Chapter 10 - Quality Improvement 2
Six Sigma and Process Improvement 2
CHAPTER 11Six Sigma and Process ImprovementTeaching NotesThis
chapter brings the Six Sigma concept into a sharp focus, and builds
on the need to integrate a performance management framework with
operational requirements in managing quality. In this chapter, we
introduce the statistical basis for Six Sigma, and outline the
requirements for Six Sigma implementation. This chapter also
extends the concepts of Chapter 10 on statistical thinking and
introduces the 7 QC Tools used for kaizen -- continuous improvement
-- Six Sigma and lean projects. Key objectives for this chapter
should be to assist students:
To fully understand the concept of breakthrough, defined as the
accomplishment of any improvement that takes an organization to
unprecedented levels of performance. Six Sigma projects often focus
on breakthrough improvements that add value to the organization and
its customers through systematic approaches to problem solving.
To become aware of the tools and techniques of Six Sigma which
represent a collection of quality improvement and statistical
methods and have been used successfully over the years in generic
TQM initiatives, ISO 9000, and in Baldrige processes. To learn that
a defect, or nonconformance, is any mistake or error that is passed
on to the customer. A unit of work is the output of a process or an
individual process step. A common measure of output quality is
defects per unit (DPU), computed as Number of defects
discovered/Number of units produced, and in Six Sigma metrics,
defects per million opportunities (dpmo) = DPU (
1,000,000/opportunities for error. A six-sigma quality level
corresponds to at most 3.4 dpmo which is equivalent to a process
variation equal to half of the design tolerance, while allowing the
mean to shift as much as 1.5 standard deviations from the target.
The sigma level can easily be calculated on a spreadsheet using the
Excel formula =NORMSINV(1 Number of Defects/Number of
Opportunities) + SHIFT or equivalently, =NORMSINV(1 dpmo/1,000,000)
+ SHIFT To learn about and practice problem solving -- correcting
deviations between what is happening and what should be happening.
Quality related problems often fall into five categories:
conformance problems, unstructured performance problems, efficiency
problems, product design problems, and process design problems. To
appreciate that a structured problem solving process provides
employees and teams with a common language and a set of tools to
communicate with each other. To develop understanding of the Six
Sigma stages of: 1) Define - the process of drilling down to a more
specific problem statement is sometimes called project scoping; 2)
Measure - collecting good data, observation, and careful listening;
3) Analyze - focuses on why defects, errors, or excessive variation
occur, and focuses on the root cause; 4) Improve - focuses on idea
generation, evaluation, and selection; 5) Control - focuses on how
to maintain the improvements.
To appreciate that projects are the vehicles that are used to
organize team efforts and to implement the DMAIC process. Being
able to manage a large portfolio of projects, as would be found in
Six Sigma environments, is vital to organizational success.
To study factors that should be considered when selecting Six
Sigma projects including: financial return, impacts on revenues and
market share, impacts on customers and organizational
effectiveness, probability of success, impacts on employees, and
strategic fit.
To introduce Six Sigma tools, in the light of two unique
features of DMAIC: its emphasis on customer requirements and the
disciplined use of statistical and other types of improvement
tools. Typical types of tools include elementary statistical tools,
advanced statistical tools, product design and reliability,
measurement, process control, process improvement, and
implementation and teamwork. To become familiar with, and learn to
apply the "seven QC tools" and related tools for quality problem
solving. The seven tools include flowcharts or process maps, run
charts, data sheets or check sheets, histograms, cause-and-effect
diagrams, Pareto diagrams, scatter diagrams and control charts. To
introduce the concept of lean production which refers to approaches
initially developed by the Toyota Motor Corporation that focus on
the elimination of waste in all forms, including defects requiring
rework, unnecessary processing steps, unnecessary movement of
materials or people, waiting time, excess inventory, and
overproduction. In service contexts, lean production is often
called lean enterprise. Some of the key tools used in lean
production are the 5Ss; visual controls; efficient layout and
standardized work; pull production; single minute exchange of dies
(SMED); total productive maintenance; source inspection; and
continuous improvement. To comprehend that tools and approaches
used in Six Sigma and lean production are different, yet
complementary. Lean is focused on efficiency by reducing waste and
improving process flow while Six Sigma is focused on effectiveness
by reducing errors and defects. Lean Six Sigma is a synthesis of
the best practices of both Six Sigma and lean production and has
gained considerable favor among practitioners in many
organizations. To appreciate that although Six Sigma was developed
in the manufacturing sector, it can easily be applied to a wide
variety of transactional, administrative, and service areas. Within
the service sector, Six Sigma is beginning to be called
transactional Six Sigma. However, differences between services and
manufacturing make opportunities in services more difficult to
identify, and projects more difficult to define. Small
organizations can use Six Sigma, although perhaps in a more
informal fashion.
The Instructors Resource section of the website accompanying
this text has a number of Baldrige video clips which give an inside
view of organizations that have received the Baldrige award. Some
of these, that are especially appropriate for this chapter, have
scenes that show how process improvement approaches can enhance an
organizations quest for world-class quality.
ANSWERS TO QUALITY IN PRACTICE KEY ISSUES
An Application of Six Sigma to Reduce Medical Errors
1.Process mapping was an essential early step for Froedtert
Hospital to take in order to identify the points at which failures
in the IV infusion process could possibly take place. When combined
with the FMEA, it provided clues as to where errors were generated,
their severity, and suggested the next steps to be taken for
measurement, in order to point the way toward control and
improvement.
2.The teams and task forces were multidisciplinary because the
processes crossed organizational boundaries. For example, the IVs
could be used in emergency rooms, surgical theatres, maternity
wards, etc. They would be of interest and concern to physicians,
nurses, pharmacists, and administrators, as pointed out in the
case. The approach has the benefit of bringing people with
different perspectives and expertise together to work on a complex
problem, which may not be fully understood by any one person or
smaller group of people.
Applying Quality Improvement Tools to an Order Fulfillment
Process
1.The DMAIC process from this chapter, and the Deming cycle and
the creative problem-solving process from an earlier chapter are
basically parallel statements of statistical thinking used to solve
problems.
Although not a perfect fit these processes can be compared in a
parallel fashion as follows:
Six SigmaDemingCreative Problem-solving
DefinePlanUnderstanding the "mess"
MeasureDoFinding facts
AnalyzeStudyIdentifying specific problems
Generating ideas
ImproveDeveloping solutions
ControlActImplementation
In the case study the team defined the problem (initially) as
discover ways to reduce order processing time so that at least 98%
of orders would be shipped on time (within 24 hours of receipt).
They then sought to understand the process by gathering facts the
data of times for processing 50 orders. They then analyzed the data
in a search for the causes for variation and select the most likely
cause. Once they discovered that order-picking was the bottleneck
operation, they could generate ideas and test them for
effectiveness in solving the problem. They found out that their
first solution did not work well. They had to use a longer cart,
rather than wider one in order to become more efficient in the
order-picking step in the process. Once they had tested their
improvement they undertook a longer study of effectiveness to prove
the solution.
2.Although not mentioned in the case, Deming would suggest that
the cycle could be repeated over and over for continuous
improvement. Therefore, if sufficient time had not been shaved off
the average time for picking each order, then additional analysis
might produce further methods improvements to reduce the time.
3.Alternatively, the packing process might be examined for
improvements. The same set of problem-solving steps should be used
to gather facts, analyze the facts, develop ideas for improvement,
test the proposed solution, and make it a part of the entire order
picking process.
BONUS MATERIALS - QUALITY IN PRACTICEImproving Patient Services
at Middletown Regional Hospital
Note: Data from Press Ganey surveys, mentioned in the body of
the case, are available in spreadsheet MRHCase.xls on the
instructors website for this chapter. Instructors may choose to
make these data available to the students for more detailed
analysis.Key Issues for Discussion
1.The Deming Cycle of Plan-Do-Study-Act is certainly covered by
the steps of Middletown Regional Hospitals Total Quality
Improvement Implementation System. Planning is done using steps 1-3
of MRHs process. Organizational awareness and environmental
transformation are used as preparation for identifying and defining
the problem. The Do part of the Deming Cycle is performed during
steps 4 and 5 of MRHs process. The Study phase of the Deming
process takes place in step 6, analysis, of MRHs process. The
Deming Act phase takes place in step 6, make recommendations, of
MRHs process. Follow-up on actions taken is the step 7 requirement
of MRHs process, to re-measure in order to assess improvement.
Coverage of Demings Act phase does not indicate whether or not a
pilot study is usually performed before action is taken to
implement the change systemwide.The DMAIC process may be an even
better fit for MRHs process. Define, Measure, Analyze, Improve, and
Control are covered by steps 1-3 of MRHs process (define); steps 4
and 5 (measure); step 6 (analyze and improve); and step 7
(control).
2.EVS has used checksheets (from the Press Ganey surveys),
control charts and correlation matrixes to analyze the problem
situation. This combination has provided more insights into the
nature of the problem and possible solutions than as single
technique would have done.
3.Another insight from analysis of graphs developed from the
Press Gainey surveys stems from the fact that there is very little
variability in results over the 11-quarter period. This indicates a
very stable process, but also indicates that changes that were made
had little impact. The root cause for not being able to raise
satisfaction levels still has not been found. Approaches to service
recovery should be explored, as well as after the fact
determination of quality deficiencies.
4.These three initiatives have helped to slightly reduce the gap
between managements goals and the actual customer perceptions. This
is what a control and improvement system is supposed to do.
However, the gap has not yet been completely closed.
5. Other approaches that EVS might try in order to close the gap
would be to benchmark the hotel industry (perhaps by visiting The
Ritz-Carlton Hotel Company), develop a fast feedback form that
would help identify strengths and OFIs and could be acted on while
the patient was still in the facility, and do in-depth interviews
with patients upon their discharge.
Six Sigma at National Semiconductor
1. An example of how National Semiconductor applied the DMAIC
process was at the South Portland, Maine, facility, which produces
advanced CMOS devices. The site saw an opportunity to improve
yields on a 0.35-m fabrication process by finding and eliminating
sources of variation in the manufacturing flow.
Define
The fab was experiencing losses at the final electrical test
(ET) operation. An 18-week baseline study of ET yield showed that
there was an opportunity for approximately 1.5 percent additional
yield for all wafers in this technology, meaning that they were
leaving more than $1 million in profit on the table. The goal of
eliminating the systematic yield loss caused by transistor leakage
and gate oxide breakdown was set. The team was also challenged to
establish a control plan for measurement of critical-to-function
parameters in the process flow that could be used to prevent
failing wafers from getting to the final ET step.
Measure
Analysis of the reasons for failed wafers at electrical test
identified three process modules that showed excessive variation in
the fabrication process. These modules included the spacer
formation, channel implants, and post-implant acid cleaning steps.
Other suspect areas were eliminated by performing commonality
studies on past events, discussing the issue with industry experts,
and re-evaluating experiments that had been previously run.
Analyze
Analysis revealed seven factors were determined to have an
effect on the three key failure modes. Utilizing the design of
experiments (DOE) function of JMP statistical software, the team
developed three experiments to determine which interactions had the
greatest effect on yield.
Improve
Prior to running the DOEs, the team had already collected enough
information to know that tighter controls were needed on certain
in-fab process parameters in the transistor modules. By taking
early action to put these controls in place, the team was able to
realize some yield improvements after only two months. Once the DOE
results came out, the team was able to validate the earlier actions
and fully characterize the process window for these critical steps.
In the improve step, the team leveraged the results of the DOEs
performed during the previous (analyze) step to make necessary
changes to specs, procedures, and equipment. These changes ensured
that variation during the spacer, implant, and acid cleaning steps
wouldnt affect yield at electrical test.
Control
To ensure that the process improvements developed by the team
became permanent, a control plan was created and transferred to the
manufacturing group, defining the requirements for every
critical-to-function parameter in terms of control charts, control
limits, sampling plans, gage capability, and out-of-control action
plans. By implementing improved process control methods, the team
was able to surpass the initial goal, provide Nationals designers
with a competitive advantage in their efforts to develop new
products for the analog marketplace, and increase yield by more
than 1.5 percent, with a projected annual savings of $1.1 million.
The project lasted about nine months.
2. National Semiconductor produces highly complex integrated
circuits that generally require clean-room technology to avoid
quality problems. With all of these complex processes, isolating
and reducing variables is a Herculean taskone that is perfect for
Six Sigma. National Semiconductor has a field defective rate of
less than 20 defective parts per million, but it still saw room for
improvement, particularly in the area of internal yield. Kamal
Aggarwal, executive V.P. of the CTMG, mandated that all CTMG units
would deploy Six Sigma.
It took a number of initiatives, in at least six stages, to
build momentum and establish the Six Sigma process. 1) A large
CTMG-wide continuous improvement (CI) effort yielded good results,
but it suffered from inconsistency. Each unit -- three wafer fabs,
three assembly sites, and the CTMG headquarters in Santa Clara,
California used its own methods to improve quality. 2) So National
Semiconductor invited GE Consulting to help with implementing Six
Sigma, but the idea met with resistance, because the employees felt
that they were already using Six Sigma tools and believed that they
were quite good at it. 3) Then, a year later, another consulting
firm that had experience in high-tech processes was brought in to
reintroduce Six Sigma. They convinced the senior V.P. of Nationals
plant in Melaka, Malaysia, to launch a pilot Six Sigma program with
10 projects. 4) Nine months later at the next CTMG Summit, a key
National Semiconductor supplier, DuPont Electronic Technologies,
made a presentation on how to quantify Six Sigma. This, coupled
with the results of the 10 Melaka pilot projects (initial savings
of nearly $900,000 and projected annual savings of nearly $2.3
million), convinced Aggarwal to implement Six Sigma across CTMG. 5)
The first wave of projects began in June 2003. Since then, 52
projects have been completed with an 84 percent success rate,
resulting in tens of millions of dollars in both hard and soft
savings. 6) In March 2004, CTMG launched its second wave of Six
Sigma programs, targeting similar savings from 44 projects. Four of
these projects target issues that exist on multiple sites, whereas
nine are aimed at customer satisfaction, quality, or safety
improvement.
Although not addressed by this case, chances are that National
Semiconductor is a high-performing company based on various
financial and quality measures (see Ch. 9). If that is the case,
research indicates that best practices should include: providing
customer-relationship training for new employees, emphasizing
quality and teamwork for senior management assessment, encouraging
widespread participation in quality meetings among non-management
employees, using world-class benchmarking, communicating strategic
plans to customers and suppliers, conducting after-sales service to
build customer loyalty, and emphasizing competitor-comparison
measures and customer satisfaction measures when developing plans.
The key to establishing the Six Sigma process at National
Semiconductor was executive support at several levels. It seems
likely that outstanding effort in the form of team participating in
quality meetings, world-class benchmarking, and communicating
strategic plans, relating to quality, to customers also played a
part in the eventual success of the Six Sigma initiative and
related projects.
Answers to Review Questions
1.A six sigma quality level corresponds to a process variation
equal to half of the design tolerance (in terms of the process
capability index, Cp = 2.0) while allowing the mean to shift as
much as 1.5 standard deviations from the target, which is the
manufacturing specification. The allowance of a shift in the
distribution is important, since no process can be maintained in
perfect control, due to natural variation.2.The recognized
benchmark for Six Sigma implementation is General Electric. GEs Six
Sigma problem solving approach (DMAIC) employs five phases:a)
Define (D)
i) Identify customers and their priorities.
ii) Identify a project suitable for Six Sigma efforts based on
business objectives as well as customer needs and feedback.
iii) Identify CTQs (critical to quality characteristics) that
the customer considers to have the most impact on quality.
b) Measure (M)
i) Determine how to measures the process and how is it
performing.
ii) Identify the key internal processes that influence CTQs and
measure the defects currently generated relative to those
processes
c) Analyze (A)
i) Determine the most likely causes of defects.
ii) Understand why defects are generated by identifying the key
variables that are most likely to create process variation.
d) Improve (I)
i) Identify means to remove the causes of the defects.
ii) Confirms the key variables and quantify their effects on the
CTQs.
iii) Identify the maximum acceptable ranges of the key variables
and a system for measuring deviations of the variables.
iv) Modify the process to stay within the acceptable range.
e) Control
i) Determine how to maintain the improvements.
ii) Put tools in place to ensure that the key variables remain
within the maximum acceptable ranges under the modified
process.
Note that this approach is similar to the other quality
improvement approaches we discussed and incorporates many of the
same ideas. The key difference is the emphasis placed on customer
requirements and the use of statistical tools and
methodologies.
3.There are significant differences between project selection
and problem definition. Project selection generally responds to
symptoms of a problem and usually results in a rather vague problem
statement. In project definition, one must describe the problem in
operational terms that facilitate further analysis. For example, a
firm might have a history of poor reliability of electric motors it
manufactures, resulting in a Six Sigma project to improve motor
reliability. A preliminary investigation of warranty and field
service repair data might suggest that the source of most problems
was brush wear, and more specifically, suggest a problem with brush
hardness variability. Thus, the problem might be defined as reduce
the variability of brush hardness. This process of drilling down to
a more specific problem statement is sometimes called project
scoping.
4.NCR Corporation defines root cause as that condition (or
interrelated set of conditions) having allowed or caused a defect
to occur, which once corrected properly, permanently prevents
recurrence of the defect in the same, or subsequent, product or
service generated by the process. As with a medical analogy,
eliminating symptoms of problems usually provides only temporary
relief; eliminating root causes provides long-term relief.
A useful approach for identifying the root cause is the 5 Why
technique. This approach forces one to redefine a problem statement
as a chain of causes and effects to identify the source of the
symptoms by asking Why? (ideally five times). In a classic example
at Toyota, a machine failed because a fuse blew. Replacing the fuse
would have been the obvious solution; however, this would have only
addressed the symptom of the real problem. Why did the fuse blow?
Because the bearing did not have adequate lubrication. Why? Because
the lubrication pump was not working properly. Why? Because the
pump axle was worn. Why? Because sludge seeped into the pump axle
this was the root cause. Toyota attached a strainer to the
lubricating pump to eliminate the sludge, thus correcting the
problem of the machine failure.
5.Several key principles are necessary for effective
implementation of Six Sigma:
a) Committed leadership from top management.
b) Integration with existing initiatives, business strategy, and
performance measurement.
c) Process thinking.
d) Disciplined customer and market intelligence gathering.
e) A bottom-line orientation.
f) Leadership in the trenches.
g) Continuous reinforcement and rewards.
6.The tools used in Six Sigma efforts have been around for a
long time. What is unique about Six Sigma is the integration of the
tools and methodology into management systems across the
organization. The topics covered may be categorized into seven
general groups:
a) Elementary statistical tools (basic statistics, statistical
thinking, hypothesis testing, correlation, simple regression)
b) Advanced statistical tools (design of experiments, analysis
of variance, multiple regression)
c) Product design and reliability (quality function deployment,
failure mode and effects analysis)
d) Measurement (process capability, measurement systems
analysis)
e) Process control (control plans, statistical process
control)
f) Process improvement (process improvement planning, process
mapping, mistake proofing)
g) Implementation and teamwork (organizational effectiveness,
team assessment, facilitation tools, team development)
7.Successful quality improvement depends on the ability to
identify and solve problems. According to Kepner and Tregoe, a
problem is a deviation between what should be happening and what
actually is happening that is important enough to make someone
think the deviation ought to be corrected. Problem solving is the
activity associated with changing the state of what is actually
happening to what should be happening.8.All quality problem-solving
can be classified into the following five categories:
a.Conformance problems are defined by unsatisfactory performance
by a well-specified system. Users are not happy with system
outputs, such as quality or customer service levels. Traditional
quality improvement tools and Six Sigma methods are often used
here.
b.Efficiency problems result from unsatisfactory performance
from the standpoint of stakeholders other than customers. Typical
examples are cost and productivity issues. Lean tools are often
used to address such problems.
c.Unstructured performance problems result from unsatisfactory
performance by a poorly specified system. That is, the task is
nonstandardized and not fully specified by procedures and
requirements. Unstructured problems require more creative
approaches to solving them.
d.Product design problems involve designing new products that
better satisfy user needsthe expectations of customers that matter
most to them. For such problems, Design for Six Sigma (see Chapter
12) tools and methods are applicable.
e.Process design problems involve designing new processes or
substantially revising existing processes. The challenge here is
determining process requirements, generating new process
alternatives, and linking these processes to customer needs. Work
systems design approaches discussed in Chapter 7 are typically
used.9.The Seven QC (quality control) Tools include flowcharts,
check sheets, histograms, Pareto diagrams, cause-and-effect
diagrams, scatter diagrams, and control charts. These tools support
quality improvement problem solving efforts.
In the CPS process, these tools will be useful in charting and
understanding messes. They can help in finding and recording facts
and in identifying specific problems. Once solution ideas have been
generated, these ideas can be added to the QC Tools to produce what
if we? scenarios. Problem solvers can use the tools in
cause-and-effect scenarios to figure out which solutions will work,
and which ones will not.
10.Flowcharts are best developed by having the people involved
in the processemployees, supervisors, managers, and
customersconstruct the flowchart. A facilitator provides
objectivity in resolving conflicts. The facilitator can guide the
discussion through questions such as What happens next?, Who makes
the decision at this point?, and What operation is performed at
this point? Quite often, the group does not universally agree on
the answers to these questions due to misconceptions about the
process itself or a lack of awareness of the big picture.
Flowcharts help all employees understand how they fit into a
process and who are their suppliers and customers. This realization
then leads to improved communication among all parties. Flowcharts
also help to pinpoint places where quality-related measurements
should be taken. Once a flowchart is constructed, it can be used to
identify quality problems as well as areas for productivity
improvement. Questions such as How does this operation affect the
customer?, Can we improve or even eliminate this operation? or
Should we control a critical quality characteristic at this point?
trigger the identification of opportunities.
11.A control chart is simply a run chart to which two horizontal
lines, called control limits are added: the upper control limit
(UCL) and lower control limit (LCL). Control limits are chosen
statistically so that there is a high probability (generally
greater than 0.99) that points will fall between these limits if
the process is in control. Control limits make it easier to
interpret patterns in a run chart and draw conclusions about the
state of control.
A run chart is a line graph in which data are plotted over time.
The vertical axis represents a measurement; the horizontal axis is
the time scale. Run charts show the performance and the variation
of a process or some quality or productivity indicator over time.
They can be used to track such things as production volume, costs,
and customer satisfaction indexes. Run charts summarize data in a
graphical fashion that is easy to understand and interpret,
identify process changes and trends over time, and show the effects
of corrective actions.
12.Check sheets are simple tools for data collection. Nearly any
kind of form may be used to collect data. Data sheets are simple
columnar or tabular forms used to record data. However, to generate
useful information from raw data, further processing generally is
necessary. Check sheets are special types of data collection forms
in which the results may be interpreted on the form directly
without additional processing.
In manufacturing, check sheets similar to Figure 11.9 are simple
to use and easily interpreted by shop personnel. Including
information such as specification limits makes the number of
nonconforming items easily observable and provides an immediate
indication of the quality of the process. For example, in Figure
11.9 a significant proportion of dimensions are clearly out of
specification, with a larger number on the high side than the low
side.
A second type of check sheet for defective items is illustrated
in Figure 11.10, which shows the type of defect and a tally in a
resin production plant. Such a check sheet can be extended to
include a time dimension so that data can be monitored and analyzed
over time and trends and patterns, if any, can be detected. Figure
11.11 shows an example of a defect location check sheet.
13.A histogram is a basic statistical tool that graphically
shows the frequency or number of observations of a particular value
or within a specified group. Histograms provide clues about the
characteristics of the parent population from which a sample is
taken. Patterns that would be difficult to see in an ordinary table
of numbers become apparent.
The Pareto principle was observed by Joseph Juran in 1950. Juran
found that most effects resulted from only a few causes. Pareto
analysis clearly separates the vital few from the trivial many and
provides direction for selecting projects for improvement. Pareto
analysis is often used to analyze data collected in check sheets. A
Pareto distribution is one in which the characteristics observed
are ordered from largest frequency to smallest. A Pareto diagram is
a histogram of the data from the largest frequency to the
smallest.14.The cause-and-effect diagram was introduced in Japan by
Kaoru Ishikawa. It is a simple, graphical method for presenting a
chain of causes and effects and for sorting out causes and
organizing relationships between variables. Because of its
structure, it is often called a fishbone diagram. The general
structure of a cause-and-effect diagram is shown in Figure 11.15.
At the end of the horizontal line, a problem is listed. Each branch
pointing into the main stem represents a possible cause. Branches
pointing to the causes are contributors to those causes. The
diagram identifies the most likely causes of a problem so that
further data collection and analysis can be carried out.15.Scatter
diagrams are the graphical component of regression analysis. While
they do not provide rigorous statistical analysis, they often point
to important relationships between variables, such as the
percentage of an ingredient in an alloy and the hardness of the
alloy. Typically, the variables in question represent possible
causes and effects obtained from Ishikawa diagrams.
16.Some of the tools and approaches used for lean operations in
organizations include: the 5Ss; visual controls; efficient layout
and standardized work; pull production; single minute exchange of
dies (SMED); total productive maintenance; source inspection; and
continuous improvement. Six Sigma is a useful and complementary
approach to lean production where the concepts might be used in
combination, in order to reduce cycle times, streamline an order
entry processes, or drill down to the root causes of the problems
and identify solutions. Because of their similarities, many
industry training programs and consultants have begun to focus on
Lean Six Sigma, drawing upon the best practices of both approaches.
Both are driven by customer requirements, focus on real dollar
savings, have the ability to make significant financial impacts on
the organization, and can be used in non-manufacturing
environments.
However, lean and Six Sigma concepts are different. They attack
different types of problems. Lean production addresses visible
problems in processes, for example, inventory, material flow, and
safety, while Six Sigma is more concerned with less visible
problems, for example, variation in performance. Another difference
is that lean tools are more intuitive and easier to apply by
anybody in the workplace, while many Six Sigma tools require
advanced training and expertise of Black Belt or Master Black Belt
specialists, or consultant equivalents. The concept of the 5Ss is
easier to grasp than statistical methods.
17.Small organizations are often confused and intimidated by the
size, costs, and extensive technical training they see in large
organizations that implement formal Six Sigma processes. Because of
this, they often they dont even try to adopt these approaches.
Small organizations are usually lean by necessity, but their
processes often operate at quality levels of two to three sigma,
and they are not even aware of it. Small organizations thinking
about adopting Six Sigma or lean production are advised to: Obtain
management commitment
Identify key processes and goals
Prioritize the improvement projects
Be systematic
Dont worry about training Black and Green Belts
Use just-in-time practices to learn the Six Sigma tools
necessary to successfully carry out specific projects
Communicate successes and reward and recognize performers
Small companies often need to bring in consultants for training
or improvement initiatives in the early stages of learning. This
can help to develop in-house expertise and put them on the right
track.
Discussion Questions
1.As with many other criticisms of quality improvement efforts,
this Fortune article focuses on what happens if organizations fail
to apply the basic quality concepts to so-called improvement
initiatives. Deming frequently said: There is no instant pudding.
in making quality work. Thus, the statements in the article are
largely bogus.
a. If Six Sigma has no noticeable impact on company financial
statements, its because not many home runs (large return) projects
have been done. One of the strengths of the Six Sigma concept is
that it documents financial impacts of the projects.
b. It is unclear why only early adopters can benefit, since
every organization has its own unique set of process and
opportunities will differ from any other organization. Thus any
organization should be able to benefit from Six Sigma process
improvements.
c. Yes, defects are more difficult to define and measure for
service organizations than for manufacturing, but numerous service
organizations (such as Baldrige winners) have proven that it is
possible and profitable to do so.
d. Six Sigma cannot guarantee that your product will have a
market, but nothing else can guarantee it, either, -- other than
continuous innovation and continuous improvement (of which Six
Sigma can be a vital part).
2. Six Sigma projects can be devised to improve activities and
processes in such areas as:
Sales and marketing projectsPerceived product value
Overall customer satisfactionSales force effectivenessComplaint
reductionGains and losses of customersCustomer
awards/recognitions
Supply chain management projects
Internal supplier quality measurements
Defect levels
Response time
Customer ratings of prod/service performance
Managing information technology projects
Internal information technology quality measurements
Defect levels
Response time
Customer ratings of service performance
Improvement of procurement process for software and hardware
Human resource projectsRoot cause analysis and reduction of
absenteeism
Root cause analysis and reduction of turnover
Measurement and improvement of employee satisfaction
Measurement and improvement of training effectiveness
Grievance reduction
Suggestion system improvement
Safety improvement
3.A set of CTQs that might influence overall service
satisfaction for service at an automobile dealership includes the
dimensions of service quality found in Chapter 4, as well as a
couple of related ones (timeliness and time):
Empathy ability to understand and empathize with the problems of
customers who may or may not have the technical know-how to explain
them in mechanical terms
Timeliness return of the car at the time promised is essential
to busy customers
Time the amount of time required is related to the size of the
repair bill
Assurance with limited technical proficiency, customers have to
feel comfortable relying on the expertise of the repair staff
Reliability customers wan the problems fixed right, the first
time, so that they dont have to keep returning, over and over, for
the same problem
Responsiveness customers would like to have answers as to how,
when and where a repair can be done, by a responsive service
representative
Tangibles last, but perhaps not least, customers would like to
see clean, attractive work areas, waiting areas, and toilet
facilities
4.Resistance to change is to be expected in introduction of a
new approach, such as Six Sigma, and plays a key part in successful
adoption of the concept. As pointed out in Chapter 6, keys to
overcoming resistance to change, more often held by managers than
by first-line employees, are: a) early involvement by all parties,
b) open and honest dialogue, and c) good planning. Managers must
believe in workers and their ability to contribute. Workers, of
course, must believe that managers will support the change to
empower them and help them to learn new skills required to be
effective contributors. Managers must also show commitment to the
practices of EI, such as training, rewards, and recognition. Thus,
adoption of Six Sigma cannot be successful in the long run without
supervisor and employee buy-in. 5.Processes that students might
encounter at their college or university due to part-time work on
campus might include accounting, budgeting, purchasing, training
and development, and research. Non-educational institutions they
might benchmark might be hospitals (accounting); relief agencies,
such as the Red Cross (budgeting); discount stores (purchasing);
consulting firms (training and development); and pharmaceutical
firms (research and development).
In addition students in this course might be assigned the
personal TQM project (see Chapter 1 for details) and asked to
identify a number of objectives for improvement. Some typical
objectives are get up on time (no snooze alarm), study chapters
before coming to class, eat no more than one junk food item out of
the vending machine each day, etc. Students might use the DMAIC
process in order to make improvements. To do so, they would need to
define the critical to quality characteristics that they desired
(such as those things that contribute to higher grades), decide how
to measure and analyze them (using a PTQM checksheet and scatter
diagrams, for example), and then deciding on required improvements
and a control process to hold the gains.
6. Companies today face incredible pressures to continually
improve the quality of their products while simultaneously reducing
costs, to meet ever-increasing legal and environmental
requirements, and to shorten product life cycles to meet changing
consumer needs and remain competitive. The ability to achieve these
goals depends on a large extent on product design. The complexity
of todays products makes design a difficult activity; a single
state-of-the-art integrated circuit may contain millions of
transistors and involve hundreds of manufacturing steps. Thus six
sigma quality levels are difficult to attain, due to multiple
variables that may affect quality. Nevertheless, improved designs
not only reduce costs, but also increase quality.7.The appropriate
tools to attack the problems would be:
a. If the significant causes of the copier jams are known, use a
flowchart to show the method for clearing each type of jam so that
users would know how to fix them.
b. Use a cause-effect diagram to identify the problem and to
focus in on the most significant one in the engineering
publications.
c. Gather data using a checksheet to identify the significant
causes of errors with the laser printer, and then use a Pareto
diagram to graph them.
d. Use a scatter diagram to see if the weights show a stable
pattern. If a weight reduction program is not needed and the
process is stable, plot the weight results on a control chart.
e. Use process simulation to determine the number of positions
of different types.
f. Use a scatter diagram and do a correlation analysis between
changes and dollar value, or changes and days between the request
for proposal and contract award.
g. Use a scatter diagram and do a regression analysis by time of
year, in order to predict staffing needs.
8.Lean concepts similar to those used in small businesses may
have applicability to classrooms, such as:
Obtain administrative commitment
Identify key processes and goals
Prioritize the improvement projects
Be systematic
Dont worry about training Black and Green Belts
Use just-in-time practices to learn the Six Sigma tools
necessary to successfully carry out specific projects
Communicate successes and reward and recognize performers
9.Ideally, the skills of technical experts (Green or Black
Belts) will complement those of team members (often called subject
matter experts, or SMEs). The two types of experts may be at odds
if they cannot agree ways to analyze problems, what their measures
show, and how to implement improvements and hold the gains through
appropriate control techniques. To prevent them from clashing in
such a way as to harm the results of the Six Sigma process, it is
useful to see that each has training and/or orientation to the
environmental factors, methods, and concepts used by the others.
Also, the project champion has responsibility to see that any
disputes are mediated and resolved in such a way as to enhance
project success. 10.The DMAIC process for a registration process
design/improvement should begin with customer needs and
expectations (expected quality), and end with what the customer
sees and believes the quality of the product to be (perceived
quality). These might include characteristics as follows:
Attributes
Technical Requirements
Convenience
Time, dates, internet, phone
Speed
Process standards
Costs
Fees
Accuracy
Error prevention
Empathy
Understanding/willingness of
personnel to solve problems
Expected quality needs to be considered in the Define stage
based on what the customer assumes will be received from the
process as a reflection of the customer's needs. The university
must focus on the key dimensions that are reflected in specific
customer needs. If these expectations are not identified correctly
or are misinterpreted, then the final product will not be perceived
to be of high quality by customers. For registration, this will
typically involve availability of classes, timeliness of the
process, time required to complete the process, etc.
Technical requirements determine the design quality of the
product. Process designers' must Measure and Analyze not only the
technical requirements for providing and registering students for
courses, but also perceptions of their needs, which may differ from
what is feasible. Other customers of the process also have some CTQ
issues that registration process designers must be aware of. For
instance, while the "average" student might need general courses,
curriculum majors may need in depth or specialized courses. If
registration process designers never have an opportunity to
interact with customers (students, academic department
administrators, faculty), the probability that they will not
understand or will misinterpret the expected delivery requirements
and expected quality is greatly increased.
After the registration process design is transferred to people
or organizational units responsible for delivering the service,
poor attention to customer needs can affect the perceived quality.
For example, if the system is not designed to assure conformance to
the technical specifications, then the actual quality produced may
not be the same as the design quality. The equation that relates
these different levels of quality is: perceived quality = actual
quality - expected quality. These characteristics must be taken
into account when the Improve and Control phases are completed.
Admission processes have many of the same process considerations
as registration. However, the needs of certain stakeholders, such
as potential students and parents of students, must also be taken
into account.11.To balance the who, what, where, when why, and how
of Six Sigma implementation, a manager, assisted by the team, must
evaluate them and select the most promising. This includes
confirming that the proposed solution will positively impact the
key process variables and the CTQs, and identifying the maximum
acceptable ranges of these variables. Since problem solutions often
entail technical or organizational changes, some sort of decision
or scoring model may be used to assess possible solutions against
important criteria. These may include: cost, time, quality
improvement potential, resources required, effects on supervisors
and workers, and barriers to implementation such as resistance to
change or organizational culture. To implement a solution
effectively, responsibility must be assigned to a person or a group
who will follow through on what must be done, where it will be
done, when it will be done, and how it will be done.
12.An argument can be made on either side of the question. On
the positive side, Welchs action could be taken as a clear signal
that Six Sigma was going to be vital process at G.E. for both
management development and continuous improvement. It is still
alive and well in G.E., as of this writing, almost 15 years later.
On the negative side, there are often unintended consequences of
threats such as these, as managers seek to carry out the letter of
such a mandate, rather than the spirit. 13.The problem of proposing
changes in one area that cancels out gains in the other area would
appear to be due to a lack of communication between the two project
teams. However, it may also have indicated a weakness in defining
the scope of the projects. The team leaders should have been aware
of the other project going on. They should then have coordinated on
boundaries so as to prevent overlap. Finally project champions
should have been consulted on how to either expand or break out the
projects before they were started, or soon afterward.
14.Phantom" capacity in service operations is the equivalent of
the hidden factory in manufacturing. By simplifying or eliminating
processes in banks or hospitals that require excessive waiting
time, maintenance, and rework, actual capacity could be increased
by 25 percent or more without investing in additional
equipment.SOLUTIONS TO PROBLEMS
1.Wellplace Insurance Company set a standard that policy
applications should be processed within three days of receipt. If,
out of a sample of 1,000 applications, 65 fail to meet this
requirement, at what sigma level is this process operating?ANS.The
defect rate is 65/1000 = 0.065. This is the same as: 0.065 X
1,000,000 = 65,000 dpmo. From Table 11.1, we see that this is
slightly better than 3 sigma with off centering of 1.5 sigma.
2.During one month, 35 preflight inspections were performed on
an airplane at Southstar Airlines. Nine nonconformances were noted.
Each inspection checks 30 items. What sigma level does Southstar
maintain if this incidence of nonconformance is typical of their
entire fleet of airplanes? ANS. To calculate the dpmo, we use 9/35
to get the number of defects per unit (DPUs). However, 30
opportunities per aircraft checked must be taken into
consideration, as shown, in order to calculate dpmo.
dpmo = (18/35) X 1,000,000/30 = 8571.4, which is less than 4
sigma with off centering of 1.5 sigma.3.Over the last year 1,054
injections were administered at the Fairhealth clinic. Quality is
measured by the proper amount of dosage as well as the correct
drug. In two instances, the incorrect amount was given, and in one
case, the wrong drug was given. At what sigma level is Fairhealths
process? ANS.We use 3/1054 to get the number of defects per unit
(DPUs). However, there are 2 opportunities per injection (wrong
drug, wrong dosage) to make an error. They must be considered, in
order to calculate dpmo.
dpmo = (3/1054) X 1,000,000/2 = 1423.1, which is slightly less
than 4.5 sigma with off centering of 1.5 sigma.
4.A few years back, the Wall Street Journal reported that about
750,000 airplane components are manufactured, machined, or
assembled for Boeing Co. by workers from the Seattle Lighthouse for
the Blind. A Boeing spokeswoman noted that the parts have an
exceptionally low rejection rate of one per thousand. At what sigma
level is this process operating?ANS. There is no indication of how
many opportunities for defects there are per component, so we will
have to assume that the defect rate is 1 per 1000 units produced.
Therefore, only 750 defective items (0.001 X 750,000) were
produced. To calculate dpmo, we see:dpmo = (1/1000) X 1,000,000 =
1000, which is slightly better than 4.5 sigma with off centering of
1.5 sigma.
5.Broadwork Electronics manufactures 500,000 circuit boards per
month. A random sample of 5,000 boards is inspected every week for
five characteristics. During a recent week, two defects were found
for one characteristic, and one defect each was found for the other
four characteristics. If these inspections produced defect counts
that were representative of the population, what is the overall
sigma level for this Broadwork process? What is the sigma level for
the characteristic that showed two defects?
ANS. To calculate the overall dpmo and sigma level, we have:
dpmo = (6/5000) X 1,000,000/5 = 240, which is approximately 5
sigma with off-centering of 1.5 sigma.
But for the one characteristic, we have:
dpmo = (2/5000) X 1,000,000 = 400, which is still good, but
somewhat less than 5 sigma with off centering of 1.5 sigma.
A Six Sigma project should be launched to determine root causes
for the defects from this one characteristic.6.Outsource
Microprocessor Corporation (OMC) sells 1500 specialized computer
processing chips each month at a price of $1200 each. Variable
costs amount to $1,000,000, and fixed costs are $400,000. Currently
the company has a defect rate of 8 percent (which are chips
returned by customers, scrapped by OMC, and replaced). Note that
the variable costs include the cost of producing the defective
chips.
a. What is the hidden cost to the company of making this rate of
defectives instead of 1500 good chips each month?
b. Suppose a Six Sigma effort can reduce the defects to a six
sigma level (assume for simplicity that the defective rate is
essentially zero). What is the impact on profitability?
ANS. In order to produce and sell 1,500 good computer chips, OMC
must start 1,500/0.92 = 1,631 chips into production. However, since
the variable cost of $1,000,000 includes the cost of making scrap,
the unit variable cost is therefore not $1,000,000/ 1,500 = $666.67
but $1,000,000/1,631 = $613.12. Thus the price paid for poor
quality, sometimes called the hidden factory, is 131 x $613.12 =
$80,319. This additional cost is incurred to make useless products
that cant be sold.
If a quality improvement initiative achieves a six sigma defect
level, the defective rate is essentially zero. This will remove the
variable cost of making the 131 defective units. The table below
shows that the $80,319 poor quality cost is eliminated from the
variable costs, and the saved money trickles falls to the bottom
line to increase profits. Thus, the profit increased to $480,319.
The 8% reduction in operational costs produced a 20% increase in
profit ($80,319/ $400,000).
Monthly BaselineMonthly Six Sigma Results
Sales1500 X $1200.00 =1800000
Sales1500 X $1200.00 =1800000
Variable Cost1631 X $ 613.12 =1000000
Variable Cost1500 X $ 613.12 = 919682
Contribution margin 800000
Contribution margin880319
Fixed cost
400000
Fixed cost400000
Net Profit
400000
Net Profit480319
Profit margin 0.222
Profit margin
0.267
7. A flowchart for a fast-food drive-through window is shown in
Figure 11.25 (see text). Determine the important quality
characteristics inherent in this process and suggest possible
improvements.
ANS. The important quality characteristics for this
drive-through window are: the machinery, materials, methods, and
people (manpower). The machinery must work well, e.g. most
important is the speaker system by which the order is transmitted
and received, the bell and its operating system must work well, the
menu sign must be readable and conveniently placed, the order
computer/cash register must be working properly to give the total
bill, and all the necessary equipment in the food preparation area
must also be working properly. The materials used in order taking
are few. However, the sign must be kept up-to-date with the latest
prices and selection of menu items. The method currently being used
is shown on the flowchart, and possible improvements are discussed
in the next paragraph. The people who take the order must be
trained to be courteous, friendly, accurate, and knowledgeable, or
the systems quality will suffer.
Possible improvements to the system might include installation
of a second window, so that the order is taken at the first window,
money is collected there, and the pickup is made at the second
window. A radio transmit/receive unit linking the customer at the
sign to the employee wearing a headset could increase the ability
of the employee to hear the order and to move around to assemble
the order while the customer is driving through. Automatic order
entry of standard selections might be built into the menu board
with push buttons (similar to an automated teller machine in a
drive-through banking operation). This would probably need to be
coupled with personal assistance from employees for special orders
via a speaker system.
8. The current process for fulfilling a room service request at
the Luxmark hotel can be described as follows. After the tray is
prepared at the room service station, the server proceeds to the
room, knocks on the door, sets up the meal, has the customer sign
the check, asks if anything else is needed, and then returns to the
room service station.
a. Draw a flowchart that describes this process.
b. From the perspective of creating a high level of customer
satisfaction from this experience, what improvements might you
suggest to enhance this process? Think creatively!ANS. 8 a.
Flowchart
8.b. A number of things could be done to enhance the service
experience and make it more memorable for the guest. Broadly, these
could be categorized as preparation, performance, and
leave-taking.
Preparation steps
Tray layout liner, flower, salt/pepper, silverware, tray
card
Food preparation salad, entree, condiments
Performance
Knock (exactly 3 times) and announce Room service.
Warm greeting and self-introduction (use names)
Permission to enter, enter, place tray, give a tour of the
meal
Provide weather report and forecast
Request guest to sign check (use name)
Leave-taking
Offer a wake-up call
Ask if anything else is needed; if yes, radio for it
Warm thanks (use name)
Exit room, return to service area
9. Placewrite, Inc., an independent outplacement service, helps
unemployed executives find jobs. One of the major activities of the
service is preparing resumes. Three word processors work at the
service typing resumes and cover letters. Together they handle
about 120 individual clients. Turnaround time for typing is
expected to be 24 hours. The word-processing operation begins with
clients placing work in the assigned word processors bin. When the
word processor picks up the work (in batches), it is logged in
using a time clock stamp, and the work is typed and printed. After
the batch is completed, the word processor returns the documents to
the clients bins, logs in the time delivered, and picks up new
work. A supervisor tries to balance the workload for the three word
processors. Lately, many of the clients have been complaining about
errors in their documentsmisspellings, missing lines, wrong
formatting, and so on. The supervisor has told the word processors
to be more careful, but the errors still persist.
a. Develop a cause-and-effect diagram that might clarify the
source of errors.
b. What tools might the supervisor use to study ways to reduce
the number of errors?
ANS. The C-E diagram, shown below, for this process analysis can
be found in cleaner format in spreadsheet Prob 11-09 on the Premier
website for this chapter.
EMPLOYEESPROCESSING METHOD
Experience
Number of clients per processor
Empowerment to contact client
Sequencing of work
Training on the equipmentError check method
Training on how to handle errors in copy
Instruction on resume writing
Format used by clients
Delivery of instructions to processor
CLIENT PROCEDURES
b. The supervisor might use flowcharts, checksheets and Pareto
analysis to classify the types of defects and their frequencies.
Then, training, crosschecking for errors, and work redesign might
be done in order to remove those error causes. Once the process is
under control, control charts might be used to hold the gains.
10.
A catalog order-filling process at Cats Catalog Company for
personalized printed products for pet owners can be described as
follows: Telephone orders are taken over a 12-hour period each day.
Orders are collected from each person at the end of the day and
checked for errors by the supervisor of the phone department,
usually the following morning. The supervisor does not send each
one-day batch of orders to the data processing department until
after 1:00 p.m. In the next stepdata processingorders are invoiced
in the one-day batches. Then they are printed and matched back to
the original orders. At this point, if the order is from a new
customer, it is sent to the person who did the customer
verification and setup of new customer accounts. This process must
be completed before the order can be invoiced. The next steporder
verification and proofreadingoccurs after invoicing is completed.
The orders, with invoices attached, are given to a person who
verifies that all required information is present and correct to
permit typesetting. If the verifier has any questions, they are
checked by computer or by calling the customer. Finally, the
completed orders are sent to the typesetting department of the
print shop.
a. Develop a flowchart for this process.
b. Identify opportunities for improving the quality of service
in this situation.
ANS. See flowchart, below, for the summary of the process at
Cats Catalog Company. The most serious problem from the standpoint
of customer service is the potential for a 12-hour delay before an
order reaches the supervisor for error checking, and another 3-4
hours may be required before entry into the computer. Obviously too
much checking and handling of the order occurred, and much of it
was many hours after the customer and order information had
originally been taken. Suggestions for improvement include: a)
processing small batches of orders (perhaps within 1-2 hours, or
less); b) building in error checking, perhaps through direct entry
of telephone orders into the computer; c) processing information
needed for customer verification and setup of new accounts at the
time the order is taken; d) having the phone department supervisor
simply audit or sample orders for errors; e) developing a
computerized method of matching orders and invoices, so that manual
verification is not required; generating an exceptions report after
step (e), with proofreading required for printing information that
cannot be computerized, if order verification and proofreading is a
vital step.
11.A Six Sigma analyst in Riverside United Bank suspected that
errors in counting and manually strapping cash into bundles were
related to the number of weeks that employees had been employed on
that job. The data found in the C11Data.xls file for Prob. 11-11 on
the student Premier website for this chapter were gathered from the
process. What do you conclude from your analysis? What do you
recommend?
ANS. The scatter diagram shows that the employees accuracy in
the strapping department of Riverside United Bank improves for
approximately the first 25 weeks. After that, it basically levels
off. The differences dont appear to be significant after about 30
weeks. (see spreadsheet Prob11-11.xls for details)
12.The times required for trainees in an electronics course at
Elecktronica Tech to assemble a component used in a computer were
measured. These are shown in the C11Data.xls file for Prob. 11-12
on the student Premier website for this chapter. Construct a
histogram to graphically show the data. What recommendations for
improvement would you give the course instructor, based on your
findings?
ANS. The histogram indicates that the assembly time for the
computer component in the electronics course at Elecktronica Tech
is concentrated in the two periods from 9 up to 15 minutes and from
16 up to 20 minutes, which shows a bi-modal distribution. The
highest frequency is for 17 minutes, representing the times of 20%
of the students. Many students (approximately 50 percent) appear to
be slower than the average (15.24 minutes) in the class. If the
assembly quality is the same for the slower group as it is for the
faster one, then the instructor should attempt to find the root
cause, by observing the methods of both groups, as well as testing
to see if there are any significant differences in abilities
between the groups. Then the techniques of the two groups may be
compared and the slower group members times may be reduced. (see
Prob11-12.xls for details)
13.The times required to prepare standard-size packages for
shipping at Pakman Shipping Company were measured. The packers were
divided into two equal groups of 20 people, each, having similar
experience in packing. These data are shown in the C11Data.xls file
for Prob. 11-13 on the student Premier website for this chapter.
Construct a scatter diagram for these data. What recommendations
for improvement would you give the section leader, based on your
findings?
ANS. The scatter diagram below (see spreadsheet Prob11-13) shows
the packing time for a standard size package for Pakman Shipping is
lowest for the first group of 20 packers, who average 13.85
minutes. In group 2, packers #22 and 37 are considerably slower
than the faster group members, such as #30 and #31. Still, the
overall packing time for a standard size package is higher for the
second group of 20 packers, who average 18.45 minutes. This
suggests that some workers, especially in group 1, are able to
perform the task much faster than the norm (mean of 16.15). If the
output quality is the same for the faster group, as well as the
slower one, then the production coordinator should attempt to find
the root cause, by observing the methods of both groups, as well as
testing to see if there are any significant differences in
abilities between the group members. If the methods used by the
first group can be taught to the slower group members, this could
increase productivity, reduce cost, and perhaps even improve
quality, simultaneously.
14.The data found in the C11Data.xls file for Prob. 11-14 on the
student Premier website for this chapter were gathered from a
process used to make PrintGear, Inc.s plastic gears for a computer
printer. The gears were designed to be 2.75 0.05 centimeters (cm)
in diameter. Construct a histogram based on the data given. What
can you observe about the shape of the distribution? What would you
recommend to the production manager, based on your analysis?
ANS. The histogram on the graph below (see spreadsheet
Prob11-14.xls for details) shows that, although the data are fairly
uniformly distributed, 18 points are above the upper specification
limit of 2.80 cm and 4 points are below the lower specification
limit of 2.70 cm. It is likely that the process needs to be
improved, with the first step being the removal of any special
causes. Also, attention should be given to reducing variation in
the process and centering it on the nominal dimension. A control
chart would have to be constructed and/or a process capability
study performed in order to get a fuller picture of the
process.
15.Ace Printing Company realized that they were losing customers
and orders due to various delays and errors. In order to get to the
root cause of the problem, they decided to track problems that
might be contributing to customer dissatisfaction. The list of the
problems found in the C11Data.xls file for Prob. 11-15 on the
student Premier website for this chapter shows their frequencies of
occurrence over a six-month period. What technique might you use to
graphically show the causes of customer dissatisfaction? What
recommendations could you make to reduce errors and increase
customer satisfaction?
ANS. It is obvious from the table and Pareto chart below that
the first two categories, accounting for 64.3% of the errors, need
improvement. See spreadsheet Prob. 11-15 for more detail
ACE PRINTING COMPANY
QUALITY ERRORS AND PERCENTAGES
PercentCumulative % Frequency
Setup delays35.00%35.00%245
No press time29.29%64.29%205
No paper11.43%75.71%80
Design delays10.00%85.71%70
Order info error5.57%91.29%39
Cust. chg, delays5.00%96.29%35
Lost order3.71%100.00%26
Total700
16.In an AcmeWidget, Inc. process, the production rate
(parts/hour) was thought to affect the number of defectives found
during a subsequent inspection. To test this theory, the production
rate was varied and the numbers of defects were collected for the
same batch sizes. The results can be found in the C11Data.xls file
for Prob. 11-16 on the student Premier website for this chapter.
Construct a scatter diagram for these data. What conclusions can
you reach?
ANS. The scatter diagram (see spreadsheet Prob11-16.xls for
details) for the AcmeWidget process, shows an interesting, and
counter-intuitive result. As the production rate increases, the
defect rate increases, then decreases. This could be because of the
"learning curve" effect in that as operators become more skilled
and familiar with the process and production runs are longer, the
defect rate can be improved.
17.The number of defects found in 25 samples of 100 Gamma Candy
Company lemon drops taken on a daily basis from a production line
over a five-week period is given in the C11Data.xls file for Prob.
11-17 on the student Premier website for this chapter. Plot these
data on a run chart, computing the average value (center line), but
ignoring the control limits. Do you suspect that any special causes
are present? Why?
ANS. The scatter diagram on the graph below (see spreadsheet
Prob11-17.xls for details) shows that samples 11 and 12 were
extremely distant from the average of 5.21. It is likely that an
assignable cause was present. Other than that, the process appears
to be stable, without excessive variations or trends up or
down.
18.Analysis of customer complaints at DOT.COM Apparel Company
revealed errors in five categories, such as billing, shipping, etc.
Data can be found in the C11Data.xls file for Prob. 11-18 on the
student Premier website for this chapter. Construct a Pareto
diagram for these data. What conclusions can you reach?
ANS. From the Pareto diagram, below (and spreadsheet
Prob11-18.xls), we can conclude that 54.2% of the problems are with
electronic charge errors and another 25.1% are due to shipping
errors, for a total in the top two categories of 79.3 %. These
categories should be improved first.
DOT.COM APPAREL HOUSE
QUALITY ERRORS AND PERCENTAGES
PercentCumulative %Frequency
Electronic charge errors54.20%54.20%5420
Shipping errors25.10%79.30%2510
Billing errors7.95%87.25%795
Delivery errors6.95%94.20%695
Long delays5.80%100.00%580
Total10000
19.A pharmaceutical company that manufactures individual
syringes is conducting a process capability study (see Chapter 13).
The data shown in the C11Data.xls file for Prob. 11-19 on the
student Premier website for this chapter represent the lengths of
35 consecutive samples. Plot these data on a run chart. Do the data
appear to come from a stable system so that a process capability
study may be conducted appropriately?
ANS. The data on the syringes, shown in the graph below (further
details can be found in spreadsheet Prob11-19.xls) show a
suspicious pattern that indicates that the process may be unstable.
Ten values, from samples 20 to 29 are alternating above and below
the average, indicating that some instability may be found in the
system, if it is carefully investigated.
20.The Monterey Fiesta Mexican Restaurant is trying to determine
whether sales of its popular Pan Con Mucho Sabor breadsticks are
correlated with the sales of margaritas. It has data on sales of
breadstick baskets and margaritas for 25 weeks, shown in the
C11Data.xls file for Prob. 11-20 on the student Premier website for
this chapter. Use the correlation utility, along with a scatter
diagram, in Microsoft Excel to analyze these data. What do they
indicate?
ANS. It can be seen on the correlation matrix (see Prob11-20.xls
for details) that breadstick orders are highly correlated with
numbers of margaritas. There is a 0.968 correlation coefficient.
The scatter diagram also shows a consistent pattern, as well.
Correlation Matrix
WeekOrdersMargaritas
Week1
Orders0.0711021
Margaritas-0.062910.9680711
Answers to Projects, Etc.1. This project is designed to help the
student to find how Six Sigma is viewed by various interested
parties and reflected on their websites. Dont be surprised to see
lack of agreement on concepts and definitions on the three
sites.
2.This project will take significant time to develop, but can
pay tremendous dividends in learning how to use the DMAIC process
and some useful quality tools for problem solving and improvement.
This would be a good term project for one or more students.
3. This project is designed to help the student to find which
Six Sigma or lean techniques are used in businesses to improve
their processes. Results will vary, but often are related to the
quality focus in the firm. Most companies now track some output
measures. Some Pareto charts and control charts may be found in
many firms. Dont expect to see cause-an-effect diagrams, scatter
diagrams, correlation and regression, or experimental design,
except in the most sophisticated quality-minded organizations (for
example, those with a Six Sigma program.)4. Results will vary. A
flowchart for exams should be prepared, based on the steps that
each student takes.
5. See the diagrams for a and b, below. The C-E diagram for c
will be similar.
Answers to Case QuestionsCase - LT, Inc.
1. There are a number of steps that could be taken to improve
the process. The Six Sigma team involved in this study recommended,
and then carried out, the following improvement steps. They
included:
Replaced the existing batch and queue system with a flow system.
Customer orders were processed one at a time as they arrived.
Individuals in each step were trained to do all activities in that
step. Automated the billing system and encouraged customers to
submit orders via the internet. Trained order processing employees
to take data correctly, verify data for accuracy and completeness,
and look for special terms. Created a system in which data was only
entered once and transmitted electronically, with passwords to
limit access.
Developed standard formats with built-in internal checks and fl
ags to alert users about incomplete data, incorrect part numbers
and faulty descriptions.
Trained shipping personnel to communicate with the billing
department in a timely manner regarding partial shipments and
returned goods.
Sent invoices only after receiving confirmation from shipping
personnel that orders were sent.
Printed the name and phone number of the customers contact
person, due dates and discounts on invoices to avoid runaround
experienced by customers. Highlighting payment due dates and
available discounts minimized customer calls and shrunk collection
intervals.
Simplified the chaotic pricing structure and computerized it.
Any deviations from the posted prices were immediately communicated
to the necessary parties.
Suggested blanket agreements be negotiated and invoices sent on
a monthly basis via the internet when possible. The free time that
resulted would be used for training, to pursue value adding
activities and to perform further kaizens.
2. Final actions development of metrics and next stepsThe Six
Sigma team developed a set of relevant performance measurements to
track the efficiency and effectiveness of the billing process. The
metrics selected were: Total number of billing documents processed
divided by the number processed per day.
Lead time to process customer order.
Cycle time to prepare a customer bill.
Average amount in accounts receivable divided by average number
of sales.
Percentage of accounts received past due.
Sum of the dollar amount in accounts receivable, multiplied by
the number of days since the sale, divided by the total number of
days.
Percentage of erroneous bills.
Average time to correct an erroneous bill.
Number of billing complaints.
Average time to resolve billing complaints.
The team was fully aware of the powerful nature of working in
cells and the success of just-in-time in manufacturing. For the
next level of improvement, the Six Sigma team suggested that LT
consolidate all activities into one cell manned by a team of
personnel from sales (covering order taking, order preparation,
order pricing and data processing), accounting (covering credit
check, sales tax and billing) and shipping. The cells would be
located in the shipping area, and cell personnel would take full
responsibility for all activities from order taking through
invoicing. Throughout the improvement project, the team
communicated with LT management and received managements
support.
Case Janson Medical Clinic
Analysis of the patient complaint data reveals that the three
top complaint categories are wait for doctor, ease of appointment,
and ease of phoning. All three of these are complicated to handle
because of the nature of the business. It does appear that the
telephone process is overly complex. The lack of empowerment of
receptionists appears to be part of the problem. In addition there
seems to be little telephone automation and no easy way of handling
routine calls. This might be improved by providing routine phone
options, such as ask about appointment date, ask about making an
appointment, and billing questions. See Pareto charts, cause-effect
diagram, and flowcharts, below. Also see spreadsheet C11JansonCase
in the instructor materials for more details.
1.The data, for the Pareto Diagram, in order, from highest
complaint level to lowest are:
Wait for doctor
13
Ease of appointment
12
Ease of phoning
10
Convenient hours
7
Courtesy of receptionist
7
Friendliness of phone receptionist 7
Responsive care via phone
5
Comfortable waiting
4
Physician listens
3
Respectful physician
2
Explanation of condition/treatment 2
Confidence in physicians ability 1
Time to register
1
Respect of nurses
0
Wait for doctor13
Ease of appointment12
Ease of phoning10
Convenient hours 7
Courtesy of receptionist 7
Friendliness of phone receptionist 5
Responsive care via phone 5
Comfortable waiting 4
Physician listens 3
Respectful physician 2
Explanation of condition/treatment 2
Confidence in physicians ability 1
Time to register 1
Respect of nurses 0
2. For the top three sources of patient dissatisfaction, the
following C-E diagram shows some possible causes. Note that the
answering method is closely related to the difficulty that patients
experience in quickly and easily obtaining appointments. The C-E
diagram, shown below, for this process analysis can be found in
cleaner format in spreadsheet JansonCase.xls on the website for
this chapter.
TELEPHONE ANSWERING METHODTELEPHONE EQUIPMENT
Lack of receptionist empowerment
Number of phone lines
Number of receptionists
Lack of telephone automation
Routine calls - same as emergencies
Inability to reach busy secretaries
Lack of refreshments in waiting area Excessive waiting time
Lengthy check-in process Magazines out of date or
unappealing
No information on length of wait
No TV in waiting area
WAITING
3Below are flowcharts that address some of the problems being
experienced with the current call answering and check-in process at
the clinic.
Proposed Patient Registration Process
Answers to Case QuestionsReadilunch Restaurant
1. What Carol has, here, is a mess. She must sort out the demand
pattern, the effects of time and day, if any, and the reasons for
customer dissatisfaction. Note that the analysis presented here is
not necessarily the only one for this complex set of data and
issues*.
The average number of available tables is slightly less than
3.5. Although the scatter diagram on the graph below (see
C11Readi1.xls on the instructor website for more details and
graphs) shows that the utilization of tables is the highest from
11:30 - 12:15, the results, are somewhat difficult to sort out.
Scatter diagrams (in C11Read1.xls, not shown here) for Mon.
-Tues. versus the one for Wed. - Fri. show some interesting
differences. The Tuesday and Friday patterns are similar to each
other in that tables appear to fill up earlier, and especially on
Friday, seem to stay filled longer. These two days appear to differ
from other days of the week and should be investigated further.
2. Analysis of the checksheet data on Complaints by Time may be
done by looking at totals and constructing line graphs to try to
discover patterns (see graph below and spreadsheet C11Readi2.xls on
the instructor website). Nothing jumps out of these views of the
data, although complaints about the long table wait peaks at 12:30.
The total complaints remain high all the way from 12:30 until 2:00
p.m.
3. Pareto analyses of the complaints by category are much more
revealing than the line graph shown above. This chart shows that
long table waits and long lines are the two top causes. However
these may both be related to inefficient service (cause 3).
Another Pareto diagram (in C11readi2.xls, not shown here) shows
that there are more complaints on Monday and fewest on Wednesday,
although the totals are not significantly different by day.4. Based
on the Pareto analysis it appears that the long wait for tables is
the most critical problem. Carol should investigate expanding the
restaurant. If that is not feasible, she should at least consider
the number of tables and their arrangement. She might also
investigate a system for taking reservations during peak
periods.
Since inefficient service (the third highest cause) may
contribute to slow turnover of tables, she should consider ways to
improve methods, including having wait people learn how to move
customers along in a friendly way, and having bussers trained to
quickly and efficiently clear tables.
To alleviate the problem of the long checkout line, Carol should
also consider either adding another cash register, or perhaps
placing the register at an exit door in such a way as to not
conflict with customers who are entering.
If you come up with an alternate interpretation that you like
better than this one, please forward it to Bill Lindsay
([email protected]). If we use it in the next edition, we will
acknowledge your contribution.
Bonus CasesCase - Implementing Six Sigma at GE-Fanuc
1. Jack Welch borrowed the initial vision of applying the six
sigma probability concept to operations at GE from Larry Bossidy,
the CEO of Allied Signal Corporation, who had initially borrowed it
from Motorola. Welchs vision was to deploy the methods and approach
down to the operating levels where it could be applied to numerous
processes. Thus, he specified that employees be trained in its
science and methodology so that they would be able to improve
efficiency and reduce variation in internal operations. This was
done, more than 100,000 people were trained, and over $1 Billion
has been saved since it was implemented.
2. Direct labor savings occur when a position can be eliminated
from a department (for example, if a dryer operator is no longer
needed because a dryer is removed). Labor cost avoidance savings
are really indirect savings because only part of the workers duties
are eliminated, thus freeing him/her to be used for other
productive duties. If there are no other duties to be performed the
worker must still be retained and paid for remaining tasks, even
though not fully productive.
3.The calculation can be verified by taking 1/7703 X 1,000,000 =
129.8 or 130 dpmo.
4. The improvement started with a feeling that the step of
testing the boards after running them through a high temperature
oven was unnecessary. However, to verify that feeling required
application of GEs systematic DMAIC problem-solving approach of
defining, measuring, analyzing, improving, and controlling the
process. After carefully defining the problem, Splauns team
(measured) collected information on costs, and test failures at
each point; analyzed what the impact of the testing was, improved
the operation, and monitored (controlled) it to ensure that there
were no unintended effects on quality. Future teams that hear of
this successful project can be assured of coming up with useful
results on their projects if they carefully follow the DMIAC steps,
as Splauns team did.Case - National Furniture
The process used by National Furniture for special orders
appears to have numerous points where orders can be misrouted or
can go wrong. The C-E diagram and process map, shown below, for
this process analysis can be found in cleaner format in spreadsheet
NatFurnCase.xls on the instructors website for this chapter.
1.Below is the process map for the current process. 2.Below is
the cause-effect diagram for why orders go wrong. Although it is
not shown, or discussed in the case, it seems obvious that things
could go wrong in the manager approval stage, or the front office
for the store. The office manager or store manager probably related
the information to the case writer!
3.The process map shows the steps in the process, while the
cause-effect diagram shows where things can and do go wrong. The
two can be used to improve the process. For example, the manager
could check the form for completeness before giving the associate
approval to process the order. The office manager could provide a
clear label on the bin where the form is to be placed for
processing, so that it is virtually mistakeproof. A Re-FAX stamp
could be placed on orders that were resubmitted after more
information was obtained. The office manager could help on
follow-up for customer information by having a logbook that lists
follow-ups to be done by order number and due dates for order
delivery.
National Furniture Store
C-E Diagram for Special Order Errors
SALES ASSOCIATE ORDER PREPARATION
Lack of complete information from customer
Incorrect FAX bin
Lack of additional customer info to office
No request for office to re-FAX order
Lost or misplaced order
Wrong merchandise ordered
Store not notified when order is due there
REGIONAL OFFICE
ORDER PROCESSING
National Furniture Store
Current Special Order Process
No
Yes
Case - Welz Business Machines
For the Welz Business Machines case problem, a comprehensive
analysis is required. The data and Pareto diagram, shown below, for
this process analysis can be found in cleaner format in spreadsheet
WelzCase.xls on the instructors website for this chapter.
Parts a. & b. See the cause and effect and Pareto diagrams,
below.
c. The cause and effect diagram and the Pareto analysis help to
determine where the problem of long telephone waiting may lie and
point the way to some possible solutions. First, the Pareto diagram
shows that the two major categories that account for most of the
customer problems are no operator (operators short-staffed) and no
receiver (receiving party not present). These reasons account for
73.1% of the customer complaints about long waits. The causes for
these problems for phone service representatives seem to be that
peak periods are not well staffed and there is no lunch break
coverage. For the "no receiver" category, the sales representatives
who are frequently out of the office seem to have no "back up"
people who stay to receive calls.
Possible solutions are to bring in extra phone representatives
for peak morning and afternoon periods, and either "stagger" lunch
breaks or bring in extra representatives, then. The sales
representatives, and/or their supervisors, should attempt to
develop cross-training, so that everyone in the office is equally
well-trained, sales representatives should be scheduled to be "on
call" in the office at certain times, and communications should be
set up in such a way that phone reps and sales reps work to
coordinate customer service and communications, perhaps even giving
the phone reps pager access to reach sales reps in the field. (See
the welzcase.xls spreadsheet for further details.)
WELZ BUSINESS MACHINES
TELEPHONE WAITING TIME REASONS
AND PERCENTAGES
PercentCumulative %Frequency
No operator51.34%51.34%172
No call rcvr.21.79%73.13%73
No understnd.18.21%91.34%61
Customer5.67%97.01%19
Other2.99%100.00%10
Total335
Prepare tray
Take tray to room
Knock, saying Room service
Enter
Set up meal
Request signature on check
Return to service area
Ask if anything else is needed
Patient dissatisfaction
Nurse gets chart and calls patient
Papers embedded in chart form
Receptionist pages nurse
Signature taken, if needed