Project Assignment for Lean Six Sigma Class of 2017
Industrial and Systems Engineering
&
Business management
Total Points: 100 and Due Back Date: 4/7/17
A company is going through a severe downturn in the business of making
automobile parts for global automobile manufacturing companies in India and abroad. It
produces aluminum pistons of mean 100 mm diameter with a tolerance band of
+/- 0.1 mm. The company has very strong manufacturing and quality control department
and historically took pride on its quality of products. Quality and its maintenance have
been the core focus of the company for last 25 years. Recently management has
changed to a new generation of executives led by the highly educated son of the
original owner of the company. During first six months in office, the new chairman
realized the business is not as it should be since the revenue has declined for last 5
years at a steady but alarming rate for the company. If such trend continues, the
company might be in its bankruptcy with 8-10 years. The chairman has heard some
good things about incorporating lean six sigma approach and went through a very quick
introductory course in one of the premiere institution. He realized its importance and
finally decided to apply such method as a facelift project for the company to turn around
in a very short time. Considering the fact he does not have time to look into every
details of the company, he decided to hire experienced personnel required to initiate the
project at the earliest possible time. Your Professor is the MBB for the project.
The lean six sigma project team was formed consisting of master black belt,
black belt and green belt team members. The team had its first meeting and went
through the vision, mission and core values of the company. The team formulated its
game plan for next 2 years to start with. The team has to come up with an execution
plan and decided to formalize it next 4-6 weeks at best.
As a first step, the team decided to develop a balanced scorecard for the
company in order to align the business activities to the strategy and monitor
performance against strategic goals of the company over time in next 10 years. Your
team was given the responsibility of developing the balance scorecard for the company.
For the balanced scorecard, your team decided to contain Financial, IBP, Customer and
Learning and growth aspects for the company while developing Vision and strategy for
the company for future. In each category, you decided to include Objective, Measure,
Targets and Initiatives for each factors in the balance scorecard.
1
The financial scores should connect financial strategies with the shareholders
with respect to growth, profitability and risk management. The customer strategy should
address the strategic goals to your customers with respect to creating values and
differentiation. The IBP strategic goals should answer questions with respect internal
business processes you employ on a day-to-day basis. IBP should consist of support
functions and generation of revenue. The learning growth portion of your balance
scorecard should answer how you sustain ability to change and grow. Each of the
values in the balance scorecards must be justified properly for your explanation to
higher executives during your first presentation. In order to identify the key performance
indicator of the company (KPI), BSC can be used as a strategic tool for mapping future
growth, mission and vision of the company.
Your MBB suggested that QFD is an excellent tool to combine multiple sets of
quantitative data and to quantify the C&E relationships between corporate goal and
individual process improvement efforts. For each of the BSC perspectives, a QFD
matrix was constructed. Recently, customers have expressed their dissatisfaction
mainly over the quality of the product. The team decided to look at the complaint at the
first level and then they expanded to second level of detail for each first level complaint.
The customer wanted friendly and knowledgeable staff, speed and accuracy of
answering the questions. The team came up with the measures for each one of these
customer needs and ranked them accordingly on a scale 1 to 10. The composite
scoring was determined. The team also looked at three competitors position against
these customer needs and included the same into the QFD chart. They also compared
the companys position against the best in the market by interviewing customers. These
benchmark findings were also included in the QFD chart. At the end, a single house of
quality linking customer needs to measurements with three vendors and the best
competitor was drawn and shared with the CEO. Your team was given the responsibility
of coming up with this assignment. You have to show and tail about this assignment
next week. Are you ready?
The team members decided that the primary cause of the customers
dissatisfaction was the quality of the products (piston) that have been put in the market
recently. The failure rate of these components, pistons, have changed substantially as
per the customer. While exploring the nature of failure rate of the product, they decided
to start from the testing results of the products. They decided to meet the test engineer
who is testing the recent products. The test engineer gave the failure data of a typical
product. Total number of components tested was 10,000. The output is given in Table 1.
In order to understand the nature of failure and failure rate, team decided to plot the
graph of failure rate vs. time in hours.
You have calculated the failure rate of the product at the give times of test
duration and also plotted the failure rate data against time. The customer expectation of
2
failure was 500 hours during the life of operation of the product. It appears from the plot
that the product is not satisfying customer wants or expectation. You have report the
deviation. Can you calculate the expected rate failure rate from the graph? If so, what is
the deviation that you are finding? Assuming the failure rate during first 200 hours is
denoted by exponential failure distribution, determine the mean and standard deviation
of the component. Once you have found the average failure rate for the component, you
have been asked to determine the probability that the component will fail within first 25
hours of operation. If the component has performed successfully for 200 hours, the
probability that it will fail during the next 25 hours. These are the values you need to
determine the effect of applying lean six sigma process to this system.
Table 1
Time of Observation, t, Hours Number of components operating
at time, t, hours
50 8880
100 8300
150 7918
200 7585
300 6968
400 6375
500 5808
600 5269
700 4760
750 4517
800 4237
900 3396
1000 2219
Next, the team decided to investigate the output of a CNC lathe that the
operators complained about its quality output. You measured physically the diameter of
the piston for a particular day of operation. The measurement data is given in Table 2.
The machine was set to produce the mean value of the piston diameters exactly at the
middle of the specified lower and upper value of tolerance limits. The machine was
supposed to follow six sigma production process to be precise. The output of the lathe
was assumed to follow normal distribution. Your first order of task was to calculate the
mean and standard deviation of the output for the day. Once you have found out the
mean and standard deviation of the process, you have to find out the probability of
producing a piston within tolerance limits of 100+0/-1, 100+3/-3, and 100+ 6 /-6.
Table-2
3
Piston Diameter, millimeters
Piston Sample Number Measured Diameter, inches
1 101.00
2 100.05
3 100.00
4 99.94
5 99.92
6 100.00
7 99.9
8 99.99
9 100.09
10 100.00
11 99.92
12 98.7
13 99.0
14 100.0
15 100.5
16 100
Accordingly, you also have to find out the rejection rate for these tolerance limits
for the piston diameters. In order to investigate the process further, you have been
asked to find the effects of the process not being centered in the tolerance range but
located at a distance of 1 from the center of tolerance range. Determine the probability
of producing a piston within specified tolerance limits for each of the three specifications
for the diameter of the piston mentioned above. Assuming the process is centered
within the tolerance limits for 6 sigma process and piston diameter specifications are
100+/- 0.1 mm, calculate the process capability or process capability index and find the
number of oversized pistons produced on the CNC lathe. Now, assuming the process
capability index of the lathe is 1.33, you have to determine the expected percentage of
oversized pistons produced. Assuming the process mean will shift by 1.5 over three
years of time in use, from the center of tolerance range of diameter of the pistons.
Next, In order to investigate the effect of the piston rings for the engine, you
decided to measure the fifteen samples, each of size 5, for the pistons as forged. The
internal diameter of the piston groove is measured accordingly and given in Table 3
below.
Table 3
Piston Groove dimensions
4
Sample Nr. Sample Values, millimeters
1 70.8 69.9 69.7 70.4 69.1
2 69.9 70.1 70.6 71.1 69.3
3 69.8 69.5 70.2 70.7 69.9
4 70.7 70.1 69.2 69.8 70.5
5 70 70.8 71.2 69.1 69.9
6 70.1 70.6 69.1 69.9 71.2
7 70.2 70.1 70.3 69.4 69.6
8 69.7 69.1 68.7 70.8 71.1
9 70.5 70.1 71.8 69.2 69.9
10 70.9 70.2 71.2 69.1 69.8
11 71.2 70.7 71 69.5 69.9
12 69.4 70.3 69.1 69.9 71
13 69.8 70.3 71.2 68.9 70.4
14 70.5 70.8 69.4 68.3 71.1
15 69.2 71.3 70.2 70.9 68.9
Establish the Xbar chart of the process. is the process under control? Your next
assignment is to plot the R-Chart for the process. Construct the R-chart for the process
also. Does the process how any abnormality in the production process? The piston was
assembled with the piston rings. The assemblies were examined for defects. The
number of defects found in the assemblies 1, 2, 3, 4, 5, 6, 7, 8, 9 and 10 were
4,6,19,9,3,10,8,5,7,6,2 and 5 respectively. You have to construct and draw the C-chart.
Is the assembly in control? Which assembly seems to be out of control? If so, do you
have to find the causes of the abnormal number of defects in this assembly?
Now, you have to find the risk control of this process. The null hypothesis of this
process is that process is under control for the assembly. Ou have to conduct a
hypothesis test. The test has a rule that indicates when to reject the null hypothesis
based on the sample data collected and measured. You have to find the value of an
acceptable value of probability of a Type I error i.e. value of . What would the value of
for a six sigma process. What would be the value of for a 2 sigma process?
The team decided to investigate the number of rejections in each day once the
piston and piston rings are assembled. You inspected a sample of 50 assemblies from
each days output. Based on the variety of attributes, you (quality inspector), classified
each assembly as acceptable or not depending on the number and nature of defects.
The experience over atypical 20-day period is summarized in Table 4.
Table 4
Rejections Noted
5
Date Number rejected Date Number Rejected
3/18 3 4/1 0
3/19 10 4/2 4
3/20 13 4/3 9
3/21 4 4/4 22
3/22 12 4/5 7
3/25 14 4/8 6
3/26 8 4/9 18
3/27 7 4/10 3
3/28 19 4/11 9
3/29 1 4/12 7
Even if the number of defects is binomial and p charts could be drawn on based
on exact binomial distribution, it is more common to use a normal approximation as your
interest is in estimating the value of p while tracking the random variable X/n. The
value of p is defined as the proportion of defectives produced and n as the number of
items sampled in each group on each day. A p chart is used to determine if there is a
significant shift in the true value of p. The expectation of X/n is p rather than X itself. For
assumed normal distribution for the process distribution, expectation is taken as mean,
, and , standard deviation is the variance of (X/n). The preliminary control chart
indicated that 4 points are out of control limits. You need not worry about the points that
fall below lower control limit, because that shows a better than expected rate of
defectives. . The production manager considers the points that are above upper control
limits. You agreed to it also. After investigation it was found out that a key employee
was absent for personal reasons. The high rate of defectives on those days was
apparently the result of a temporary employees lack of experience with media-plating
equipment. You termed it as assigned cause for the defect. You wanted to reconstruct
the p chart eliminating the points for the points beyond UCL. Is the process under
control after the effect of the assigned causes were eliminated? Answer your stand
giving further details.
To close this measurement process for the project, you also decided to construct
the c chart for the process using the data given in Table 5. You realized that c char is
based on Poisson distribution. If c represents the true mean number of defects in a unit
production, the likelihood that are there are k defects in the piston-ring assembly is
given by Poisson distribution.
The sample size remain the same at each inspection. Using these 20 data points
as a baseline, you have to determine UCL and LCL that include the universe of
6
observations with probability of 0.95. Also, you have to determine the control limits
resulting from using a normal distribution approximation of the Poisson distribution.
Table 5
Number defects observed
Unit Number Defects Unit Number Defects
Observed Observed
1 4 11 2
2 3 12 3
3 3 13 6
4 0 14 1
5 2 15 5
6 5 16 4
7 4 17 1
8 2 18 1
9 3 19 2
10 3 20 2
Now, your team would like to meet the MBB and BB to discuss the progress of
the project. Your team has decided to discuss the following reports:
Balanced Score card
House of Quality
Failure Rate and Bath Tub Curve
Statistical Details for the CNC lathe
Effect of Mean shift for CNC lathe
Process Capability Index for CNC lathe
Six Sigma Process details for CNC Lathe
Xbar, R, c and p Chart for the process
Null Hypothesis for the Charts
UCL and LCL for the part to be manufactured in lathe
Are you ready for the meeting?
