Six Sigma programs: An implementation model

Document Sample
Six Sigma programs: An implementation model Powered By Docstoc
					                                                        ARTICLE IN PRESS
                                                        Int. J. Production Economics 119 (2009) 1–16



                                                   Contents lists available at ScienceDirect


                                             Int. J. Production Economics
                                          journal homepage: www.elsevier.com/locate/ijpe




Six Sigma programs: An implementation model
Satya S. Chakravorty Ã
Department of Management and Entrepreneurship, Michael J. Coles College of Business, Kennesaw State University, 1000 Chastain Road, Kennesaw,
GA 30144-5591, USA




a r t i c l e in fo                                   abstract

Article history:                                      Despite the pervasiveness of Six Sigma program implementations, there is increasing
Received 19 April 2008                                concern about implementation failures. One reason many Six Sigma programs fail is
Accepted 15 January 2009                              because an implementation model on how to effectively guide the implementation of
Available online 25 January 2009
                                                      these programs is lacking. Using a successful Six Sigma program in a Network
Keywords:                                             Technology company, the purpose of this research is to develop an effective
Six Sigma                                             implementation model which consists of six steps. The first step is to perform strategic
Process improvement                                   analysis driven by the market and the customer. The second step is to establish a high-
Implementation model                                  level, cross-functional team to drive the improvement initiative. The third step is to
Case study
                                                      identify overall improvement tools. The fourth step is to perform high-level process
                                                      mapping and to prioritize improvement opportunities. The fifth step is to develop a
                                                      detailed plan for low-level improvement teams, and the sixth step is to implement,
                                                      document, and revise as needed. Important for both practitioners and academicians,
                                                      implications of our implementation experience along with directions for future research
                                                      are provided.
                                                                                                    & 2009 Elsevier B.V. All rights reserved.




1. Introduction                                                                Delphi Automotive have implemented Six Sigma programs
                                                                               (Treichler et al., 2002), and claimed that these programs
    Many characterize Six Sigma programs as the latest                         have transformed their organizations. Six Sigma
management fad of improvement tools and techniques                             programs are heavily promoted in practitioners’ books
(e.g., Watson, 2006). It is well known that Six Sigma                          on Six Sigma (e.g., Harry and Schroeder, 2000), and in
programs involve a host of critical decisions and many                         academicians’ books on Operations Management (e.g.,
researchers have contributed to the existing literature. For                   Jacobs and Chase, 2008). The American Society of Quality1
example, Schroeder et al. (2008) have identified many                           offers Six Sigma certifications; major corporations (e.g.,
critical decisions or elements of Six Sigma programs such                      General Electric Company, 2005) provide Six Sigma training,
as management involvement, improvement specialists,                            and a plethora of websites2 advertise Six Sigma solutions.
performance metrics, a systematic procedure, and project                          Despite the immense popularity and the wide-spread
selection and prioritization. Six Sigma programs improve                       adoption of Six Sigma, there is an increasing concern
operational performance in order to enhance customer                           across industries regarding the failure of Six Sigma
satisfaction with a company’s products and services                            programs. One reason many Six Sigma programs fails is
(Rajagopalan et al., 2004). Over the years, many compa-                        because an implementation model detailing the sequence
nies, such as General Electric, Allied Signal, Raytheon, and                   of Six Sigma elements/activities is not available. The
                                                                               existing literature identifies many elements of Six Sigma


  Ã Tel.: +1770 423 6582; fax: +1770 423 6606.
                                                                                   1
    E-mail addresses: satya_chakravorty@kennesaw.edu,                                  www.asq.org.
                                                                                   2
schakrav@kennesaw.edu.                                                                 www.isixsigma.com.

0925-5273/$ - see front matter & 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2009.01.003
                                                ARTICLE IN PRESS
2                                    S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16


which does enhance our understanding of Six Sigma                       facilitating Six Sigma implementation (Gitlow and Levine,
programs. However, the success of Six Sigma programs                    2005; Snee and Hoerl, 2003). Antony et al. (2007)
hinges on the sequence of many Six Sigma elements/                      emphasize management’s involvement in on-going pro-
activities or a model for implementation. To put it                     jects for sustainability of Six Sigma programs. Second,
differently, it is well known that one needs many                       improvement specialists are trained or hired at different
ingredients for Chicken Curry. However, in order to cook                Sig Sigma competency levels (e.g., Black Belt or Green
a delicious Chicken Curry, according to Jaffery (2003), the             Belt). Their primary responsibility is to provide technical
recipe requires a sequence of ingredients/activities (e.g.,             expertize and leadership in facilitating a specific Six Sigma
heat oil, fry onion and garlic, add Indian spices, put in               implementation (Pyzdek, 2003). Third, as Keller (2005)
chicken, pour water, and let it simmer). Any unreasonable               points out, Six Sigma programs have performance metrics
deviation from the recipe will lead to less than positive               and measurements based on cost, quality, and schedules.
experience. In the absence of a recipe or a model, it is not            Fourth, Six Sigma implementation uses a systematic
surprising that many implementations of Six Sigma                       procedure; a five-step DMAIC (Define, Measure, Analyze,
programs have failed. A survey of aerospace companies                   Improve, and Control) methodology. A detailed descrip-
concluded that less that 50% of the respondents were                    tion of DMAIC methodology can be found in Pyzdek
satisfied with their Six Sigma programs (Zimmerman and                   (2003) or Keller (2005). Fifth, project selection and
Weiss, 2005). Another survey of healthcare companies                    prioritization is an important element of Six Sigma
revealed that 54% do not intend to embrace Six Sigma                    programs. The prioritization of projects is determined by
programs (Feng and Manuel, 2007). Companies such as                     many criteria, such as a cost benefit analysis or the Pareto
3M and Home Depot were not satisfied with their                          priority index (Banuelas et al., 2006).
implementation of Six Sigma programs (Hindo, 2007;                          While we are familiar with many elements of Six Sigma
Hindo and Grow, 2007). Considering this, many authors                   programs, we lack an understanding of the sequence
question the return on investment of Six Sigma programs                 of these elements/activities, or a model for effectively
(e.g., Gupta, 2008). The real question is not whether Six               guiding the implementation of these programs. Because
Sigma programs have value, but why do so many Six                       there is no implementation model, practitioners have
Sigma programs fail? One reason many Six Sigma                          encountered tremendous difficulty in implementing these
programs fail is because we lack a model on how to                      programs, and there are reports of wide-spread Six Sigma
effectively guide the implementation of these programs                  failures. Zimmerman and Weiss (2005) found that less
(Wurtzel, 2008).                                                        than 50% of the survey respondents from aerospace
    Using a successful Six Sigma program in a Network                   companies expressed satisfaction with their Six Sigma
Technology company, the purpose of this research is to                  programs. Mullavey (2005) described the top 10 reasons
develop a model to effectively guide the implementation                 why Six Sigma implementations fail. Berg (2006) reported
of these programs. In the next section, we provide                      that their Six Sigma program was expensive and did not
the theoretical underpinnings of a Six Sigma implementa-                yield expected results. Concerned about Six Sigma’s
tion model. Following a description of our research                     problems, Sutton (2006) described nine ways to get the
methodology, we present our Six Sigma implementation                    best out of Six Sigma programs. A national survey of Six
experience. Then, we provide implications of our imple-                 Sigma programs in healthcare companies revealed that
mentation experience including directions for future                    54% do not intend to embrace Six Sigma programs (Feng
research. Finally, we provide the conclusion of our                     and Manuel, 2007). At 3M, a Six Sigma program that was
research.                                                               not correctly implemented almost stifled their creativity
                                                                        and innovation (Hindo, 2007). Home Depot’s Six Sigma
                                                                        program negatively affected employee performance, and
2. Literature review                                                    yielded Home Depot’s worst Consumer Satisfaction Index
                                                                        ranking (Hindo and Grow, 2007). Angel and Pritchard
2.1. Six Sigma                                                          (2008, p. 41) reported that ‘‘nearly 60% of all corporate Six
                                                                        Sigma initiatives fail to yield desired resultsy’’. According
   Over the years, many researchers have studied Six                    to Gupta (2008, p. 22), at times, Six Sigma ‘‘yimprove-
Sigma programs and identified many critical decisions of                 ment programs cost more than the improvement they
these programs. For example, see the previous research of               drive because of incorrect applicationy’’. While reporting
Antony and Banuelas (2002), Coronado and Antony                         cash flow problems of Six Sigma programs in small
(2002), Lakhavani (2003), Lynch et al. (2003), Mcadam                   companies, Foster (2007, p. 19) claims that if these
and Evans (2004), Gijo and Rao (2005), Szeto and Tsang                  programs are not ‘‘skillfully implemented; the benefits
(2005), Ladani et al. (2006), Savolainen and Haikonen                   of Six Sigma may be marginal’’. According to Chandra
(2007), and Davison and Al-Shaghana (2007). Recently, Zu                (2008), one reason Six Sigma programs fail is because
et al. (2008) studied the evolving theory of quality                    these programs are not correctly implemented.
management and the role of Six Sigma. While defining
Six Sigma programs and uncovering the underlying
theory, Schroeder et al. (2008) identified five elements                  2.2. Implementation model
of these programs. First is management’s involvement in
performing many Six Sigma functions, such as selecting                    As discussed before, a model for effectively guiding the
improvement specialists, identifying project selection, and             implementation of Six Sigma programs is not available.
                                                ARTICLE IN PRESS
                                     S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16                              3


The existing literature research related to Six Sigma and               step after securing management commitment and perso-
other improvement initiatives (e.g., Lean or Theory of                  nal involvement in a quality improvement initiative was
Constraints) are utilized to isolate steps of implementa-               to identify the tools for improvement. According to Keller
tion. Although suggested in different studies, these steps              (2005), Six Sigma programs have many tools for improve-
can connect with each other to hypothesize an imple-                    ment, which include Histograms, Pareto Charts, Statistical
mentation model. In describing a successful lean (e.g.,                 Process Control (SPC), and Analysis of Variance (ANOVA).
manufacturing cells) implementation, Chakravorty and                    Foster (2007) claimed that a common process for
Hales (2004) found that the first step in implementing an                implementing improvement tools is the DMAIC metho-
improvement plan was to perform a customer and market                   dology, which is similar to Edward Deming’s ‘‘Plan-Do-
driven strategic analysis. The purpose of this analysis was             Check-Act’’ problem solving approach. Lee-Mortimer
to direct the operational improvement effort to gain a                  (2006) considered the DMAIC methodology to be essential
competitive position in the market. Schonberger (2008)                  to Six Sigma programs and appropriate for delivering
points out that the objective of Six Sigma programs is to               business improvements. According to Chakravorty and
create a higher perceived value of the company’s products               Franza (2009), a form of DMAIC methodology, Define-
and services in the eyes of the customer. Antony et al.                 Measure-Analyze-Design-Verify (DMADV), was central to
(2005) indicated that linking Six Sigma to business                     a new product development experience. Mast and Bis-
strategy and customer needs is critical for successful                  gaard (2007) considered DMAIC methodology as the
implementation. According to Harry and Schroeder (2000,                 scientific method in Six Sigma programs. Keller (2005)
p. 39), Jack Welch said in GE’s 1997 annual meeting:                    provides a detailed description of each step of DMAIC
                                                                        methodology, and of various levels of training (e.g., Black
   y[Six Sigma implementation] begins not inside the                    Belt or Green Belt).
   business, but outside it, focused on answering the                       In order to implement Deming’s style of quality
   questions, ‘How can we make the customer more                        management, Hales and Chakravorty (2006) also found
   competitive? What is critical to the customer’s suc-                 that after identifying the tools for improvement to be
   cess?’ Learning the answer to the question and then                  used, the next step was to understand the overall
   learning how to provide the solution is the only focus               operations, and to set priorities for the project. One way
   we need.                                                             to understand overall operations is by developing a
                                                                        process map. According to Pyzdek (2003, p. 252):
    Andel (2007) claims that Six Sigma programs should be
implemented with a clear objective of improving compe-
                                                                            yA process map is graphic representation of a process,
titive positioning and of increasing the company’s value as
                                                                            showing the sequence of tasks using a modified
perceived by the customer. All activities related to Six
                                                                            version of standard flowcharting symbols. The map of
Sigma implementation should be approached from that
                                                                            a work process is a picture of how people do their
perspective.
                                                                            work. Work process maps are similar to road maps in
    In order to effectively manage bottleneck operations,
                                                                            that there are many alternative routes that will
Chakravorty and Atwater (2006) found that the second
                                                                            accomplish the objective. In any given circumstance,
step in that operation was to form a cross-functional team
                                                                            one route may be better than others. By creating a
to guide the implementation process. Pande et al. (2000)
                                                                            process map, the various alternatives are displayed and
point out that a cross-functional team is necessary to
                                                                            effective planning [to improve the process] is facili-
implement Six Sigma programs and the purpose of the
                                                                            tated.
team is to provide an on-going involvement of manage-
ment in the implementation process. According to Harry
and Linsenmann (2007), the CEO of DuPont committed                         According to Meredith and Mantel (2003), project
complete management support for implementing Six                        selection and prioritization is the process of evaluating
Sigma programs, and ensured that management learned                     projects, and then choosing to implement some set of
Six Sigma methodology by requiring that managers                        them so that the objectives of the organization are
themselves become Green Belt certified. At DuPont the                    achieved. Pande et al. (2000) claim that proper selection
Six Sigma program was not merely a methodology to get                   of Six Sigma projects and their prioritization could
results, but was a management culture created to ensure                 substantially improve the potential benefits of Six Sigma
long-term transformation of the business units. Gopal                   programs. Many authors have proposed a variety of
(2008) found that one reason Six Sigma implementation                   models of Six Sigma project selection, prioritization
failed in many companies was due to the lack of                         procedures, and tools (e.g., Breyfogle et al., 2001; Adams
commitment from management. Management simply                           et al., 2003; Kumar et al., 2008).
pushed Six Sigma programs out to employees, and did                        Finally, in describing a successful lean (e.g., manufac-
not become personally involved in the implementation                    turing cells) implementation, Chakravorty and Hales
process. As Mullavey (2005) points out, in order to                     (2004) describe a detailed plan to successfully orchestrate
successfully implement Six Sigma programs, management                   activities related to the implementation. The plan should
must understand Six Sigma methodology, must provide                     include a sequence of execution activities including lower-
leadership, and must guide the implementation process.                  level team improvement exercises, a training schedule, and a
    While implementing Deming’s style of quality man-                   detailed plan of all necessary changes related to the imple-
agement, Hales and Chakravorty (2006) found that the                    mentation. The plan provided a basis for implementation,
                                               ARTICLE IN PRESS
4                                   S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16


documentation, and revisions needed to assure contin-                  with participants and stake-holders such as managers,
uous improvement.                                                      engineers, and employees. We conducted the interviews
                                                                       in an open-ended nature, which implied that the
3. Methodology                                                         respondents provided objective opinions of the events as
                                                                       well as insights into certain occurrences. We followed the
3.1. Case study                                                        general guidelines provided by Fontana and Frey (2000).
                                                                       Fourth, qualitative data were collected in an observation
   Our case study was conducted in a regional provider of              mode involving many decisions during the implementa-
Network Technology services in the United States. These                tion. Fifth, since the researcher was involved with decision
regional centers provide provisioning, service assurance,              making during the implementation it was possible to
and customer care functions for all the products offered.              collect data in a participant-observation mode. Sixth,
While none of these centers create finished goods from                  additional quantitative data were collected for mistakes in
raw materials, the locations are often considered small                order entry, long setup times, and mistakes in network
internal factories. At the time of implementation, the                 transfer. These results provided clues to determine the
company held an estimated 5% of the domestic market.                   reasons for requests for manual activities which were
   According to Yin (2008) there are three reasons why                 occurring in the service operations. During the study the
a case study research methodology is appropriate for                   researcher kept a research log that documented each
our study. First, when ‘‘how’’ and ‘‘why’’ questions are               problem encountered during the implementation, in
investigated, the case study approach provides an ex-                  addition to the thoughts and insights gained during the
planation of linkages among events and their frequencies               process.
of occurrence. This cannot be done through simulation
or statistical models based on survey. Second, the case
study approach is preferred when a real world event is                 3.3. Data analysis
examined. Since many companies are actively engaged in
implementing Six Sigma programs, it is a natural way to                   The primary form of data analysis in case study is the
document emerging models from such an implementation                   reflection by the researcher on his own experience. The
environment. According to Stake (2000), real world                     researcher identified patterns and common themes by
studies are valuable for refining theory and suggesting                 analyzing the experiences of themselves and other
complexities for further investigation. Chakravorty and                participants. Content analysis worked well for identifying
Hales (2008) emphasize the need for real world based                   possible root causes and prioritizing alternative solutions.
research in order to help practicing managers on their                 In essence, content analysis is the counting of words,
journey to process excellence and prudent decision                     sentences, or ideas within categories of interest. In this
making. Third, the case study approach is appropriate                  study, we used it to collect ideas on how Six Sigma was
because this approach makes use of variety of evidence,                implemented. Three general guidelines are applicable to
including documents, archival records, interviews, and                 content analysis. First, two judges were used for perform-
direct observation. According to Meredith (1998, p. 443):              ing the analysis so that the consistency of results could be
                                                                       estimated. Second, the categories of interest must be
    ythe importance of direct observation [first source
                                                                       applicable to the research objectives. In this study,
    (seeing it oneself) rather than second (speaking or
                                                                       we collected data specifically on the implementation
    writing to someone who saw or experienced it) or
                                                                       and use of Six Sigma. Third, the units of analysis must
    third, or sometimes no source at all], the role of the
                                                                       be appropriate for representing the topic under examina-
    context in which the phenomenon is occurring, and the
                                                                       tion (Rosenthal and Rosnow, 1991). In this study, the unit
    dynamics of the temporal dimension through which
                                                                       of analysis was the operational/department level where
    the events of the phenomenon unfold help to under-
                                                                       Six Sigma was designed to be used.
    stand the how and why elements of the phenomenon.
                                                                          To examine the flow of information through the
                                                                       system; we employed process mapping. Each activity in
3.2. Data collection                                                   a business process is shown on a two-dimensional scale.
                                                                       The processes are then connected with arrows showing
    Multiple sources of evidence were used to validate                 the direction of service flows. Average waiting times and
data. Yin (2008) identifies six major sources of evidence.              processing times were estimated for each stage of the
We employed all six in this study. First, qualitative data             process. These maps helped identify where services had
were collected through documentation obtained in the                   breakdowns or long wait times and were processed
form of letters, memoranda, minutes of the meeting,                    through redundant or unnecessary activities.
progress reports, and strategic planning reports, etc.                    Each evening the researcher reviewed the information
Second, quantitative data were collected in the form of                and data for the current day. They spent several hours
archival records of financial data, customer complaint                  brainstorming ideas and reflecting on what had happened
reports, ordering processing, quality reports, purchase                during the day’s events with several of the management.
orders, operational data (such as personnel utilization),              They grouped the ideas using content analysis and
routing information, performance measurements (such as                 mapped the order processes which would be addressed
annual sales and responsiveness). Third, additional quali-             the next day. The ideas with the greatest potential for
tative data were collected through extensive interviews                success and most support from the team responsible for
                                                ARTICLE IN PRESS
                                     S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16                             5


changes or from other company executives were prior-                    occurred and manual intervention was required, some
itized for implementation. Results of the ideas implemen-               of the benefits of the system were lost due to costs
ted during the current day were reviewed for                            associated with manual repair. If the RMAs were reduced,
performance. Those that provided improvement were                       these expenses would be avoided and customer satisfac-
noted, as well as our preliminary insights into why they                tion would be improved by reducing the circuit provision-
worked. Other areas in the order process that could                     ing interval. Therefore, it was very important to the BNISC
benefit from the same solutions were identified for                       to reduce RMAs to decrease the variability, meet or exceed
implementation on the following day. Each morning,                      customer expectations, and achieve the original purpose
the team chose the next problems to attack based on                     and savings of deploying an automated solution.
consensus. The planned solutions were implemented                           The company adopted the Six Sigma program and the
during the first shift operations of the company. During                 Chief Broadband Officer (CBO) championed the program.
this period, the researcher focused solely on assisting with            The program was developed in cooperation with internal
the implementation efforts and documenting the process.                 resources and by levering best practices from the private
These processes occurred on an iterative basis throughout               industry. The program included application of statistical
the study.                                                              methods to bring rigor into daily business decision
                                                                        making. The program was an extensive deployment
3.4. Limitation of case study approach                                  within the company as a first attempt to reduce variation
                                                                        or a form of waste throughout the company. The CBO was
                                                                        aware that a company-wide improvement strategy ex-
    There are three limitations of case study based
                                                                        tending over several months had to be developed. While
research. First, case study fundamentally assumes that
                                                                        the team had considerable engineering experience and
an espoused theory should adequately specify action,
                                                                        qualifications, cumulatively they did not have the experi-
which is rarely the case (Kemmis and McTaggart, 2000).
                                                                        ence or the background to deploy a Six Sigma implemen-
Rather, the best theories are parsimonious and do not
                                                                        tation program and execute the program throughout the
claim to replicate reality. Second, the participants are
                                                                        company.
exposed initially to the ideas of the researcher which can
                                                                            The CBO facilitated improvement efforts and estab-
create bias when identifying root causes or explaining
                                                                        lished change objectives. The researcher-client agreement
events. Three, the conclusions from a single study may
                                                                        stipulated: (1) that the client was committed to improving
have limited generalizability, and therefore, contributing
                                                                        existing order processes and was willing to provide all
little to developing or informing a theory. Other research-
                                                                        pertinent data to the researcher, (2) that the researcher
ers are encouraged to test these findings by conducting
                                                                        would document the application improvement solutions
further research in other environments.
                                                                        in a research log and provide findings to managers as
                                                                        requested, (3) that the researcher would provide training
4. Description of implementation experience                             and participate in improvement efforts in exchange for
                                                                        publishing the results, and (4) that management would be
4.1. Perform strategic analysis                                         available for at least two hours per day to reflect and
                                                                        provide feedback to the researcher. Following the initial
   When customers use their Automated Teller Machine                    discussion between the researcher and management, a
(ATM) card, several transactions need to be completed                   cross-function team was established.
simultaneously. These transactions are completed by
Broadband Network Infrastructure Support Center (BNISC)
by performing provisioning translations for ATM custo-                  4.2. Form cross-functional improvement team
mers. These translations help complete the transactions
quickly and include assigning logical layer (software)                      With approval from the CBO, a cross-functional
information for use with the physical layer (equipment) of              team was formed, consisting of managers, engineers, and
the desired circuit. In an effort to improve operational                consultants. This team was called the ‘‘Management
efficiency, an automated system was designed and                         System of Operational Change’’ (MSOC). While leadership
implemented entitled ‘‘Broadband Network Management                     skills were considered, eight members of MSOC were
System’’ (BBNMS). This internally developed application                 primarily chosen for their technical skills. This included
was designed to automate the flow of information from                    participants from the customer service organization (non-
the customer’s service order through the provisioning                   access), the circuit fulfillment group (access), and the
process. This was a completely manual process prior to                  application development team that originally designed
implementation. Thus, the primary intent of the BBNMS                   BBNMS. By definition, ‘‘access’’ orders were only those
was to eliminate manual work, improve operational                       circuits that require any type of long distance connection
efficiencies, and improve order accuracy.                                which crosses one or more footprints and do not
   One of the features of the BBNMS system was a                        require a long distance connection. The BNISC had the
functionality set known as ‘‘Request for Manual Activity’’              responsibility to prepare translations for both access/non-
(RMA). An RMA occurred when an order failed to                          access ‘‘flow-through’’ and access/non-access ‘‘non-flow-
automatically flow through the provisioning system. This                 through’’ orders. MSOC decided that the ‘‘non-flow
indicated that an error occurred somewhere in the                       through’’ orders impacted circuits managed out BBNMS
process, thereby indicating a defect. Any time a defect                 and were therefore not included as a part of initial Six
                                                 ARTICLE IN PRESS
6                                     S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16


Sigma implementation (deployment). Initially, Six Sigma                  improvement ideas, such as Lean and Theory of Con-
implementation was to identify ‘‘flow through’’ access and                straints principles, were included to enrich the training
non-access orders that generated one or more RMAs,                       programs (Appendix B). To reinforce understanding
measure the RMAs and their causes, and reduce/eliminate                  among managers, engineers, and associates in the appli-
RMAs using the Six Sigma tool kit.                                       cation of DMAIC methodology, we developed examples of
   MSOC was responsible for developing a charter for all                 DMAIC applications, and supervised many sessions where
the improvement initiatives, a timeline for carrying out                 the managers and the workers applied the methodology
the implementation, and a budget to support the changes.                 to reduce variation or waste from the process.
MSOC was also responsible for analyzing the existing
process, planning the change with active participation
                                                                         4.4. Execute high-level process mapping and
from all departments, implementing the change in
                                                                         prioritize improvement
process, establishing performance matrices, and following
up the effects on performance. MSOC began their analysis
                                                                             MSOC performed a high level process mapping of
by interviewing personnel (customers, engineers, and
                                                                         information flow with the objective of identifying and
technicians) from different departments in existing pro-
                                                                         prioritizing improvement opportunities. This was de-
cess and by examining customer complaint reports, cost
                                                                         signed to understand the existing flow of information
reports, types of RMAs, etc. MSOC initial assessment was
                                                                         and its associated value creation. While performing
that the deteriorating performance was due to variation in
                                                                         process mapping, the MSOC discovered numerous exam-
the process and/or many wasteful activities at each stage
                                                                         ples of variation (wasteful) activities throughout the
of the existing process. The company was determined to
                                                                         entire order process. Specifically, MSOC found outdated
find a long-term solution that would improve their
                                                                         information in the systems database, redundant proce-
operations and satisfy their customers. The main objec-
                                                                         dure, duplicate and confusing paper work, long setup
tives were to improve customer service by eliminating
                                                                         times for the network, mistakes in order entry, mistakes in
waste from the existing process.
                                                                         network transfer, etc. The MSOC could not effectively
                                                                         address all of the problems simultaneously and struggled
                                                                         deciding which ones to address first. MSOC documented
4.3. Choose improvement tools
                                                                         many wasteful activities from different departments and
                                                                         segregated them using content analysis. After several
    MSOC deliberated to adopt an improvement tool
                                                                         iterations, they were able to agree to segregate the
and an instrument to document and communicate the
                                                                         problems based on possible outcomes, i.e. cost savings
implementation of process improvement ideas. The
                                                                         and/or increased revenue generation. When applicable,
company adopted Six Sigma process improvement in-
                                                                         the potential cost savings were computed in terms of
itiatives and was keen on a unified deployment of
                                                                         reduction in duplicate activities, reduction in returns,
improvement initiative. In the past, the company had
                                                                         reduction in payroll, etc. When applicable, the revenue
initiated many process improvement or change programs
                                                                         generation ideas were calculated in terms of increased
(e.g., Total Quality Management, Business Process Reen-
                                                                         sales, faster response time, and better utilization of
gineering, Team Building, etc.). One problem with these
                                                                         network bottlenecks. Using the two potential outcomes,
initiatives was that they were isolated in different
                                                                         MSOC prioritized Six Sigma variation (waste) reduction
departments and rarely worked together to improve the
                                                                         projects based on those with the strongest support from
performance or the overall value creation process. One
                                                                         the content analysis (i.e. those that were mentioned
objective of this implementation was to integrate all
                                                                         most by engineers and managers), keeping in mind the
of these strategies into one simple core strategy to drive
                                                                         company’s long-term objectives. MSOC then held a meet-
operational excellence from the bottom. MSOC recognized
                                                                         ing with all managers, including the President, for their
the importance of statistical tools such as Histograms,
                                                                         input and approval. After deliberations, MSOC developed
Pareto Charts, Statistical Process Control, Regression, and
                                                                         the final priority of all Six Sigma projects to reduce
Design of Experiments (DOE), etc. The general census
                                                                         variation from the process.
was that the vast majority of the engineers did possess
requisite analytical and/or mathematical skills to fully
understand and implement such tools.                                     4.5. Develop detailed implementation plan
    MSOC worked diligently to develop the contents of
training program with illustrations, examples, and inter-                   Using the implementation priority, a detailed imple-
active statistical software to provide training in the basics            mentation plan consisting of six major activities was
of process improvement. The researcher recommended                       developed. First, each department was designated to
that Six Sigma training be provided in four different levels             setup a lower level improvement team from the depart-
(Appendix A). First level was designed for executive                     ment. After meeting and discussing with the supervisor, a
overview of Six Sigma for upper level management and                     team (6–10 engineers and associates) of shop workers
engineers for one day. Second level was designed for                     (and/or process owners) was established to carry out the
directors and managers working directly as champions of                  implementation. The entire company was divided into
Six Sigma projects. Third level was designed at the Green                30 teams. Second, once teams were identified, the next
Belt level, and fourth level was designed at the Black Belt              step was to select supervisors or champions for managing
level. In addition, several concepts from popular process                the teams during implementation. MSOC spent inordinate
                                                ARTICLE IN PRESS
                                     S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16                              7


amounts of time selecting supervisors and champions.                    associate’s questions openly and honestly and solicited
MSOC believed that the success of the implementation                    suggestions to improve the process. After several weeks,
was dependent on the champions and their ability to                     MSOC and the workers were freely exchanging ideas and
manage the teams. While technical skills of champions                   working with greater cohesiveness.
were required, their human skills were more important
for this role. They came with varying levels of experience
(5–25 years), professional qualifications (e.g., Electronics             4.6. Implement, document, and revise
Engineering to Human Resources to Sales or Accounting),
and education (e.g., no college degree to Masters in                        While implementing, MSOC encountered two major
Engineering Management). Following many days of                         findings that impacted the implementation of Six Sigma
deliberations, MSOC selected 10 champions or supervisors                and its DMAIC methodology. First, when the problem
to manage 30 teams (average three teams per champion).                  was easy and clearly identified, DMAIC proceeded in a
After establishing communication protocol with MSOC,                    sequential and rational manner—beginning with phase
the champions or supervisors were responsible for                       one and ending with phase four. Although the amount of
implementing Six Sigma programs with the teams. Third,                  time spent in each stage varied, MSOC experienced the
teams were formed to become completely familiar with                    succession from one stage to another smoothly. After each
the department and its operations. Each team was                        iteration of DMAIC, MSOC documented the process,
responsible for obtaining the latest department functional              interacted to reflect on why the solution worked (or did
flow drawings, the department’s scope or boundaries                      not work), and tried to identify other areas where the
(where it begins and where it ends), and information flow                solution might apply. On reflection, the researcher
throughout the department. These teams would walk                       believed that when the problem was clearly identified,
through the process and observe information flow and the                 MSOC and the departments felt that everyone behaved in
use of various sub-systems and tools. Fourth, once a                    a predictable and rational manner and the company’s
‘‘good’’ understanding of a department and its operations               resources (money and personnel) were not wasted. One
were developed, we needed to verbalize the possible                     example of a Six Sigma project is provided in Appendix C.
problem statement. Typically, the first time a problem                       Second, when a problem was difficult and not clearly
statement is written there is no clear agreement among                  identified early in the process, DMAIC proceeded in a
the team members. Fifth, a room was secured for meeting,                cyclic and reflective manner. In other words, the problem
invitations were sent to management to attend the event,                definition (Define Stage) may not be complete until
and suggestion boxes were set up near the department. A                 Analyze Stage. This process appeared counterintuitive
room was secured close to the target department in order                and irrational to participants and observers in the process.
for engineers and associates to meet frequently and                     During this period, the amount of time spent in each stage
openly discuss ways to improvement the process. Sixth,                  varied greatly. In addition, the succession from one stage
as we know people learn differently (some learn better                  to other was not smooth. In fact, when consensus could
with text and others learn better with visuals), the                    not be reached on either the problem or the solution,
training was delivered by MSOC in an interactive manner                 MSOC proceeded through DMAIC cyclical process (i.e. they
using presentation slides with audio and video enhance-                 painfully jumped from one stage to other with little
ments.                                                                  rationale). This resulted in speculation on root causes of
    MSOC was aware that in the past many improvement                    problems and proposed solutions being implemented
initiatives had failed, not because the programs were                   through ‘‘trial and error’’ with little solid justification.
flawed, but because the company did not identify                         The researcher attempted to guide the efforts; however,
improvement champions and failed to communicate with                    some members became emotional and would not at first
the shop floor employees. In other words, there was no                   accept guidance.
general agreement on what to do, which program to use,                      To stabilize the process, we recommended initially
no adequate training, no communication of what was to                   delaying deliberation on problems which were not fully
be done, no development of trust between workers and                    identified, in favor of addressing those for which the
managers, and no visual support of top management.                      problem was easier to recognize. During the daily review
The MSOC resolved these critical problems to prepare the                sessions, we proposed a revision to the DMAIC training
company for the changes. For example, MSOC knew that                    program. We suggested revising the DMAIC training to
the workers did not trust the management’s commitment                   include both sequential methods as suggested in the
and were not interested in participating and giving their               literature, as well as cyclical methods. In other words, we
honest responses (or suggestions) to improve the process.               proposed an additional one or two-day training session
Over four weeks MSOC held a series of formal and                        where we would discuss how to address a difficult event
informal meetings with workers from different depart-                   where the root cause was difficult to identify. Although
ments. During the last week of the meetings, MSOC                       not specifically for the DMAIC process, previous research
provided a week-long training session on process excel-                 indicates that revised training of a cyclical nature does
lence through Six Sigma. By the time they completed their               reduce average time to solve problems and drive process
training program, MSOC and workers had met twenty                       improvement (e.g., Chakravorty et al., 2008).
times. Upper managers were involved at every stage                          Fig. 1 shows the Six Sigma implementation model.
and exhibited visual commitment to the effort. For the                  First four steps were: Strategic Analysis; Form High-Level
first time in the company’s history, managers answered                   Cross-Functional Team; Establish Improvement Tools; and
                                               ARTICLE IN PRESS
8                                   S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16



                                      Strategic Analysis
                Start               Market/Customer Driven

                                               Change objectives


                                     Form High-Level Cross
                                  Functional Improvement Team

                                               Develop a charter to guide the improvement initiative
                                               Develop timeline to complete all improvements           Strategic Decisions
                                               Establish budget for all the implementation
                                               Establish performance matrices

                                       Establish Overall
                                      Improvement Tools

                                               Develop appropriate training material


                             Perform a High-Level Process Mapping
                              Prioritize Improvement Opportunities

                                                 Establish communication with employees
                                                 Identify improvement champions
                                                                                                       Tactical Decisions

                                  Detailed Plan and Form                               Implementation, Documentation
                               Low-Level Improvement Teams                                     and Revision

                                                                Change training material
                                                                Revise implementation plan
                                                                Continuous Improvement




                                              Fig. 1. Six Sigma implementation model.


Perform high-Level Process Mapping and Prioritization of                   (2005, p. 65) point out, that first step in Six Sigma
Improvement Opportunities. These steps were considered                     implementation:
strategic decisions, implying a top down approach, where
management was primarily involved in decision making.                           yis voice of customer. Yet, in practice is focus often
Last two steps were: Detailed Plan and Form Low-Level                           perverted or diluted. A survey conducted by Greenwich
Improvement Teams; and Implementation, Documenta-                               Associates. Strangely absent from most user responses
tion, and Revision. These steps were considered tactical                        was any mention of the customer. When asked on
decisions implying bottom up approach where engineers                           an unaided basis to define a successful project, only
or technicians were primarily involved in decision mak-                         three out of the 13 companies mentioned customers
ing.                                                                            as critical success factors, despite the first rule of Six
                                                                                Sigma—listening to the voice of customer.

5. Implications of Six Sigma implementation                                    In our implementation experience, customer expecta-
                                                                           tions were fairly evident. Often, this may not be the case,
    There are several important points worth discussing                    and we may have to employ rigorous techniques, such as
about the implementation model. The first step of the                       Quality Function Deployment (QFD) or traditional orga-
model is to perform Strategic Analysis, which needs to be                  nizational assessments. According to Evans and Lindsay
market/customer driven. Our implementation experience                      (2008), the QFD approach incorporates customer require-
shows that the reason for Six Sigma implementation was                     ments and performs competitive evaluation to identify
to improve customer expectations through operational                       design requirements which drives the design of produc-
excellence. Many Six Sigma programs are implemented to                     tion systems. Many companies such as Toyota, General
gain operational efficiency. Unfortunately, many of these                   Motors, Ford, and Mazda have utilized a QFD approach
operational gains do not directly provide enhanced                         to plan and design their products. Recent case studies
customer satisfaction or value. Bendell (2006) claims                      (e.g., Hsieh et al., 2007; Edgeman et al., 2005) involving
that Six Sigma is a strategic approach and improvement                     real world application, have found that a QFD approach
projects should be selected based on improving customer                    could be applied to orchestrate Six Sigma implementation.
satisfaction and operational efficiency. In reality, a                      More research is necessary to connect the QFD approach
majority of the improvement projects are selected based                    to pinpoint customer expectations, and Six Sigma im-
on cost minimization and, therefore, the approach                          plementation. In addition, traditional organization assess-
becomes suboptimal. According to Andel (2007, p. 1) the                    ment-Strengths, Weaknesses, Opportunities, and Threats
cost minimization approach usually translates into a                       (SWOT) analysis can be performed to realize optimum
cutting headcount exercise and when that happens, ‘‘any                    alignment in a business situation (Andrews, 1971). Cheng
program, Lean or Six Sigma is bound to meet stiff,                         (2006) claims that by performing organization assess-
self-protecting opposition.’’ As Zimmerman and Weiss                       ment or SWOT analysis, companies can assess their
                                                ARTICLE IN PRESS
                                     S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16                               9


current situation and develop a strategy for Six Sigma                  analytical and/or mathematical skills to fully understand
implementation. More research is necessary to support                   and implement such tools in improving operations.
this claim.                                                             The projects required application of statistical tools. It
   The second step of the model is to Form High-Level                   is important to understand that often improvement
Cross-Functional Improvement Team. Our implementation                   projects do not require rigorous statistical tools. As Gopal
experience shows that a high-level cross-functional team                (2008, p. 1) says:
was effective in guiding Six Sigma initiatives and keeping
management involvement. Previous research supports our                      Although Six Sigma as a methodology boasts a multi-
implementation experience. According to Jepperson                           tude of robust statistical tools, one needs to be wary of
(2005) Raytheon launched their Six Sigma transformation                     getting trapped by these tools. The practitioner needs
by forming a team consisting of experts from different                      to use these tools to drive improvement and not the
parts of the organization. While implementing a Six Sigma                   other way round. In fact, according to one survey, 60 to
program in an Automotive company, Kumar et al. (2007,                       70 percent of improvement projects do not require
p. 862) found that:                                                         advanced statistical tools. While driving projects, the
                                                                            larger goal of improvement and the essence of the
   yThere was also a need to have effective organiza-                       methodology should be kept intact. Using tools just for
   tional infrastructure in place to support the Six Sigma                  the sake of using them because they are part of the Six
   introduction and development program within the                          Sigma toolkit is not smart or efficient in terms of time.
   organization. Thus, a cross-functional deployment                        Consistent with our implementation experience, there
   team comprising people from middle-level manage-                     is a good bit of confusion among companies attempting to
   ment (executives from Department, Quality Assurance                  select a particular improvement tool such as Six Sigma,
   Department, Sales and Marketing Departments, etc.)                   Lean, or Theory of Constraints for their specific business
   and workers involved on the floor were formed. The Six                environment. The problem is that managers waste their
   Sigma initiative was led by Divisional Manager of the                time in adopting the newest improvement tool, using
   Foundry Shopy . This was followed by formation of y                  consultants, and spending very little time in driving
   other team members who represented different de-                     process improvement from the bottom. In other words,
   partments of the organization.                                       they do not develop deep problem solving capability
                                                                        inside their companies. Hayes et al. (2004) argue that
    There are two points worth mentioning regarding a                   improvement tools fail to show a significant increase in
high-level, cross-functional team to guide Six Sigma                    performance because prior to implementing them; most
implementation. First, in choosing members of the team;                 companies do not develop deep problem-solving capabil-
leadership skills were considered, but no formal psycho-                ities in their employees. They refer to the development of
logical profiles were considered. There are many such                    these capabilities as ‘‘Learning by Doing’’. As the name
tests available. See, for example, Stevens and Campion’s                implies, managers and workers must practice the art and
(1994) ‘‘Knowledge, Skill, and Ability Test,’’ Kembel’s                 science of problem solving to use it effectively. Further,
(1996) ‘‘Rational, Organized, Loving, and Energized Test,’’             they write that ‘‘yyou are not likely to be able to make
and Kolbe’s (1994) ‘‘Measure for Instinctive Behavior of                substantial progress through analysis and detailed im-
Individuals,’’ which could assist in quick decision-making.             provement programs, no matter how many consultants
While theoretical research (e.g., Askin and Huang, 2001)                you hirey’’ There is some research on how to develop
provides strong argument for the consideration of psy-                  deep problem-solving capability in various real-world
chological factors when introducing change programs;                    settings (e.g. Chakravorty et al., 2008) and why this
additional insights are necessary in this area. Second, the             capability is helpful in implementing Lean (Chakravorty
researcher used his implementation experience to deter-                 and Hales, 2008) and Six Sigma (Chakravorty and Franza,
mine a maximum size of eight to lead Six Sigma                          2009). Instead of developing sophisticated statistical
implementation. There is no research to suggest the                     tools; future research needs to develop methods to
optimal number of team members in a steering commit-                    implement deep problem solving capability inside com-
tee. We believe that many factors such as the business                  panies.
environment, the complexity of implementation, and the                      The fourth step of the model is to Perform a High-
skill levels of teams will play an important role in                    Level Process Mapping and Prioritize Improvement Opportu-
determining the optimal number of team members. More                    nities. Previous research supports our implementation
research is necessary to support this claim. A previous                 experience. For example, according to Mader (2008) a
claim (e.g., Kleasen, 2007) has emphasized the need to                  company’ process converts inputs (e.g., material, informa-
conduct research on the role and the appropriateness                    tion) into outputs desired by the customers (products
of cross-functional improvement team for Six Sigma                      or services). A typical approach is to first identify
implementation.                                                         processes within the company and then perform project
    The third step of the model is to Establish Overall                 selection and their prioritization. Thiraviam (2006, p. 40)
Improvement Tools. In our implementation Six Sigma                      said that:
training program for employees included rigorous statis-
tical tools such as Statistical Process Control, etc. Since                 The first tool we used in our case was process mapping.
this was primarily an engineering company, vast majority                    In our Six Sigma based technique, the first step was to
of the engineers or technicians did possess requisite                       understand, identify and define all opportunities
                                                ARTICLE IN PRESS
10                                   S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16


     for failure in a given manufacturing process.                      and Grow, 2007).               According   Angel   and   Pritchard
     We mapped processes from incoming inspection                       (2008, p. 41):
     through shippingy .
                                                                            Examples like Home Depot and 3 M show that
    We experienced considerable difficulty in identifying                    companies cannot focus on implementing Six Sigma
and prioritizing Six Sigma projects in alignment with                       in isolation. Clearly Six Sigma is set of process tools
overall process improvement objectives. In order to                         that should be part of a more holistic process
prioritize our projects, we used cost reduction and                         improvement strategy. For any of these tools to be
revenue generation (or throughput based) potential, along                   used effectively, employee behavior change must be an
with input from workers and management. It is important                     integral part of the programy . A behavior-focused
to learn more about how to identify projects and how to                     approach makes change sustainabley Further, it keeps
prioritize them. One reason many Six Sigma improvement                      us ever aware that a technically sound change designed
programs fail is because improvement projects are not                       by Six Sigma, lean or similar applications could be at
correctly identified and prioritized (Zimmerman and                          risk of failing unless supported by the appropriate
Weiss, 2005). Over the years, many researchers have                         behavior change.
worked on prioritizing improvement projects by mixing
tools such as Six Sigma, Quality, Lean, or Theory of                        Previous research (e.g., Chakravorty and Hales,
Constraints tools. For example, Chakravorty and Atwater                 2008) has shown that effective management of human
(1998) showed how to prioritize quality improvement                     aspects is critical for a successful lean implementation.
projects using Theory of Constraints. Chakravorty and                   Zimmerman and Weiss (2005) point out companies
Sessum (1995) showed how to prioritize Lean improve-                    need to pay attention to the human side of Six Sigma
ment projects using Theory of Constraints. Chakravorty                  implementation. The human side of Six Sigma implemen-
(1996) mixed Lean and Theory of Constraints concepts to                 tation is an important area for future research.
improve the performance of manufacturing operations.                    This research will be greatly helpful for practicing
Recent empirical research (e.g., Banuelas et al., 2006)                 managers wanting to effectively implement Six Sigma
found that companies prioritize improvement initiatives                 programs to achieve sustained results in their business
by mixing these tools. More research is necessary on how                environment.
to mix these tools to correctly identify and prioritize
improvement projects.                                                   6. Conclusion
    The last two steps of the model are to Develop Plan
and Form Low-Level Improvement Teams; and, Implementa-                     Using a successful Six Sigma program in a Network
tion, Documentation, and Revision. In addition to having                Technology company, the purpose of this research was to
a detailed plan for implementation, our Six Sigma                       develop an implementation model which consists of six
implementation was successful because we developed                      steps. The first step is to perform a strategic analysis
an appropriate human infrastructure to drive improve-                   which is customer/market driven. The second step is to
ment from the bottom up. Our approach was to                            establish a high-level, cross-functional team to drive the
establish low-level improvement teams in each depart-                   improvement initiative. The third step is to identify the
ment and choose champions or supervisors to manage                      overall improvement tools. The fourth step is to perform
the teams. In addition to communicating with the                        high-level process mapping and to prioritize improve-
management; champions or supervisors worked with                        ment opportunities. These four steps are considered
the improvement teams on various Six Sigma projects.                    strategic decisions implying a top down approach where
This approach is well supported in the literature.                      management was primarily involved in decision making.
For example, Carlson and Wilmot (2006) claim that                       The fifth and sixth steps are to develop a detailed plan and
low-level improvement teams and champions are                           form low-level improvement teams and to implement,
essential to drive improvement projects from the bottom                 document, and revise as needed. The last two steps are
up. According McManus (2007, p. 20) for Six Sigma                       considered tactical decisions, implying a bottom up
implementation:                                                         approach where engineers or technicians were primarily
     Every organization should have a process improvement               involved in decision making. In addition, important for
     team approach in place that engages a high percentage              both practitioners and academicians, several areas of
     of the work force. Such an approach is mandatory for               future research are also discussed regarding the imple-
     sustained progress toward process excellencey If you               mentation model.
     want to sustain your improvement efforts over timey                   Lastly, this research provides a model to effectively
     give a training [to the improvement teams], and turn               guide the implementation of Six Sigma programs
     teams loose.                                                       to reduce variation or waste from the operations.
                                                                        This is particularly relevant because today’s competitive
   There is growing concern that Six Sigma or other                     environment demands that companies reduce variation
process improvement programs fail because they                          (waste) to meet or exceed efficiency and responsi-
do not consider the human side of implementation.                       veness requirements of customers. There is increasing
For example, Six Sigma implementation negatively                        pressure to pursue new ways of thinking as a source
affected employee morale at Home Depot and stifled                       of competitive advantage. More research in this area is
creativity and innovation at3M (Hindo, 2007; Hindo                      necessary to contribute to the science and practice of
                                                        ARTICLE IN PRESS
                                             S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16                                              11


implementation of Six Sigma or any other process                                MEASURE
improvement model, to reduce waste and create                                   Establish performance metrics
                                                                                Develop data collection and sampling plan
value.
                                                                                Conduct measurement systems analysis
                                                                                ANALYZE
Appendix A. Six Sigma training                                                  Benchmark the process or product
                                                                                Analyze the process map theory of constraints
                                                                                Determine and summarize root cause
                                                                                IMPROVE
Executive overview for Six-Sigma                                                Determine breakthrough solutions, costs and benefits
Maximum of 5 attendees—1 day course                                             Implement and measure solution effectiveness
This course is for people from service backgrounds. It is designed for          CONTROL
executives working directly with Champions, Black Belts, and Green              Determine type of control needed
Belts on Six-Sigma projects.                                                    Implement and validate controls



This course includes an overview of Six-Sigma concepts, including:              Black Belt—process excellence for Six-Sigma
What is Six-Sigma?                                                              5–10 attendees—2 week course
How Six-Sigma fits into your company?
Integration of Lean manufacturing into Six-Sigma deployment
DMAIC road map
Expected benefits of Six-Sigma                                                   This course focuses on product production excellence. It is designed for
Management infrastructure for Six-Sigma                                         those individuals working directly on Six-Sigma projects as Black or
Business goals, objectives, and dashboards for Six-Sigma                        Green Belts.
DMAIC breakthrough methodology                                                  This course covers Six-Sigma philosophy and tools. Students will learn, at
Implementation issues and timelines                                             both a basic and statistical level, the DMAIC methodology and apply it to
Six-Sigma deployment model                                                      their manufacturing projects.
Project selection and Theory of Constraints
Black Belt selection
Reviewing progress and maintaining accountability                               DEFINE
Next steps                                                                      Define resource/stakeholder analysis
                                                                                Project leadership
                                                                                MEASURE
Champion training for Six-Sigma                                                 Determine critical Xs and Ys
5–10 attendees—2 day course                                                     Understand foundation of statistics
This course is for people from service backgrounds. It is designed for          ANALYZE
directors and managers working directly as champions of Six-Sigma               Establish causal relationships using regression analysis
projects.                                                                       Visualize the problem and determine root cause
                                                                                IMPROVE
                                                                                Measure solution effectiveness
                                                                                Rapid DOE (Design of Experiment)
This course includes an overview of Six-Sigma concepts, including:              CONTROL
What is Six-Sigma?                                                              Statistical process control
Expected benefits of Six-Sigma                                                   Develop transfer plan and close project
DMAIC breakthrough process
Management and organizational infrastructure for Six-Sigma                      (1) Keller, P., 2005. Six Sigma: Demystified. McGraw-Hill. New York, NY.
Project identification and prioritization                                        (2) Pyzdek, T., 2003. Six Sigma. Handbook. McGraw-Hill. New York, NY.
Financial analysis
Consider painful processes
Examining areas of waste
Look at the cost of quality
Customer satisfaction
                                                                                Appendix B. Training program enhancements
Link projects to strategic objectives
Prioritizing projects and Theory of Constraints
DMAIC breakthrough methodology and deliverables
Black Belt and Green Belt selection and management                              Lean Operations principles categorize activities into value added and
Six-Sigma implementation roadmap                                                non-value added activities. The focus of these principles is on
                                                                                eliminating all forms of non-value added activities. There are seven
                                                                                forms of non-value added (or waste): waste of over production, waste of
                                                                                inventory, waste of defects, waste of motion, waste of processing, waste
Green Belt—process excellence for Six-Sigma
                                                                                of waiting, and waste of transportation. Lean Operations tools include:
5–10 attendees—1 week course
This course focuses on product production excellence. It is designed for
those individuals working directly on Six-Sigma projects as Green Belts.            Value stream mapping
                                                                                    Lean events
                                                                                    SMED (a.k.a quick changeover)
                                                                                    Manufacturing cell design
This course covers the Six-Sigma philosophy and tools. Students will                Poka-Yoke (a.k.a. mistake proofing)
learn the DMAIC methodology at a basic level and apply it to their                  5S plus one (Sort, Straighten, Shine, Standardize, Sustain, and Safety)
manufacturing projects.
                                                                                Theory of Constraints (TOC) challenges the belief of functional
                                                                                independence with the view of an organization as a chain of dependent
DEFINE                                                                          activities. According to TOC, every organization has a constraint or
Define problem, objective, goals, and benefits                                    bottleneck and the existence of a bottleneck provides an opportunity for
Develop project plan and map the process to identify waste                      process improvement priority. In addition, TOC provides a set of
                                                                      ARTICLE IN PRESS
12                                                      S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16


measures that link local actions to the measures of organization wide                                the service orders and due to critical role in the process
measures. TOC tools include:                                                                         also had two other RMA types that would be in scope:
  Bottleneck analysis                                                                               cirmisact and unknactivity. The second highest overall and
  Drum buffer rope
  Buffer management                                                                                 the highest RMA for Systems Management (SM) was
  Buffer sizes and locations                                                                        interprocesserr. Finally, the largest RMA for Service Order
  Throughput maximization                                                                           Management (SOM) was rcidnotfnd. This rcidnotfnd type
  VAT operations analysis                                                                           was later eliminated from consideration due to a system
                                                                                                     dependency. Two RMA types were deemed redundant
(1) Womack, J.P., & Jones, D.T., 2003. Lean Thinking. Free Press, New York,                          (ddmissed and dvamissed) and other there were thrown
NY.                                                                                                  out due to small number of observations. Table 1 provides
(2) Goldratt, E.M., & Cox, J., 2004. The Goal. North River Press, Great                              the detailed log of RMAs for the weeks specified and the
Barrington, MA.
                                                                                                     Pareto Chart that summarizes the observation data over
                                                                                                     the same period (Fig. 2).
                                                                                                        The team initially reviewed raw RMA data but after
                                                                                                     considering counting constraints it was decided that
Appendix C. Implementation of a Six Sigma project                                                    simplifying the measure to RMA per Service Order
                                                                                                     (RMA/SO) would produce the most meaningful results.
C.1. Define                                                                                           The project objective was changed to reduce RMA/SO by
                                                                                                     50% by December 31. The team calculated the expected
   During scope definition, the project team evaluated all                                            objective by averaging four weeks of data after all system
Requests for Manual Activity (RMA) categories and related                                            functionality was put into place. The project used a
sub-groups which define the group owner. After tracking                                               baseline of September 29. The team proceeded by
and recording RMA events over a week period, the team                                                counting two weeks past the baseline order to ensure
was ready to review the results to determine the proper                                              that the numbers included all the data. This was required
course of action to drive corrections. The team elected to                                           since RMA/SO are measured on service order due date
pursue the top RMA generates for each sub-group and the                                              which typically occurs within two weeks of the service
next two RMA types with the highest contribution.                                                    order submission. The resulting baseline was 4.6 RMA/SO
Overall, the highest RMA was missfid which impacted                                                   calculated between 10/13 and 11/7. Therefore, the objec-
the Service Order Processing (SOP) team. This team wrote                                             tive was established at 2.3 RMA/SO.
                                                                                                        Simple count, sore, and Pareto analysis were use to
                                                                                                     identify the RMA that occurred most frequently. In
Table 1
Initial RMA observation count.                                                                       addition, field experience was a determining factor in
                                                                                                     evaluating which RMA took the longest to resolve. This
RMA type           RMA sub-group 9–22 9–23 9–24 9–25 9–26 Total                                      practice was input to RMA priority which then honed
                                                                                                     further using root cause analysis. After the root causes
missfd             SOP               214        176          510 308        264        1472
interprocessserr   SM                  85        48            72 76          66         347         were isolated, and improvement plan was designed
ddmissed           Date                   0      85         134 83             57       359          based on the priority and system capabilities. The project
unknactivity       SOP               104         63            47 41          38        293          was subject to following complications and constraints
pvcpserr           SOM                  14       29            19      5       14          81        (Table 2).
cirmisact          SOP                 72        38            36 67           18        231
rcidnotfnd         SOM                 25        28            46 30           30        159
dvamissed          Date                   0       13           31 14           17          75        C.2. Measure
socpxfail          SOM                 40        26            22 22          28         138
iportpserr         SOM                    2         5           8      4         8         27
....               ....              ....      ....       ....    ....      ....      ....              In order to fully understand the RMAs, the team
Total              ALL RMAs          739       669        1099 807          647       2489           prepared BBNMS process flows. This permitted the team
                                                                                                     to conceptualize breakdowns and refer to the correction

                          3000                                                                                                                 1.00
                                                                                                                                               0.90
                          2500                                                                                                                 0.80
                          2000                                                                                                                 0.70
                                                                                                                                               0.60
                          1500                                                                                                                 0.50
                                                                                                                                               0.40
                          1000                                                                                                                 0.30
                            500                                                                                                                0.20
                                                                                                                                               0.10
                               0                                                                                                               0.00
                                      d


                                                   d




                                                                           ity



                                                                                                 t

                                                                                                d


                                                                                                                l




                                                                                                                                d
                                                               rr




                                                                                                                       rr




                                                                                                                                          rr
                                                                                                             ai
                                                                                          ac
                                     sf

                                               se




                                                                                             tfn




                                                                                                                              se
                                                             se




                                                                                                                       e




                                                                                                                                        se
                                                                           iv




                                                                                                         xf

                                                                                                                    ps
                                   is




                                                                                      is
                                              is




                                                                                                                              is
                                                                                           no
                                                                        t
                                                          ss




                                                                                                                                     rtp
                                                                                                       cp
                                                                     ac
                                 m




                                                                                     rm




                                                                                                                   c
                                           m




                                                                                                                            am
                                                        ce




                                                                                                id




                                                                                                                pv
                                                                                                     so




                                                                                                                                      o
                                                                 kn
                                        dd




                                                                                 ci




                                                                                                                                   ip
                                                                                           rc




                                                                                                                       dv
                                                     ro

                                                               un
                                                   rp
                                                te
                                              in




                                                                                     Fig. 2. Pareto chart.
                                                          ARTICLE IN PRESS
                                               S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16                              13


Table 2                                                                            time recorded for repair. The team then determined which
System complications and constraints.                                              RMAs to measure by examining the RMA histogram. Upon
                                                                                   reviewing the special causes, it was fund that interpro-
Condition                             Description
                                                                                   cesserr RMA contained embedded error codes. The other
Redundant RMA counts                ddmissed (due date missed) and                 three RMA types (missfid, cirmisact, and unknactivity) did
                                    dvamissed were that were                       not contain this information and manual investigation
                                    generated because of other RMAs                was required.
Initial delay due to system         All states were not initially
unavailability                      represented
                                                                                       The team then conducted a detailed review of each of
Total service orders not counted by Sigma number unavailable                       the four remaining RMAs. The sample size was deter-
OMS (Order Management System)                                                      mined by taking each RMA, beginning and missfid,
RMA/SO numbers based on RMA         To be determined                               calculating the number of service orders with that RMA
due date (dd) vice when the RMA
                                                                                   in one week, and comparing that total with the total
initially falls out
Team member availability            Workload dependent and not 100%                number of service orders. Table 3 shows the sample size
Accurate time measurements          Extreme variability in manual                  (n) was derived.
                                    rework                                             Each of the RMA data sets was randomly sampled by
Existing records issue              Addressed in improve phase                     computer for 62 unique entries for missfid, cirmisact, and
                                                                                   interprocesserr. Since the data set was small for unknac-
                                                                                   tivity only 40 samples were taken. Once the service order
Table 3
Sample size calculations.                                                          samples were identified, the results were tabulated
                                                                                   in excel. The summarized results from unknactivity are
Week     Total# SOs     #missfid SOs     Z value (90%)    p        c     n          shown in Table 4.
                                                                                       The interprocesserr RMA differs from the service order
11–3     2860           1033            1.64             0.36     0.1   62
                                                                                   management and processing RMAs. This type occurs when
                                                                                   BBNMS encounters an unexpected event, something that
Table 4                                                                            was not previously defined. Since these RMAs related to
Unknactivity results.                                                              developed software, the developer devised a method for
                                                                                   identification and tracking of these specific cases. In a
Order problem                   Quantity                     Percentage (%)
                                                                                   similar fashion as the other RMA type, the team collected
Circuit                          6                            15.00                and prioritized the data by frequency of receipt. The total
Records                         11                            27.50                service orders were available for interprocesserr RMA so U
Special assembly                10                            25.00                Charts were generated based on a two week period
Missing action                  13                            32.50
Total                           40                           100.00
                                                                                   calculation. Only the top ten interprocesserr error codes
                                                                                   included amounting for 86.99% of all errors (Figs. 3–6).

                            7
                            6
                            5
                            4
                            3
                            2
                            1
                            0
                                 70-153                 154-237               238-320           321-404          405-488
                                                                   missfid RMAs 10/13-11/7

                                                                  Fig. 3. Missfid histogram.



                            9
                            8
                            7
                            6
                            5
                            4
                            3
                            2
                            1
                            0
                                  32-65                 66-98                 99-132            133-165          166-199
                                                                cirmisact RMAs 10/13-11/7

                                                                 Fig. 4. Cirmisact histogram.
                                                            ARTICLE IN PRESS
14                                             S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16


                      12
                      10
                       8
                       6
                       4
                       2
                       0
                                  19-65                    66-112              113-158                159-205     206-252
                                                                  intrprocesserr RMAs 10/13-11/7

                                                                 Fig. 5. Interprocesserr histogram.



                      9
                      8
                      7
                      6
                      5
                      4
                      3
                      2
                      1
                      0
                                  2-16                     17-30                  31-45               46-59        60-74
                                                                  unknactivity RMAs 10/13-11/7

                                                                  Fig. 6. Unknactivity histogram.



                          0.080
                          0.070
                                                                                                                            UCL
                          0.060
                          0.050
                          0.040
                          0.030
                          0.020
                          0.010
                                                                                                                            LCL
                          0.000
                                   1       2       3   4     5      6    7    8     9     10 11 12 13 14 15 16
                                                                endpoint collision 11/17-12/12

                                                                     Fig. 7. End point U chart.




Table 5                                                                                 condition/constraint and was caused by BBNMS procedur-
Improvements.                                                                           al change in how a service order is prepared. It was
                                                                                        determined that this information would be noted as a
RMA type          SS team              Other               Total
                  initiated            initiated           improvements                 change to the order preparation process. This field is
                                                                                        necessary input into the automated flow through process
missfid            7                    3                   10                           and will be listed as a required field. This change would
cirmisact         6                    0                    6                           only benefit future orders. The other RMA categories
unknactivity      5                    0                    5
interprocesserr   5                    3                    8
                                                                                        include in-flight software code changes (36.33%), pending
                                                                                        future change requests (9.34%), and training issues
                                                                                        (12.46%).
                                                                                           The RMA cirmisact was similar to missfid as 38.81% of
C.3. Analyze                                                                            the samples were records related. In this instance, a
                                                                                        greater portion of the sample of RMAs has either had edits
   Upon analyzing the missfid RMA type and categorizing                                  completed or in-process (44.78%). The remaining issues
the results, the data indicated that the majority of orders                             included training (5.97%), future software release (4.48%),
(41.87%) produced RMAs due to an existing records issue.                                parser sequencing (2.99%), and unidentified problems
This issue was originally identified as a pre-existing                                   (2.99%).
                                                         ARTICLE IN PRESS
                                              S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16                                              15


Table 6
Cost of quality.

08/06-01/02        RMA     RMAs 10/ RMAs     # Sos     Min/ Total  Total Total     Avg cost       Cost 10/    Cost 12/ Six Sigma         Annualized Six
                   Total   13-11/7  12/8-1/2 (Tot/     Ord Min ($) Hr ($) cost ($) RMA ($)        13-11/7 ($) 8-1/2 ($) savings ($)      Sigma savings ($)
                                             2.85)

cirmisact       4634 1965            289      1626    12      19,512 325       14,006              5,939           873      5,066        20,263
interprocesserr 7399 1558            651      2596    25      64,904 1,082     46,590              9,810         4,099      5,711        22,845
missfid         15007 4571           1890      5266    12      63,187 1,053     45,358             13,816         5,712      8,103        32,413
unknactivity     289 492             285      1017    12      12,206 203        8,762              1,487           861        626         2,503
Grand total    27329 8586           3115     10505    61     159,809 2,663    114,716             31,052        11,547     19,506        78,023




    The RMA unknactivity was also comparable to missfid                           References
and cirmisact where the records issue was significant
accounting for 27. Zero percent of the random sample. The                        Adams, C., Gupta, P., Wilson, C., 2003. Six Sigma Deployment. Butter-
remaining issues included training (32.5%), special as-                              worth-Heinemann, Oxford, UK.
                                                                                 Andrews, K.R., 1971. The Concepts of Corporate Strategy. Dow Jones-
sembly (25%), and circuit order-related (15%).                                       Irwin, New York, NY.
    The interprocesserr RMA analysis evaluated the histo-                        Andel, T., 2007. Lean & Six Sigma Traps to Avoid. Material Handling
grams from the top ten error codes. As seen in other                                 Management, March /http://www.mhmonline.com/viewStory.asp?
                                                                                     nID=5423S.
RMA types, a large amount of special cause variation was
                                                                                 Angel, D.C., Pritchard, C., 2008. Behavior tests Six Sigma. Industrial
evident. This variation was captured using Poisson U                                 Engineer (August), 41.
Charts. The data was then analyzed per error code and a                          Antony, J., Antony, F.J., Kumar, M., Cho, B.R., 2007. Six Sigma in service
                                                                                     organisations: benefits, challenges and difficulties, common myths,
sample U Chart from the most frequently reported error,
                                                                                     empirical observations and success factors. International Journal of
end point collision, is shown in Fig. 7.                                             Quality & Reliability Management 24 (3), 294–311.
    Improve: One of the biggest challenges during project                        Antony, J., Banuelas, R., 2002. Key ingredients for the effective
execution was managing change requests to incorporate                                implementation of Six Sigma program. Measuring Business Excel-
                                                                                     lence 6 (4), 20–27.
new edits, defects, and other desired improvement                                Antony, J., Kumar, M., Madu, C.N., 2005. Six Sigma in small- and medium
actions. Table 5 summarizes the types of changes and                                 sized UK manufacturing enterprises: Some empirical observations.
references where they were initiated or influences by the                             International Journal of Quality & Reliability Management 22 (8),
                                                                                     860–874.
Six Sigma project team or from an external source. All                           Askin, R.G., Huang, Y., 2001. Forming effective worker teams for cellular
changes summarized in this table were acted upon so 29                               manufacturing. International Journal of Production Research 39 (11),
total changes were introduced from this initiative. Overall,                         2431–2451.
                                                                                 Banuelas, R., Tennant, C., Tuersley, I., Tang, S., 2006. Selection of Six
the project delivered a 52.08% process improvement from                              Sigma projects in the UK. The TQM Magazine 18 (5), 514–527.
baseline to completion.                                                          Bendell, T., 2006. A review and comparison of Six Sigma and the lean
    Control: While the project ended, the team instituted a                          organizations. The TQM Magazine 18 (3), 255–262.
                                                                                 Berg, M., 2006. Six Sigma shortcomings. Industrial Engineer (October),
lifecycle improvement process to monitor activities for
                                                                                     45.
future identification and elimination of RMAs. A workflow                          Breyfogle, F., Cipello, J., Meadows, B., 2001. Managing Six Sigma. Wiley
analyst position was created to track RMA numbers                                    Inter-Science, New York, NY.
                                                                                 Coronado, R.B., Antony, J., 2002. Critical success factors for the successful
on a daily basis and identify and investigate when
                                                                                     implementation of Six Sigma projects in organisations. The TQM
RMA numbers fall outside the control limits. In addition,                            Magazine 14 (2), 92–99.
a lifecycle manager was given responsibility to lead                             Carlson, C.R., Wilmot, W.W., 2006. Innovation: The Five Disciplines for
regular recurring service order meetings to manage the                               Creating What Customer Want. Random House, Inc., New York, NY.
                                                                                 Chakravorty, S.S., 1996. Robert Bowden, Inc.: a case study of cellular
work effort related with the identification and correction                            manufacturing and drum buffer rope implementation. Production
of RMA issues. The project enabled the lifecycle team with                           and Inventory Management Journal 3rdQtr, 14–18.
many new tools in the form of histograms, Pareto Charts,                         Chakravorty, S.S., Atwater, J.B., 1998. Implementing quality improvement
                                                                                     programs using the focusing steps of theory of constraints.
Statistical Process Controls charts, etc. to assist with this                        International Journal of Technology Management 16 (4/5/6/),
analysis.                                                                            544–555.
    Cost of quality: By implementing improvements across                         Chakravorty, S.S., Atwater, J.B., 2006. Bottleneck management: theory
                                                                                     and practice. Production & Planning Control 17 (5), 441–447.
for RMA types from 29 requests, this project determined                          Chakravorty, S.S., Hales, D.N., 2004. Implications of cell design imple-
the Total cost of Quality as $318,635 which reflects                                  mentation: a case study and analysis. European Journal of Opera-
prevention, appraisal, and internal failure costs (no                                tional Research 152, 602–614.
                                                                                 Chakravorty, S.S., Hales, D.N., 2008. The evolution of manufacturing cells:
external failure numbers were available). This includes
                                                                                     an action research study. European Journal of Operational Research
costs associated the original process through the comple-                            188, 153–168.
tion of the project. Many of these numbers were estimates                        Chakravorty, S.S., Hales, D.N., Herbert, J.I., 2008. How problem solving
                                                                                     really works. International Journal of Data Analysis Techniques and
since no individual data was recorded. To reflect the
                                                                                     Strategies 1 (1), 44–59.
Annualized Total Cost of quality ($479,237), the above                           Chakravorty, S.S., Franza, R.M., 2009. Implementation of design for Six
number was adjusted in scale from five months originally                              Sigma (DFSS): a development experience. International Journal of
represented into 12. The saving generated from this                                  Product Development, in press (Special Issue on: Transformations in
                                                                                     Industrial Design: Emerging Trends, Approaches, and Challenges).
initiative was $19,506 and would be $78,023 if annualized                        Chakravorty, S.S., Sessum, J.L., 1995. Developing effective strategies to
(Table 6).                                                                           prioritize set-up reduction in multi-machine production system.
                                                         ARTICLE IN PRESS
16                                            S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16


     International Journal of Operations and Production Management 15            Kemmis, S., McTaggart, R., 2000. Participatory action research. In:
     (10), 103–111.                                                                  Denzin, N.K., Lincoln, Y.S. (Eds.), Handbook of Qualitative Research,
Chandra, A., 2008. Five Six Sigma deployment mistakes- and how to                    second ed. Sage Publications, Thousand Oaks, CA, pp. 567–606.
     avoid them. /http://www.isixsigma.com/library/content/c080414a.             Ladani, L.J., Das, D., Cartwright, J.L., Yenker, R., Razmi, J., 2006.
     aspS.                                                                           Implementation of Six Sigma system in celestica with practical
Cheng, J.L., 2006. DMAIC integration necessary for success. ASQ SIX                  examples. International Journal of Six Sigma and Competitive
     Sigma Forum Magazine 5 (4), 22–28.                                              Advantage 2 (1), 69–88.
Davison, L., Al-shaghana, K., 2007. The link between Six Sigma and               Lakhavani, S.T., 2003. Six Sigma implementations: trials, tribulations,
     quality culture—an empirical study. Total Quality Management 18                 and lessons learned. ASQ’s Annual Quality Congress Proceedings 57,
     (3), 249–265.                                                                   643–647.
Edgeman, R., Bigio, D., Ferleman, T., 2005. Six Sigma and business               Lee-Mortimer, A., 2006. Six Sigma: a vital improvement approach when
     excellence: strategic and tactical examination of IT service level              applied to the right problems, in the right environment. Assembly
     management at the office of the Chief Technology Office of                        Automation 26 (1), 10–18.
     Washington, DC. Quality and Reliability Engineering International           Lynch, D.P., Bertolino, S., Cloutier, E., 2003. How to scope DMAIC projects.
     21, 257–273.                                                                    Quality Progress 36 (1), 37–41.
Evans, J.P., Lindsay, W.M., 2008. Managing for Quality and Performance           Mader, D.P., 2008. Lean Six Sigma’s evolution. Quality Progress 41 (1),
     Excellence, seventh ed. South-Western, Mason, OH.                               40–48.
Feng, Q., Manuel, C.M., 2007. Under the knife: a national survey of Six          Mast, J.D., Bisgaard, S., 2007. The science in Six Sigma. Quality Progress
     Sigma programs in US healthcare organizations. International                    40 (1), 25–29.
     Journal of Health Care Quality Assurance 21 (6), 535–547.                   Mcadam, R., Evans, A., 2004. Challenge to Six Sigma in a high technology
Fontana, A., Frey, J.H., 2000. The interview: from structured questions to           mass-manufacturing environments. Total Quality Management 15
     negotiated texts. In: Denzin, N.K., Lincoln, Y.S. (Eds.), Handbook of           (5/6), 699–706.
     Qualitative Research, second ed. Sage Publications, Thousand Oaks,          McManus, K., 2007. The trouble with teams. Industrial Engineer 39 (10),
     CA, pp. 567–606.                                                                20.
Foster, S.T., 2007. Does Six Sigma improve performance? The Quality              Meredith, J., 1998. Building operations management theory through case
     Management Journal 14 (4), 7–20.                                                and field research. Journal of Operations Management 16, 441–454.
General Electric Company, 2005. Six Sigma Customer Black Belt Training:          Meredith, J., Mantel, S., 2003. Project Management: A Managerial
     Six Sigma Customer Solutions. Cleveland, OH.                                    Approach. Wiley, New York, NY.
Gijo, E.V., Rao, T.L., 2005. Six Sigma implementation—hurdles and more           Mullavey, F., 2005. Shackled by bad Six Sigma. Quality Digest, September
     hurdles. Total Quality Management 16 (6), 721–725.                              /http://www.qualitydigest.com/sept05/articles/03_article.shtmlS.
Gitlow, H.S., Levine, D.M., 2005. Six Sigma for Green Belts and                  Pande, P.S., Neuman, R.P., Cavanagh, R., 2000. The Six Sigma Way: How
     Champions: Foundations, DMAIC, Tools, Cases, and Certification.                  GE, Motorola, and Other Top Companies are Honing their Perfor-
     Prentice-Hall, Upper Saddle River, NJ.                                          mance. McGraw-Hill, New York, NY.
Gopal, R., 2008. 10 ways to failure for a new Six Sigma program. /http://        Pyzdek, T., 2003. The Six Sigma Handbook: A Complete Guide for Green
     www.isixsigma.com/library/content/c071203a.aspS.                                Belts, Black Belts, and Managers at All Levels. McGraw-Hill, New
Gupta, P., 2008. Reducing the cost of failures. Quality Digest 47 (1), 22.           York, NY.
Hales, D.N., Chakravorty, S.S., 2006. Implementation of Deming’s style of        Rajagopalan, R., Francis, M., Suarez, W., 2004. Developing novel catalysts
     quality management: an action research study. International Journal             with Six Sigma. Research—Technology Management 47, 13–16.
     of Production Economics 103, 31–148.                                        Rosenthal, R., Rosnow, R., 1991. Essentials of Behavioral Research.
Harry, M.J., Linsenmann, D.R., 2007. The Six Sigma Fieldbook: How                    McGraw-Hill Publishers, Boston, MA.
     DuPont Successfully Implemented the Six Sigma Breakthrough                  Savolainen, T., Haikonen, A., 2007. Dynamics of organizational learning
     Strategy. Doubleday Business of Random House, Inc., New York, NY.               and continuous improvement in Six Sigma implementation. The
Harry, M.J., Schroeder, R., 2000. Six Sigma: The Breakthrough Manage-                TQM Magazine 19 (1), 6–17.
     ment Strategy Revolutionizing the World’s Top Corporations. Ran-            Schonberger, R.J., 2008. Best Practices in Lean and Six Sigma Process
     dom House, Inc., New York, NY.                                                  Improvement. John Wiley & Sons, Hobokon, NJ.
Hsieh, C.T., Lin, B., Manduca, B., 2007. Information technology and Six          Schroeder, R.G., Linderman, K., Liedtke, C., Choo, A.S., 2008. Six Sigma:
     Sigma implementation. The Journal of Computer Information                       definition and underlying theory. Journal of Operations Management
     Systems 47 (4), 1–10.                                                           26 (4), 536–554.
Hayes, R.H., Pisano, G.P., Upton, D.M., Wheelwright, S.C., 2004. Opera-          Snee, R.D., Hoerl, R.W., 2003. Leading Six Sigma. Wiley, Hoboken, NJ.
     tions, Strategy, and Technology: Pursuing the Competitive Edge. John        Stake, R.E., 2000. Case studies. In: Denzin, N.K., Lincoln, Y.S. (Eds.),
     Wiley & Sons, Indianapolis, IN.                                                 Handbook of Qualitative Research, second ed. Sage Publications,
Hindo, B., 2007. At 3M, a struggle between efficiency and creativity.                 Thousand Oaks, CA, pp. 435–454.
     Business Week, June 11.                                                     Stevens, M.J., Campion, M.A., 1994. The knowledge, skill, and ability
Hindo, B., Grow, B., 2007. Six Sigma: so yesterday. Business Week, June              requirements for teamwork: implications for human resource
     11.                                                                             management. Journal of Management 20 (2), 303–530.
Jacobs, F.R., Chase, R.B., 2008. Operations and Supply Chain Manage-             Sutton, C., 2006. Getting the most out of Six Sigma. Quality Management
     ment: The Core. McGraw-Hill Irwin, New York, NY.                                (March), 46–48.
Jaffery, M., 2003. Indian Cooking. Barron’s Educational Series, Inc.,            Szeto, A.Y.T., Tsang, A.H.C., 2005. Antecedent to successful implementa-
     Hauppauge, NY.                                                                  tion of Six Sigma. International Journal of Six Sigma and Competitive
Jepperson, M., 2005. Focused journey of change. Industrial Management                Advantage 1 (3), 307–322.
     (July/August), 8–13.                                                        Thiraviam, A.R., 2006. Simple tools for complex systems. Quality Progress
Keller, P., 2005. Six Sigma: Demystified. McGraw-Hill, New York, NY.                  39 (6), 40–44.
Kembel, J., 1996. Role Play—Personalities in Action. Life Change Partners,       Treichler, D., Carmichael, R., Kusmanoff, A., Lewis, J., Berthiez, G., 2002.
     Boston, MA.                                                                     Design for Six Sigma: 15 lessons learned. Quality Progress 35, 33–42.
Kleasen, K., 2007. Building human resources strategic planning, process          Watson, G., 2006. Building on Six Sigma effectiveness. ASQ Six Sigma
     and measurement capability: using Six Sigma as a foundation.                    Forum Magazine 5 (4), 14–15.
     Organizational Development Journal 25 (2), 37–41.                           Wurtzel, M., 2008. Reasons for Six Sigma deployment failures. BPM
Kolbe, K., 1994. Pure Instincts. Random House, New York, NY.                         Institute. June /http://www.bpminstitute.org/articles/article/article/
Kumar, U.D., Nowicki, D., Ramirez-Marquez, J., Verma, D., 2008. On                   reasons-for-six-sigma-deployment-failures.htmlS.
     the optimal selection of process alternatives in a Six Sigma                Yin, R.K., 2008. Case Study Research: Design and Methods. Sage
     implementation. International Journal of Production Economics 111,              Publications, Thousand Oaks, CA.
     456–467.                                                                    Zimmerman, J.P., Weiss, J., 2005. Six Sigma’s seven deadly sins. Quality
Kumar, M., Antony, J., Antony, F.J., Madu, C.N., 2007. Winning customer              44, 62–66.
     loyalty in an automotive company through Six Sigma: a case                  Zu, X., Fredendall, L.D., Douglas, T.J., 2008. The evolving theory of quality
     study. Quality and Reliability Engineering International 23,                    management: the role of Six Sigma. Journal of Operations Manage-
     849–866.                                                                        ment 26, 630–650.

				
DOCUMENT INFO