1
Use of Lean Six Sigma tools in DfSS Projects
by
Andreas Stoikos University of Groningen Faculty of Economics and Business
MSc Supply Chain Management August 24th, 2017
Supervisor: Dr. J.W.J. Timans
Co-assessor: Prof. Dr. Ir. C.T.B. Ahaus
Esdoornlaan 580 9741MG Groningen
2
Preface
This thesis is the last part of accomplishing my Master in Supply Chain Management at the University of Groningen. I could not be able to complete this thesis without the active support, guidance,
cooperation and encouragement of people whom I wish to express my gratitude and appreciation.
First of all I would like to deeply thank Dr. J.W.J. Timans who is both my supervisor and thesis mentor. His continuous guidance and encouragement throughout the whole process ensured
accomplishing this thesis. Besides, he provided the contact information of potential participants of this research. Without this help, it would have been much harder to find participants in my research. My deeply gratitude goes to him for his continuous support from the initial phase of the project on and for teaching me the core values of Lean Six-Sigma and Design for Six-Sigma projects. Moreover, my appreciation goes to my co-assessor, Professor Dr. C.T.B. Ahaus, professor in Supply Chain
Management at University of Groningen. He helped me to find a research topic close to my interests and he provided me valuable support and encouragement, especially at the beginning of the process.
This thesis would not have been possible without the participation of all the GBs and BBs involved in the single company case study. I truly want to thank all the employees participated in this research for the inspiring information provided during the interviews. Next to that, I would like to thank my fellow student Rutger Van Ruiten who had a different Master thesis topic related to DfSS projects. We have created together the Interview Protocol and we have encouraged each other during the whole process.
On a personal note, my last but true gratitude goes to my family and friends who directly and indirectly encouraged and supported me in order to complete this master thesis.
3
Abstract
The purpose of this paper is to discover how LSS tools are used within DfSS projects. In this regard, the methodology explored in these projects is the IDOV (Identify-Design-Optimize-Verify) or similar methodologies. IDOV is the most popular methodology for New Product Development processes. It has been explored which are the key tools deployed within the DfSS projects, what are the reasons to use these tools and which are the deliverables in each of the IDOV phases in order to proceed to the following one. Next to that, this research has investigated if the findings support or not the IDOV methodology described by Antony.
A single case study within a Dutch manufacturing company has been conducted. Company A is working at Lean Six Sigma environment and it has experience on DfSS projects. The semi-structured interviews have been answered by 4 BBs and 3 GBs working at that manufacturing company. Those engineers are experienced carrying out DfSS projects and they have provided insightful information on how LSS tools are used in each phase of IDOV methodology. This kind of methodology is applied in the design of new products for household and personal care.
Company A follows the DIDOVM methodology which is a variation of IDOV. Define which is the first phase is a higher level management phase in order to setup the whole project plan. Monitor which is the last phase of that methodology, it is performed by the production where it is checked if the production is running well over time. The deliverables of the rest of the phases are supporting the article of Antony (2002) in a generic way. The appropriate usage of tools could facilitate the implementation of the whole DfSS-project. The main tools which are used in all the projects of company A, independently of the context are: CTQ flowdown, QFD, V-model, Design Dashboard, FMEA, Gage R&R, DoE and Pugh Matrix.
Comparing the literature case with the findings, there are two alternative tools in order to select the best design concept. Those are AHP (analytical hierarchy process) and Pugh matrix. In practice Pugh matrix is used more as it is a simpler tool. Furthermore, a selection can be made between Design Dashboard and DfSS scorecards. DIDOVM methodology is building quality into the design and it takes also into account the business needs and the production monitoring as well.
The paper describes how the DIDOVM methodology is applied and provides more in depth details about the tools usage in each phase and the main deliverables which permit to proceed to the following phase. The absence of standardized decision processes for selecting the most suitable tools in a
specific context can cause misuse of tools, which not only impede the successful implementation of the project, but also make the problem even worse.
4
Table of contents
1. INTRODUCTION 5 2. THEORETICAL BACKGROUND 8 3. METHODOLOGY 166 4. RESULTS 199 5. DISCUSSION 288 6. CONCLUSION 322 REFERENCES 344APPENDIX I: INTERVIEW PROTOCOL 377
APPENDIX II: CONFERENCE ACCEPTANCE 40
List of figures
FIGURE 4.1 V-MODEL 21
List of tables
TABLE 2.1 TOOLS USED IN LITERATURE CASES 14-15
TABLE 3.1 LIST OF PROPERTIES OF INTERVIEWEES 17
5
1. Introduction
For many years companies applied Lean manufacturing and Six-Sigma as separated approaches for continuous improvement. Toyota Production System (TPS) introduced Lean production in Japan around 1950 and from there is originated Lean manufacturing (Shah et al. 2007). Liker et al. (2000) defined Lean manufacturing as (p.82) “a philosophy of manufacturing that focusses on delivering the highest quality product at the lowest cost on time”. The concept of Six Sigma is originated by
Motorola Inc. in the USA, about 1985. Linderman and Schroeder (2003, p. 195) defined Six Sigma as “an organized and systematic method for strategic process improvement and new product and service development that relies on statistical methods and the scientific method to make dramatic reductions in customer defined defect rates.” In their overview of Six Sigma, Montgomery and Woodall (2008) (p. 343) state that organizations could reach high levels of process performance by the simultaneous and optimal implementation of Six Sigma, DMAIC (Define-Measure-Analyze-Improve-Control)-project-cycle, Design for Six Sigma (DfSS) and Lean.
Timans et al. (2014) highlight that Lean and Six Sigma are not independent continuous improvement approaches. This happens because waste, variability and poor flow between processes may weaken the process performance; though poor process performance might generate problems in the flow between processes and as a result waste and variability might occur. Therefore the employability of Lean and Six Sigma simultaneously and in an integrated way could lead to the identification of the major causes of poor performance. The combined terminology of Lean Six Sigma (LSS) has been introduced around 2000 and Shah et al. (2007) (p.801) state that “Lean-sigma is being forwarded as a
management philosophy based on integrating lean production principles and practices with Six Sigma tools”. Examples of tools used in a continuous improvement project for product and process design involve: Gage R&R, Benchmarking, QFD (Quality Function Development), CTQ flowdown (critical-to-quality), DoE (Design of Experiments), Pugh matrix, FMEA (Failure Mode and Effect Analysis) and Robust Design.
6
project, but also make the problem even worse (Hagemeyer et al. 2006). Consequently, the appropriate usage of tools in a DfSS project is unclear and highly important. Thus, it is crucial to deploy the most suitable tools and techniques, at each stage of a DfSS project in order to successfully accomplish it. As DfSS is not a well-documented area of research, this study can build knowledge on how to use the most important tools in each stage of a DfSS project.Bañuelas and Antony (2004) state that Six Sigma basically focusses on decreasing the potential variability from processes and products by the usage either of a continuous improvement methodology or a design/ redesign approach. Antony (2002) highlights that companies which have implemented the six-sigma methodology realized that top-quality can only be reached, when the product design is at a high level and this can be achieved by DfSS projects. Shahin (2008) defined DfSS as (p.49) “a
powerful approach to design products, processes and services in a cost-effective and simple manner to meet the needs and expectations of the customer while driving down quality costs.” The author claims that the utilization of helpful statistical tools is involved in order to predict and improve quality before making prototypes. Therefore, when the proper tools are deployed, the DfSS approach can build quality into design and at the same time reduces the need for later inspection and reworking of the product.
While in industry, LSS projects at the production phase are usually carried out following the steps of the DMAIC-project cycle, in the design/redesign phase a different approach is deployed with different steps in a DfSS-project cycle. Chung and Hsu (2010) state that the DfSS activities have a positive impact on the performance level of product development. Sokovic et al. (2010) demonstrate that for the process design the DMADV (Define-Measure-Analyze-Design-Verify) methodology is suitable. As for the design of new products, the authors suggest the IDOV (Identify-Design-Optimize-Verify) methodology. There are more acronyms describing similar methodologies but according to Park (2003) IDOV is one of the most popular for NPD (New Product Development) processes. Antony (2002) suggests that IDOV is a methodology which makes the introduction of new products more efficient and reliable therefore it is an adequate method for meeting high customer requirements. The DMAIC-cycle is mainly a problem solving method and there are hardly examples of such projects applied in NPD.
Sokovic et al. (2010) demonstrate the high importance of selecting the appropriate tools and
7
studies carried out using the DMAIC-project cycle, there are significantly much less related to DfSS projects. Therefore the following research question is formulated:-How Lean Six Sigma tools are used in Design for Six Sigma projects?
This research focusses on exploring the usage of tools in DfSS projects, following the IDOV or similar methodologies which are applied in LSS manufacturing companies. Descriptive-exploratory case-study research has been conducted in order to explore and build knowledge on what tools are used in the projects following the IDOV methodology and similar methodologies, which tools are used practically in all the cases and what are the selection criteria for these tools. A single case study within a Dutch manufacturing company working in LSS environment has been conducted. Semi-structured interviews have provided the data to base the findings of this study.
8
2. Theoretical Background
Within this section, a main generic idea about Product Design Processes (PDP) is presented and after that the prominent approach for SS-projects is illustrated, that is the DMAIC-project-cycle and why this is not a suitable methodology for NPD processes. Next to that, a more appropriate methodology for NPD processes is demonstrated that is the DfSS-approach and specifically the IDOV methodology. Then the research question is going to be refined into sub-research questions. In the end of this
section, table 2.1 is demonstrating the tools described in DfSS projects found in the literature as well as the assignment of these tools per each stage.
Generic Product Design Process
Within this sub-section it is described a generic product design process (PDP). Cooper and Kleinschmidt (1986) structured the process of designing new products into stages with different deliverables. The four basic stages or phases are (Pahl and Beitz, 1996): Planning and Clarifying the Task, Concept Design, Embodiment Design and Detail Design.
Planning and Clarifying the Task should generate a needed product idea and also it should be in accordance with the present market situation, needs of the company and economic outlook. This product idea should be accessible before the beginning of the product development project (Pahl and Beitz, 1996).
Pugh (1996) breaks down the Concept Design phase into two consecutive components which are: Concept Generation and Concept Screening and Improvement. According to Thorton (2004) the Concept Generation stage leads to rough design layouts. Within this sub-phase, many concepts with several solutions should be generated. All those concepts are assessed in a screening process. Ullman (1997) demonstrates that concepts not meeting customer needs are screened out whereas the remaining concepts are developed even more. Pugh (1996) illustrates that techniques for generating and
evaluating concepts are used constantly until a winning solution comes up for Embodiment. According to Hasenkamp (2010, p. 319) “The Embodiment Design defines the arrangement of
assemblies, components and parts, as well as their geometrical shape, dimensions and materials.” Pahl and Beitz (1996) stated that usually certain embodiment designs are needed before a specific design could emerge. The design developed in this phase should be continuously improved in the next phase (Hasenkamp, 2010).
Hasenkamp et al. (2007, p. 354) state that “In Detail Design the arrangements, forms, dimensions and surface properties of all the individual parts are finally determined, the materials are specified,
9
produced.” The outcome of the Detail Design stage is a specification of production (Pahl and Beitz, 1996).LSS stages with DMAIC
DMAIC is the dominant problem solving approach in Six-Sigma projects. Jirasukprasert et al. (2014) highlight that DMAIC-cycle simulates the Deming’s continuous learning and process improvement model PDCA (Plan, Do, Check, Act) whereas Watson et al. (2013) state that PDCA has been replaced by the Six Sigma problem solving method DMAIC. Sokovic et al. (2010, p.480) state that the DMAIC cycle is a data-driven approach to Six Sigma projects for improving processes. The authors specify that DMAIC-cycle “is systematic and fact based and provides a rigorous framework of results-oriented project management”. Shahin (2008) states that (p.51), when DMAIC-cycle is deployed, “Six Sigma teams tend to achieve constant incremental improvements by reducing or minimizing the causes of variation in the existing processes.” DMAIC-cycle is an integral part of LSS and it is the acronym for five interconnected phases which are: Define, Measure, Analyze, Improve and Control (Sokovic et al., 2010; Jirasukprasert et al., 2014).
De Koning et al. (2005) provide the basis of the rational reconstruction of the functionality of these phases by combining selected sources. Specifically, the generic definitions are described of all the five stages which are the following (p.773):
-Define: Define the problem needs to be solved, incorporating customer impact and potential benefits.
Generic definition: Selection of the problem and analysis of the benefits
-Measure: Determine the critical-to-quality characteristics (CTQs) of the product or service. Validate the capability of the measurement. Measure the existing defect rate and establish improvement goals.
Generic definition: Transformation of the problem into a measurable mode, and measurement of the existing situation.
-Analyze: Recognize the root causes of why defects happen; diagnose key process variables which cause defects.
Generic definition: Recognition of factors and causes which affect and determine the CTQ’s behavior.
-Improve: Define the way to intervene in the process in order to greatly reduce the defect levels.
10
-Control: Implement current measures and actions in order to sustain improvement.
Generic definition: Ensure the sustainability of the improvements by adjustments on the process management and control system.
However, the DMAIC-cycle is the best project-cycle for improvement of quality in the production phase, and is less fit for usage in the design phase. Measures to improve quality in the design phase have the largest impact on the final product but those measures are harder to identify. Alternatively, measures to improve quality discovered after the design phase, although are easier to identify, are often costly to solve(Berryman, 2002; Hahn et al., 2000). Consequently, a low design-quality cannot be repaired in the production process but it brings more problems in the production process. Hoerl et al. (2010) indicate that DMAIC is not the best approach for improving process beyond its design capabilities. Besides that, Timans et al. (2014) state that in LSS implementation, the major weakness of use of the DMAIC-project-cycle is that these projects are mainly focused on finding solutions in existing problems, but not on designing completely new products such as projects aiming at a NPD.
DMAIC-project-cycle mainly focusses on reducing defect rates in existing products and processes. Hahn et al. (2000, p.319) state that “applying standard SPC (Statistical Process Control) approaches to a manufactured product is likely to lead to only a fraction of the possible improvement”. Therefore, the nature of designing something new requires a totally different approach than fixing something which already exists. According to Koch et al. (2004), it has been recognized that the quality or performance of a product is dependent on earlier design decisions. Consequently, the limitation of DMAIC-project-cycle involves activities related to find and fix problems in existing processes whereas DfSS focusses on designing error-free processes, when fixing the existing processes would not lead the quality into the desired level.
The stages of the DMAIC-cycle do not fit to the stages of a project in which a completely new product has to be designed. Specifically, Define stage could be used in the design project. However, Measure and Analyze stages of DMAIC-cycle are related to the context of existing production processes. As it is highlighted in the generic definition of the DMAIC-cycle of De Koning et al. (2005), Measure stage is related on an existing situation and on Analyze stage the recognition of the root causes showing why defects happen is crucial. Therefore the above mentioned stages are trying to solve problems on existing processes/ products. The IDOV project-cycle is specifically fit for new product development. This will be explained in the next section.
11
IDOV is a project-cycle for DfSS projects. Watson and De Jong (2010, p.13) defined DfSS as “a process to define, design and deliver innovative products that provide competitively attractive value to customers in a manner that achieves the critical-to-quality characteristics for all the significantfunctions”. Edgeman and Duncan (2008) illustrate that DfSS is used when new designs or significant redesigns of products and their processes are needed in order to meet the customer expectations from the beginning. Bañuelas and Antony (2004) show that DfSS using the IDOV methodology aims to predict and improve quality before products are launched. It can also be used to redesign existing products. The authors claim that the IDOV methodology can improve product and process
effectiveness apart from efficiency and it also helps to bypass potential problems at manufacturing stages. Besides, the understanding of customers’ needs is increased and the quality of the product is enhanced. According to Ericsson et al. (2014, p.649) “The general idea of DfSS is to systemize the development process from the initial customer survey to the final product release by requiring the application of specific tools.” This research focusses on the procedures and tools used within DfSS projects that aim at NPD. IDOV is the appropriate methodology to build quality into the design phase of NPD processes. Therefore, this research aims at exploring the usage of tools within IDOV and similar methodologies.
Montgomery and Woodall (2008) demonstrate that DfSS could take the variability reduction and process improvement upstream which means from manufacturing into the design process, where new products are designed. Park et al. (2003) state that IDOV methodology has been suggested by General Electrics and it is practically used most often. IDOV tends to make products that are (Harry and Schroeder, 2000, cited by Bañuelas and Antony, 2002, p.251):
-“resource efficient;
-capable of reaching very high yields, regardless of complexity and volume; -robust to process variability; and
-highly linked to customer demands.”
The four stages of the IDOV methodology are described below (Antony et al. 2002, p.7-8):
12
-Design: After the recognition of the design parameters by the organization, these shall be transformed into the actual effective design. On this stage, companies should analyze the design requirements and parameters, as well as their relationship with CTQs. Specifically the organization should research the relationship between the design parameters and CTQs at sub-levels in complex systems. Besides, design alternatives should be presented and the best design option should be selected.
-Optimize: Within the third stage, the organization is trying to be certain that the product can be manufactured within the identified design parameters and the defined budget. The identification of sources of variability and the minimization of product performance sensitivity to all the sources of variation are involved. Furthermore the tolerance design is applied for critical design parameters as well as the optimization of the design for manufacturability. At the end, the design capability related to the design specifications is determined.
-Validate: Within the final stage, the completion and validation of the process is controlled and also that the requirements are practically met. Specifically, the verification of the design is involved in order to secure that it meets the set requirements. The assessment in terms of performance, reliability and capability is also included. Finally the process control plan for the mean and variance of CTQs in production is developed.
Literature search for DfSS case studies
A search for case studies on the design of new products has been carried out to find information about the tools that have been used in those cases. The research has been conducted in international journals, books and magazines which are related to six sigma practices, continuous improvement methods and total quality management. The timeframe of the research is from 1995 and the selected case studies are situated in the context of international manufacturing firms. Four cases were found. Those cases demonstrate that big manufacturing companies apply the DfSS methodology. The first DfSS project was carried out in the home laundry appliance division of one of the largest suppliers of white goods in Europe (Bañuelas and Antony, 2004). In this project the IDOV methodology is followed. The second DfSS project has been carried out by Ford Motor Co. after the company has launched
13
RDIDOV methodology (Park, 2003). The team basically used the IDOV methodology, however RD stages (Recognize-Define) are added before IDOV. The two extra stages R and D are added as a preparation before entering the Identify stage, for the successful implementation of the DfSS project. Specifically, Recognize is the phase related to the preselection of the project and Define is the phase where business needs are clarified.It is hard to find in literature typical case studies which are explaining the tools applied in the design process of new products. Table 2.1 demonstrates the tools that are used in the four literature case studies. The table presents all the tools that are used per stage of the project and per company as well. There are some tools which have been used in more than one case. Specifically, Quality Function Deployment (QFD) has been deployed in all the case studies. QFD is a tool responsible for identifying and quantifying the requirements of the customers and translate them into key design requirements. In all the literature cases, this tool has been used at the Identify or Define phase. The following tool which is used in all the case studies is benchmarking. Six-Sigma Benchmarking is a practical and cost-effective way in order to introduce best practices at a company. In all the literature cases, this tool has been used at the Identify phase and in the case of Ford it is used at Define phase of the DCOV-cycle. Robust Design (RD) is the next method used in most of the cases. Hasenkamp et al. (2007) state that the core of RDM (robust design methodology) is to design products which are insensitive to sources of uncontrollable variation. This method has been used at the Design and Optimize phase. Design for manufacturability (DFM) is a methodology which has been used in most of the cases. DFM is focusing on designing a product with the aim of facilitating the manufacturing process in order to reduce its manufacturing costs. DFM is applied at the Optimize and Verify phase. The next tool which is used in most of the cases is Design of Experiments (DoE). Hagemeyer et al. (2006) define DoE as (p.465) “a systematic set of experiments that permit the evaluation of the effect of one or more factors on a response.” DoE is a formal statistical tool which helps to ensure that the testing phase of the project produces data that would be beneficial for further improvements to the process. The two kinds of DoE are full factorial and fractional factorial. In the case of Bañuelas and Antony (2004), this is a full-factorial DoE. As for the rest of the cases, there is no information about the kind of DoE.
The main research question has been expanded to four research subquestions. This happens in order to illustrate the tools described in the literature cases, the tools used in practice and whether the results from literature support or not the IDOV stages described by Antony. Given these points, the main research question is expanded to the following research questions:
- What are the tools that are used with the IDOV or similar methodologies, described in all or most of the literature cases?
14
- What are the reasons/objectives behind the tools usage in practice?- How the results from literature and practice support or not support the IDOV stages described in Antony’s article (2002)?
Table 2.1, Tools used at literature case studies
Stage Tools Description of the tool
White Goods Europe (Bañuelas and Antony, 2004) Ford (Soderborg, 2004) Samsung SDI (Park, 2003) R&D Samsung (Park, 2003)
Identify Team charter
First step in SS methodology. The team charter can break down a successful project.
Identify/Define QFD
Tool used to identify and quantify customers' requirements and translate them into key critical parameters.
Identify/Define Six Sigma Benchmarking
Practical and cost effective way in order to introduce best practices at a company.
Identify CTQ flowdown
Specific method to translate customer requirements into technical specifications. Identify/ Define VOC (Voice of the Customer)
Statement made by the customer on a particular product/service.
Identify/Characterize/OptimizeFMEA Tool used to mitigate potential risk. Characterize/Identify Measurement System Analysis
Experimental and mathematical method of determining how much the variation within the measurement process contributes to overall process variability.
Identify Gap analysis
SS quality control tool that compares actual performance with the potential performance of a business.
Define Market research Gathering information about customers, so company will be able to produce and promote products/ services suitable for such customers. Define Brand analysis Analysis and planning on how the brand is perceived by the market.
Define Noriaki Kano model
Theory of product development and customer satisfaction which classifies customer preferences.
Define Quality History: survey
Historical data demonstrating quality by surveys from the customer
Define Quality History: repairs
Data coming from customers who are sending back the appliances to the company to repair that.
Define Quality loss function
Graphical representation of how an increase in variation within specification limits leads to an exponential increase in customer dissatisfaction.
Design AHP Decision making technique for design decisions.
Design Regression analysis
It allows predictions to be made based on the presence of a mathematical relationship between two variables.
Design/ Verify Baselining The average long-term performance of an output characteristic (Y) when all the input variables (x) are running in an unconstrained fashion. Design/ Verify Rational subgroups It identifies and seperates special causes of variation. Design/Optimize Design for assembly Process by which products are designed with ease of assembly in mind. Characterize-Optimize/Design Robust Design Methodology
Aims at designing products insensitive to sources of uncontrollable variation.
Design Capability flow-up
It is controlled if the technical requirements are actually met or not.
Design/Optimize DoE/ Full factorial DoE/ DoE &ANOVA
15
Continue of Table 2.1Stage Tools Description of the tool
White Goods Europe (Bañuelas and Antony, 2004) Ford (Soderborg, 2004) Samsung SDI (Park, 2003) R&D Samsung (Park, 2003) Design Geometric Dimensioning& Statistical Tolerance
System for defining and communicating engineering tolerances.
Design CAD Software for detailed engineering to design products.
Design Cause & effect matrix A matrix to understand and correlate customer requirements to process input variables Design Simulation, capability study Simulation is the imitation of the operation of a real-world process/system over time.
Design Tolerance design
A method to set up proper limits to design parameters.
Characterize P-diagram It is used to illustrate interplay of different parameters.
Characterize Axiomatic design
Systematic approach ensuring that design factors do not affect several product characteristics simultaneously.
Characterize Dimensional variation analysis
A process which simulates the effects of component part and assembly variation on a system allowing the probable dimensional variation behaviour of the entire system to be determined.
Optimize Response surface methodology
It employs experimental design to discover the shape of the response surface and then uses geometric concepts to take advantage of the relationships discovered.
Optimize Multiple response optimizer
When the optimization procedure involves more than one response, it is not possible to optimize each one in a separate way. In the optimization of a process, the overall solution must be included in an optimal region, leading to a certain degree of compliance with the
proposed criteria for each variable of the system. Optimize TRIZ (Theory of Inventive Problem Solving)
It is a methodology for stimulating and generating innovative ideas and solutions for problem solving.
Optimize Gage R&R
Statistical tool which measures the amount of variation in the measurement system. The measurement variation should be expressed as a fraction of the total observed variation Optimize/Verify Design for Manufacturability
It is the general engineering practice of designing products in such a way that they are easy to manufacture.
Optimize Parameter design Methods to determine the best process parameter set-points (using DoE) Optimize
Tolerance design/ Statistical
tolerancing Method to determine tolerance specifications. Optimize/Verify- Validate Process Capability Assessment
It demonstrates the short-term and the long-term capability of the processes.
Validate Process Mapping It helps a team approach improvements with a common understanding of a process.
Verify/ Validate Reliability Study
Mathematical methods to predict the reliability of the product and to identify the right methods for maintenance.
Identify/DCOV DfSS Scorecard
16
3. Methodology
Defining the method
This research uses the case study research methodology in order to investigate how tools are applied in DfSS projects. Specifically, the aim of this research is to explore uncovered areas and Karlsson et al. (2016) state that case studies can be used for exploration. Ketokivi and Choi (2014) highlight that in the early stages of research programs, exploration is needed to develop research ideas and questions. By exploratory case study, a list of research questions and areas can be generated in order to gain deeper understanding as there are not many references available on DfSS case studies. Consequently, case study methodology is suitable for this study which will reveal new knowledge on the usage of tools in DfSS projects and on the prerequisites to proceed to the next project-phase.
Research Setting
The setting which has been applied is in LSS environment specialized in DfSS project. Company A is a manufacturing company in the Netherlands which uses this kind of methodology in the design of new products for household and personal care. This research is investigating how tools are applied in the design phase of new products, as it is rather unclear what in design processes the selection criteria of the tools usage actually are.
Case selection
The design department of company A was willing to contribute to our investigations. The environment of this department is following the LSS methodology and it is experienced working on DfSS projects. Table 3.1 below demonstrates the properties of the interviewees in terms of educational level, years at this organization, years at this company and belt level on DfSS. The engineers of the design
17
has been conducted because the availability of companies that apply the DfSS methodology on a sufficient level in the Netherlands is very limited.Table 3.1, Properties of the interviewees
Data collection
This research is based on semi-structured interviews with 4 BBs and 3 GBs, working at the advanced development (AD) department, and new product innovation (NPI) department of company A. A well-structured research protocol has been constructed based on the recommendations of Blumberg et al. (2014) which serves as a check-list for the interviewer during the interviews. The interview protocol was conducted with a fellow student who performed a different research project in relation to DfSS. By this, the reliability and validity of case research data is improved. The core of the protocol is the set of questions which have been used during the interviews. The general questions in part A of the interview are used for both researches, part B is related to this research whereas part C questions are used for the other research. Specifically, part A focusses on general questions about the properties of the interviewees. Questions of part B mainly focus on the tools that are used at each stage of the project, whereas part C aims at the relationship between DfSS projects and lean product development.
Interviewee Educational Level
Years at the Organization
Years at the
same Position Level on DfSS A Mechanical Engineering
Master in Mechatronics 8 months 8 months GB B
2 Master in Mechanical Engineering and
Industrial Design 3.5 years 2.5 years GB C Master in Mechanical
Engineering 18 years 6-7 years BB
D Master in Mechanical
Engineering 12 years 5-6 years
BB(* Training the new GB trainees.) E
Bachelor Mechanical Engineering (University
of applied sciences) 28 years 7 years BB F
Communications (University of applied
sciences) 4 years 4 years GB
G
Bachelor Process and production control (University of applied
sciences) 15 years 15 months
18
The protocol is included in “Appendix A: Interview Protocol”. The results of it are going to bediscussed in the next section.
The interviews aim at exploring the criteria for the tool’s usage in each stage of the DfSS project. The interviews were held at the working location of the interviewees and lasted around 80 minutes. An outline of the protocol has been sent in advance to the interviewees so that they could be well prepared. Apart from this, the interviews have been recorded after receiving the relevant permission. After that, the interviews have been transcripted accurately into text and they have been sent back to the interviewees in order to be checked on completeness and potential errors. Thereafter the
interviewees have been contacted by follow-up e-mail to confirm that the interview transcripts were correct or that changes were needed.
Analysis
19
4. Results
This section presents the interview results from company A. Each subsection demonstrates a phase of DIDOVM (Define-Identify-Design-Optimize-Verify-Monitor) methodology. Company A executes NPD projects using the DIDOVM methodology which is similar to IDOV. The last subsection
illustrates some of the most inspiring elements and prominent tools used with DIDOVM methodology. DIDOVM scheme is presented at table 4.1. This material has been received by the company before the interviews. This is a translation of DIDOVM methodology in a generic way. The main deliverables and the most prominent tools of each phase have been included. Some of the tools mentioned in the scheme (Table 4.1), did not come up during the interviews. Even though those tools are not directly connected to NPD processes, they are included in the table in order to demonstrate a more completed view of the tools used in a DfSS project.
Define
Define phase is responsible for the setup of the project plan. It is on a higher-level management therefore stakeholders should make agreements about business results, required resources and team composition within the projects, with the deadlines and milestones included.
Table 4.1 shows the starting point of this phase, called “roadmap”. Roadmap is the first step to define the project. It is based on what should be accomplished in a long-term and which projects need to be characterized in order to achieve that. For instance, an electronic platform for the next generation of an appliance can be built as a short-term solution. But it is a contingency that this platform cannot be reused for after next generation. So if a road-map would have been set up first, an electronic platform for more generations might have been built from the beginning.
The following important tool mentioned at table 4.1 is the so called SIPOC (Supplier-Input-Process-Output-Customer) diagram. According to interviewee A and F this tool is mainly used by BBs. It is used in order to make visually clear in a complex project, how a product is build up and how the parts of it are assembled. The inputs to the process come from the suppliers of the process. The customers are those who receive the process output. The inputs are necessary to reach the desired output, so SIPOC can give insight on the requirements of the suppliers and the customers. In that sense,
according to interviewee G: “SIPOC diagram contributes to reach a narrowed scope of the project.”
20
definition is planned. Within the Identify phase, an extended analysis of all the customer’s needs is carried out where those needs are translated into CTQs of the project.The interviewees A, B, C, D and G confirmed that the whole project plan is the most critical delivery of this phase. High level requirements should be defined which are related to what needs to be achieved. Furthermore, time-plan of the project, team composition, high-level risks and financial resources need to be characterized. When the above mentioned issues are agreed upon, the project can proceed to the main deliverable which is the project sign-off. To quote interviewee G: “Project is
signed by all the stakeholders who sponsor it and then the project proceeds to Identify phase where there is in depth focus to the requirements and the technical details of the project.”
Identify
Within this phase there are two types of customers. Customer could be the end user but also could be the retailer. Therefore, the first roadblock needs to be passed is the identification of the fulfilled requirements in order to reach the retailers’ shelf. Next to that, there are certain requirements of the end-user which need identification. The block of customer needs identified & prioritized (Table 4.1) starts with analyzing previous reviews, doing consumer interviews, doing home visits and checking how consumers use the product. There, all the information available about product rating and reviews is analyzed. There is also attention to customer complaints files. This is when people are sending back the appliances to the shop where they bought them. These complaint files are collected by company A and they are investigated on what went wrong. In the case of company A, usage environment is the bathroom. The product research center department has a lot of historical data from consumers and that input is useful to define what the user wants. Consequently, all those inputs from business,
stakeholders, retailers and users should be included into the project.
Within this phase a critical tool is the CTQ flowdown. To quote interviewee E: “CTQ flowdown is
used at all the DfSS projects of company A.” CTQ flowdown is a specific method to translate customer
requirements into technical specifications. It begins as high level needs and those needs are translated into specific measurable customer requirements. It is applied within the V-model in order to
21
CTQ flowdown is reached, focus is assigned to the capability flow-up in order to validate the receipt of information back. Within the capability flow-up it is controlled whether the requirements are actually met or not. System test is done by determining process set-points and tolerances. Within CTQ flowdown is critical to clarify which are the end-CTQs that are going to be delivered.Figure 4.1, V-model
Gage R&R is the next tool used in all the projects. Gage R&R which stands for repeatability and reproducibility is a statistical tool which measures the amount of variation in the measurement system in relation to the total observed variation. Within company A there are two types of Gage R&R (Table 4.1, I-phase and V-phase). The first type is used for test and development (T&D) equipment. This equipment is used during the development stage of a product. The second is Gage R&R Industrial which follows the same methodology but there are much more accurate requirements on measurement equipment related to the production release of the product. Repeatability means when the
22
within 100 microns and there is a R&R of 10%, the measurement variation should be less than 10 microns.Table 4.1 DIDOVM scheme
Phase Define Main Delivera ble Project identified with sign-off Main Tools
used Vision Strategy Roadmap SIPOC
Pareto Charts/ Histograms Gantt and PERT chart Phase Identify Main Delivera bles Customer needs/ CTQs/Specs Main Tools used Survey/ Focus
groups Kano diagrams
CTQ flowdown- QFD1
Gage R&R (for T&D equipment) Design Dashboard Phase Design Main Delivera bles Design complete/ Predicted capability Main Tools
used Pugh Matrix
Benchmarking/ Historical regression
CTQ flowdown- QFD2/ Design
Dashboard DoE/ Simulations FMEA
Sensitivity Analysis Phase Optimize Main Delivera ble Targets met on predictions/ Prototypes Main Tools used Sensitivity Analysis Multiple Response Optimization Robust Design Methodology/ SPSS/ Monte Carlo simulations/
23
Design Dashboard is the next tool used in all the projects at A. This is the build-up of all the CTQs and according to interviewee G: “Design Dashboard provides information to all the departmentswhere the CTQs are.” There is decided how each CTQ is going to be realized properly, which process
steps should be executed and which are the underlying CTQs that need to be delivered. To quote interviewee B: “the Design Dashboard should translate into certain key values all the information
generated in a project which could easily be tracked by the higher level management.” Within table
4.1, Design Dashboard is developed during the Identify phase and it is mainly utilized at Design and Verify phases. Within Identify phase a prediction is given about the design capability because the design has not been finalized. Within Design phase, the final product has been designed whereas within Verify phase, it is verified whether the final design can really be produced without problems according to specifications. In the end, Design Dashboard with all the CTQs are translated in a measurable way for the production.
The main deliverables in order to proceed to the Design phase (Table 4.1) is that the user requirements should be precise which means the value proposition has been identified and translated into low level CTQs. Therefore the customer needs, Design Dashboard and the CTQ flowdown have been built up before proceeding to the Design phase.
Design
During this phase, it should be explored not only which is the best way to make a certain product but also a selection should be made among alternative design concepts. Several tools and techniques are used within this stage. Those tools are presented at table 4.1. The tools which are used at all the projects independently of the context are: FMEA, CTQ flowdown and Design Dashboard. Besides, most of the times Pugh matrix is used. Benchmarking is used when you have data from a predecessor appliance or a competitor. DoE technique is also used in all the substantial projects, substantial in terms of resources, time needed and profit. For simple and small projects, it is not necessarily used.
24
outcome, the project goes just one step back, instead of returning to the beginning”. The Pugh matrix
tool can also be used for supplier selection.
In a deeper level within the CTQ flowdown, it is demonstrated the relationship between high level requirements and underlying requirements. DoE is used to prove design capabilities during the
development of a new design. It is linked to the CTQ flowdown because combing those tools, it can be proved that the project is carried out as defined. In DoE it is needed to have representative samples, otherwise it does not make any sense. To quote interviewee C: “With the use of DoE, insight on the
expected capability could be achieved without making a lot of effort”. Sometimes simulations are
made instead of the use of DoE. Simulations are used when there are not enough practical data. DoE and simulations are both tools to gain insight on the expected capability in order to comply with the requirements.
The kind of DoE used depends on the project. Generally, full-factorial is preferable but sometimes there are so many parameters thus it would take too much time to make a full-factorial DoE.
Subsequently, the most important parameters should be selected. When the project is substantial and there are too many parameters, statistics specialists are hired in order to create a reduced DoE. The complexity of DoE depends on the number of parameters. The main software packages which are used for the application of DoE techniques are Minitab, tolerance design and ANOVA.
Within this stage, FMEA is an important tool. There are different kinds of FMEA. Specifically there is User-FMEA which is used to investigate what can go wrong when the consumer uses the product. There is Design-FMEA which is focusing on functions that the product has to fulfil. Finally, there is Process-FMEA which is used to investigate potential risks related to the production process. In the Design phase is more related to user and design FMEA whereas in the Optimize phase is more related to Process FMEA. In every phase FMEA should be updated as it is one of the most important tools used in process development. After a certain point of the project, FMEA is included in the to-do-list. It aims at mitigating risks. It is updated at least every month whereas in some projects there is FMEA session weekly. If there is a very high risk at the Design phase and the project proceeds to Optimize phase, the risk is transferred to the following phase.
Table 4.1 indicates that at the end of this phase, the predicted design capability should have been determined. All the CTQs should be agreed and the design should be fixed. Capable design is needed, which means all the high risks not only have been identified but also they have been mitigated to a lower level. To quote interviewee E: “If the right design is done from the beginning, and its
25
OptimizeWithin this stage it is estimated how much defectives are made, like D.P.M.O. (defects per million of opportunities). This is a way of addressing quality standards and within company A, this is normally expressed by Cpk figures. Cpk stands for process performance capability and it demonstrates the short-term capability (capability assessment, table 4.1). Next to that a sensitivity analysis has to be carried out. Sensitivity is always an issue. According to interviewee A:“Within Design phase
sensitivity analysis is related to higher level CTQs of a certain design. Within Optimize phase, there is already a certain design prototype. The purpose is to learn how sensitive the design is for noise influences that impact on lower level CTQs”.
Often it is necessary to make optimizations simultaneously on more than one CTQs (Multiple Response Optimization, Table 4.1). Few things can be done. Firstly there is consultation with the stakeholders in order to be informed about the new situation. It is investigated how specifications could be fulfilled. Root cause analysis is done and depending on how big the mismatch is, it might be decided either to change the specifications or to fulfill them. To quote interviewee E: “It is discussed
whether that tight specification is necessary. This happens by going back to the internal customer which in mycase is the innovation department and discuss if that tolerance is crucial or not. So after doing risk assessment and depending on how important that specification is, the requirements could be modified”. Specifications on micro-level make it expensive to meet them. The more tolerances are
sharpened, the more expensive it will be.
Next to that, there is the robust design methodology which is often used because of the tight requirements. Normally, designers are working on that and it is expected to be used at the Design phase instead of Optimize. Firstly this is more about robust processes which means that processes stay away from spec limits. When the processes are robust, the production becomes robust too. For
example, there are tests for drop resistance (Highly Accelerated Life Testing, Table 4.1). A product is inserted into a turning barrel which is rotated for a number of cycles. Depends on how the product looks like afterwards, failure modes are assessed and they are tried to be reduced. Minitab is the main statistical software tool used for RDM. Also SPSS, Monte Carlo simulations and Excel can be used. Those are also important software packages used at a design project. Another tool which is used within this stage is mistake proofing. It is connected to the Industrialization department and is done by Poka Yoke mistake proofing. This is the use of any automatic method which either makes it
26
The deliverables of this phase are mentioned to the Table 4.1. Specifically, product needs to come close to the requirements. This means FMEA, test results from Optimize phase and all the necessary modifications have been done in order to bridge the gaps. At the end of the Optimize phase the quality levels of all the CTQs should be such that the specifications are met.Verify
In Verify phase, the goal is to ramp-up and provide the industrial release of the product. When the verification starts, it means that there is a proof that the requirements are going to be met because after that the products are tested at an independent quality lab.
Within this phase it is common to be confronted with non-normal probability distributions, when the first products are manufactured during pilot production. Causes of rejection could be found in the non-normality of the distributions. So then, agreements are needed on how to deal with that. Sometimes it is done automatically in Minitab. Minitab assesses normality and it shows if a Box Cox transformation is needed. Box Cox transformation is a method to normalize a data set so that statistical tests can be performed to evaluate it properly. But it is due to the nature of some CTQs that they have non normal distribution. When a technical CTQ is related to pull-force and pull-forces are measured on a large sample of products then the distribution cannot be normal. Normality would mean that there could be a minus pushing force which is impossible. Subsequently, it is not always needed to have normal probability distributions because when they are transformed, it is difficult to interpret them.
Monitor
A NPD project starts with the advanced development department, and then it is progressed by the new product innovation department. At a certain point industrialization continues and after a certain moment the machines and the production processes are stable enough which means what comes out of the production line is complying to quality requirements. To quote interviewee G: “Monitor phase is
performed by the production and it is checked if the production is running well over time”. Besides
this, within this phase market feedback is monitored which reveals the satisfaction level of consumers.
General results
27
which are assembled to one module and production processes. After all these, Optimization and Verification phases continue. On each level it is checked if the requirements are met and in the end you have the product which fulfills the needs of the business and the different customers.Regarding the results of the DfSS training, it seems that after the certification project, BBs do not carry out any other project completely in the same way as has been done to reach BB or GB certification. However parts of the DIDOVM methodology are embedded in their regular way of working.To quote interviewee C: “In company A for really new products we have the NPI (new
product innovation) department. That means we have our standard way of working which is linked to DIDOVM methodology and it is embedded on that.”
The tools are highly important because the users know which tools could be used to tackle certain problems. The tools as such were already there, but when they are applied with this logical methodology, then they could also be re-used for the next generations of a product.
28
5. Discussion
The discussion section is structured to provide answers to the RQs. The aim of this discussion is to show the most prominent tools described in literature and in the case of company A. Next to that, the main differences on the tools usage between the literature and result section are going to be presented. The last subsection of this chapter is discussing whether the findings support or not the IDOV
methodology introduced by the article of Antony’s (2002).
Tools used with the IDOV or similar methodologies described in all or most of the literature cases
There are a few tools which have been observed in all or most of the literature cases of IDOV or similar methodologies. QFD which is applied at the early stages of the DfSS projects, which means at Identify or Define phase. Benchmarking has been observed at all the literature cases at Identify or Define phase. Robust Design is the methodology applied in most of the cases at Design and Optimize phase. Furthermore, Design for Manufacturability is used in most of the cases at Optimize phase. The literature cases demonstrated that this is a tool mainly used at the production phase. The last technique which is deployed in most of the literature cases is DoE. This technique is used at the phases of Design and Optimize.
Tools used in practice
As far as what are the tools used in practice, how these tools are used and what are the objectives of the tools usage, there are several tools which are used in all the DfSS projects of company A. Specifically those are: CTQ flowdown which translates the customer needs into technical requirements. QFD which helps translating high level customer demands into lower level design requirements. V-model which creates and controls the relationship between CTQ flowdown and capability flow-up (figure 4.1). Design Dashboard which is demonstrating that the whole CTQ flowdown should be performed according to the plan. Gage R&R that quantifies the measurement error in relation to the total observed variation, in the measurement system. FMEA is a key tool within DfSS projects which focusses on assessing and mitigating different kind of risks. This tool is aligned to the DfSS core principles because the latter a risk could be assessed, the more costly it would be to mitigate it. The last tool which is deployed to all the substantial DfSS projects at company A is DoE. It is a tool to gain insight on the expected capability. According to the input that can be gathered, DoE or simulations are used.
29
In a generic way, DIDOVM methodology is complying with the characteristics of IDOVmethodology. There were no clear deviations from Antony’s article but the information received from company A provides much more in depth information about DIDOVM compared to Antony’s article (2002). Specifically Antony, describes that within Identify phase the customer requirements should be translated into key technical characteristics at a detailed level. This is supported by Company A where the main objective of this phase is to Identify customer needs and translate them into low level CTQs. This is achieved by the development of the CTQ flowdown within the V-model and the Design
Dashboard which is the built up of all the CTQs. Following that Antony (2002) outlines that within the Design phase those key technical characteristics should construct the actual effective design. This is achieved by building relationship between design requirements and CTQs at a low level. In addition, design alternatives are needed in order to select the best one. Company A implements this phase in a very similar way where there should be an agreement at all the CTQs which leads to the predicted design capability. It is illustrated that in order to achieve that all the high risks need to be identified and mitigate them. Furthermore, the ranking and evaluation of design alternatives is achieved by the deployment of Pugh matrix.
Antony indicates that the objective of Optimize phase is that the product can be manufactured
according to the defined design parameters and budget. Variability should be identified and minimized in order to achieve robustness. Within this phase, company A focusses on all the essential
modifications to bridge the design gaps. The engineers target at minimizing the variation at all the critical design processes. The quality of the CTQs should be on the agreed level in order to meet the specifications. Lastly, within Verify phase of Company A, the similar characteristics of the Validate phase of Antony (2002) are supported. Both phases aim at meeting the set of requirements in order to verify the design.
DIDOVM methodology has two more steps than IDOV which are Define at the beginning of the project and Monitor at the end of the project.The Define phase is concentrating on the higher level business case and within this phase the project plan should be clarified. This means that all the relevant stakeholders should make agreements on the business needs, resources, deadlines and composition of the teams. On the other hand, Monitor phase is carried out in long-term gathering information if specifications are being met. In fact this long-term monitoring is done during regular production, after the product launch. CTQs are monitored in a long term and the performance is continuously compared with what has been specified.
Main differences in tools usage between literature case studies and company A
30
technique for design decision which can be used as an additional tool for Pugh matrix analysis.Company A uses Pugh matrix in order to rank and assess alternative design concepts. According to Cervone (2009), Pugh matrix is a simple decision support model which focusses on minimizing complex mathematical computations. In the research of Burgren et al. (2015), where AHP and Pugh Matrix are compared as optimal models, it is stated that Pugh matrix is so popular tool because it is easy to use and subsequently it saves time. Yet the use of AHP-methods could be very useful, adding information at a more detailed level.
Next to that, scorecards are used at most of the literature case studies. In the case of white goods Europe(Bañuelas and Antony, 2004) and Samsung SDI (Park, 2003) the scorecard prediction is built at Identify phase, whereas in the case of Ford (Soderborg, 2004), DfSS scorecard is used throughout the whole DCOV methodology. According to Antony and Bañuelas (2002), DfSS scorecard (p.26) “defines what to improve, by how much and how to accomplish the improvement”. It is also mentioned that the measures selected for the DfSS scorecard should translate the goals of the
company. The authors highlight that in the case of GE (General Electrics), the DfSS approach can be easier implemented by developing a strategy based on the DfSS scorecard. Then the DfSS scorecard should be broken down for product development and process development. In the case of company A, Design Dashboard is built during Identify phase where it is the prediction for design capability. Design Dashboard is the build-up for all the CTQs and that should be translated into certain key values on all the information generated in a project. Consequently, the reasons to develop and use of those two tools are the same.
Similarities in tools usage between literature case studies and company A
There are a few tools which are used in most of the literature cases and in company A as well. Specifically, QFD is the tool used in all the literature cases and in company A. This is a critical tool which is used in all the projects of company A. This is the practical tool used within the CTQ
31
any information clarifying what kind of FMEA used in literature cases. Comparing these with the interview results, at Optimize phase is mainly used Process-FMEA in order to investigate potential risks related to the production process. At the early stages of the DfSS project, Design FMEA and User FMEA are mainly used. The last key tool which is observed in one literature case and in company A is Gage R&R. In the case of Samsung SDI it is used at Optimize phase whereas within company A, it is used at Identify and Verify phases. From the interview results is learnt that the latter the stage this tool is used, the narrower the requirements on measurement equipment are.There are also few tools which are mentioned at Table 4.1 but did not come out during the interviews. It is understandable why no one mentioned Kano diagram, for example. Kano diagram is related to new features in customer preferences about the product. Probably this tool is mainly used by the marketing department, which is located at long distance from the manufacturing plant of company A. As for the Statistical Process Control (SPC, table 4.1), there are different methods in order to
continuously follow the process data distributions, following if the data are between their control limits and continuously calculating actual values for Cp and Cpk, Pp and Ppk (capability assessment, table 4.1).
Next to that, the results bring out that BBs do not carry out any other project in the same way as the BB certification project. A potential reason could be that, the BB-course is organized outside of the location of company A. Company A has its regular way of working and parts of DIDOVM
32
6. Conclusion
Returning to the main objective of this research, the most prominent tools of a DfSS project using IDOV or similar methodologies have been presented. Specifically the most important tools found both in literature cases and in practical findings are: CTQ flowdown, QFD, Benchmarking, Robust Design, DoE, FMEA, Gage R&R and Pugh Matrix. Within the previous sections, it is described how these tools are used and what are the objectives behind the tool’s usage. Furthermore, the main objectives of each phase of DIDOVM methodology used at DfSS projects of Company A are supported by the objectives of IDOV methodology introduced by Antony (2002). IDOV methodology brings structure to the project. Within each phase, the members of the project recognize what is needed to be done. Universal language among all the employees contributes to the successful implementation of the project.
Theoretical and managerial implication
This research started by combing both theoretical and practical need. One of the theoretical
implications of this research is the presentation of a different acronym for a DfSS project-cycle, which is the DIDOVM methodology. There is not any case in literature which describe this acronym. Within this research, it is described what is done at each phase of DIDOVM methodology. Specifically, the most prominent tools of each phase are explained and also what are the tollgates to pass to the following phase. DIDOVM scheme (table 4.1) contributes to the theory a different scheme in which business needs and monitoring in the production phase are taken into account. The main managerial implication is related to the alternatives on the tools usage. AHP could be combined with Pugh matrix in order to rank and assess alternative design concepts, their values and risks. AHP is a tool which could give more accurate results compared to Pugh matrix, on the ranking and assessment of
alternative design concepts. Next to that, DfSS scorecards can be built and deployed instead of Design Dashboard. The logic of those two tools is the same. Antony and Bañuelas (2002) highlight that if a strategy is developed based on the DfSS scorecard, then the implementation of DfSS approach would be easier.
Limitations
33
about the industry of the company. Company A which has participated in this research, is amanufacturing company which uses this kind of methodology in the design of new products for household and personal care. It is a contingency that companies from different sectors which are applying this kind of methodology in the design of their products could use the tools in a different way.
Recommendations for further research
34
References
Antony, J. (2002). Design for six sigma: a breakthrough business improvement strategy for achieving competitive advantage. Work Study, 51(1), 6–8.
https://doi.org/10.1108/00438020210415460
Antony, J., & Bañuelas Coronado, R. (2002). Design for Six Sigma. Manufacturing Engineer,
81(1), 24–26. https://doi.org/10.1049/me:20020102
Bañuelas, R. and Antony, J. (2002), “Critical success factors for the successful implementation of six sigma projects in organisations”, The TQM Magazine, Vol. 14 No. 2, pp. 92-9.
Bañuelas, R., & Antony, J. (2004). Six sigma or design for six sigma? The TQM Magazine, 16(4),
250–263. https://doi.org/10.1108/09544780410541909
Berryman, B. M. L. (2002). DFSS and Big Payoffs. Six Sigma Forum Magazine, 2(1), 25–28. Blumberg, B., Cooper, D., & Schindler, P. (2014). Business research methods (Fourth edition. ed.). London: McGraw-Hill Education.
Burgren, M., & Thorén, L. (2015). Comparing the Outcomes of Two Decision Support Models : The Analytical Hierarchy Process and Pugh Matrix Analysis.
Cervone, H.F. (2009). Applied digital library project management: Using Pugh matrix analysis in complex decision-making situations. OCLC Systems and Services, 25 (4), 228-232.
Chung, Y.-C., & Hsu, Y.-W. (2010). Research on the correlation between Design for Six Sigma implementation activity levels, new product development strategies and new product development. Total Quality Management, 21(6), 603–616.
Cooper R, Kleinschmidt E. (1986) An investigation into the new product process: Steps, deficiencies and impact. Journal of Product Innovation Management; 3:71--85.
De Feo JA. Viewpoint: Surviving the present and the future. Measuring Business Excellence 2002; 6(3):62--64.
Edgeman, R. L., & Dugan, J. P. (2008). Six Sigma from products to pollution to people. Total Quality Management & Business Excellence, 19(1/2), 1–9.
https://doi.org/10.1080/14783360701601918
Ericsson, E., Gingnell, L., & Lilliesköld, J. (2014). Implementing Design for Six Sigma in large Swedish product developing organisations – an interview study. Total Quality Management &
Business Excellence, 3363(April 2015), 37–41. https://doi.org/10.1080/14783363.2013.870782
Hagemeyer, C., Gershenson, J.K. and Johnson, D.M. (2006), “Classification and application of problem solving quality tools: a manufacturing case study”, The TQM Magazine, Vol. 18 No. 5, pp. 455-83.
Hahn, G. J., Doganaksoy, N., & Hoerl, R. (2000). THE EVOLUTION OF SIX SIGMA Six Sigma Review Basic Concept and Initial Implementation. Quality Engineering, 12(3), 17–326.
https://doi.org/10.1080/08982110008962595
Hasenkamp, T., Adler, T., Carlsson, A., & Arvidsson, M. (2007). Robust Design Methodology in a Generic Product Design Process. Total Quality Management & Business Excellence, 18(4), 351–
