be expected from its application. It is clear that the Six
Sigma methodology is highly beneficial to improve the performance of any manufacturing plant.
44
References
1) Kumar, P. (2002) “Six Sigma in manufacturing”, Productivity Journal, Vol. 43, No. 2, pp.196–
202.
2) Harry, M.J. and Schroeder, R. (1999) “Six Sigma: The Breakthrough Management Strategy
Revolutionizing the Worlds Top Corporations”, New York, NY: Double Day.
3) Henderson, K.M. and Evans, J.R. (2000) “Successful implementation of Six Sigma:
benchmarking: General Electric Company”, Benchmarking: An International Journal, Vol. 7,
No. 4, pp.260–281.
4) Mathew.H, Barth.B, and Sears.B, (2005) “Leveraging Six Sigma discipline to drive
improvement”, Int. J. Six Sigma and Competitive Advantage, Vol. 1, No. 2, pp.121–133.
5) Park, S.H. (2002) “Six Sigma for productivity improvement: Korean business corporations”,
Productivity Journal, Vol. 43, No. 2, pp.173–183.
45