Facelift: The J-Nose Job
Quality Cost Reduction and Six Sigma
Project Success Improvement
University of Groningen
Faculty of Economics and Business
MSc Technology Management
Participants
Supervisors Stork Fokker
Aerospace ing. P.J.L. Bleiji (Patrick) ing. J. van Muijen (Jaques) Lean Six Sigma Black Belt Senior Process specialist First and second
supervisors University of Groningen dr. X. Zhu (Stuart) dr. L. Zhang (Linda) Department of Operations Technology Management
Executor ing. S. Murina (Senad) Student Technology Management
Company Stork B.V.
Stork Fokker Aerospace Edisonstraat 1
7903 AN Hoogeveen The Netherlands
Tel: +31 (0)528-285000 www.fokker.com
Institute University of Groningen
Faculty of Economics en Business Technology Management Landleven 5 9747 Ad Groningen The Netherlands Tel: +31 (0)50-3634444 www.rug.nl/feb
Student Senad Murina
Violenstraat 20-29 9712 RJ Groningen The Netherlands
Tel: +31 (0)6-19001508 St. Number: s1634542 Start date July 27, 2009
Management Summary
This report is written after six month of internship at Stork Fokker BV to acquire the title Master in Science of Technology Management at the University of Groningen. The internship was held at the Aerospace division of the company. The thesis has two objectives. The first is to give advice on how to reduce the high non-quality costs –content deleted--. The second purpose is to give advice on how the company can obtain more successful Six Sigma projects.
The high non-quality costs are caused due to the production of non-conformances. Analysis revealed that the non-conformances are mainly caused by wrinkles/ buckles in the end products. Subsequently, the reasons for wrinkles/ buckles (causes of variation) are investigated. Variance is most likely to happen in the laminating process because of its dependency on human factors. Analysis of all the variables in this process led to the following results and conclusions:
The first factor of influence was –content deleted--. Engineers trained for this type of question will take over the suggestions based on this research to ultimately improve the line yield. The second factor of influence was –content deleted--. Cooperation between Stork and the
supplier to perform small tests to incrementally improve the line yield are the result.
The third factor of influence was the differences between the laminators/operators. Differences between certain operators are found and discussed among them. Additional data will be gathered and analysed to underpin the ‘best’ practice.
Finally –content deleted--.
Acknowledgement
People say “knowledge is power”. Well, at the age of 25 and a total of 21 years sitting in different school desks it is fair to say that I have gained some knowledge. This report is one of my last pieces of work that I have to accomplish to get my degree of master in science of Technology Management. (Wow, who had imagined that when I was a kid.) I strongly believe that once I have my degree, real learning will start. The same as learning to drive after obtaining your driving licence. Perhaps then I will have the great power to truly add value to this world.
First I would like to thank my family and relatives.
Most of all I would like to thank my parents for giving me the opportunity to choose and do whatever I want. You guys always let me take my own decisions and despite of not agreeing always, you supported me. Thank you for that.
Now I would like to thank my beautiful girlfriend Eeke Kuiken for supporting me during my studies and especially during this internship. I hope you will not talk the ears off my head during your internship!
Of course I would also like to thank my both sisters, my brother in law and all my friends for their support during these wonderful years.
When I began my internship at Fokker, Maarten Heere was my supervisor. I would like to thank him for giving me this great opportunity to do my thesis in a worldwide well respected company. Soon Patrick Bleiji became my supervisor as Maarten Heere was positioned on another job. Patrick, as a real people person with enormous professionalism, has taught me to look beyond scientific literature and to be aware and see the company dynamics. Thank you for supervising me during my thesis.
Then I would like to thank my Six Sigma project team colleagues at Fokker. Especially Jaques van Muijen who was one of my supervisors as well. Last but not least I would like to thank Aad van der Geest my neighbour colleague whom I had good discussion and fun.
From the university I would like to thank my first supervisor Stuart Zhu for his sharp comments and refreshing insights. Finally I would like to thank my second supervisor Linda Zhang for her good comments and time.
List of abbreviations
5S Sorting, Straighten, Sweeping, Standardizing and Sustaining
AC Autoclave
AESP Aerospace
BPR Business Process Reengineering CoC Certificate of Conformity CSF Critical Success Factor CTQ Critical to Customer
DMAIC Define, Measure, Analyse, Improve and Control DOV Diagnose, Design and Modify
--CONTENT DELETED-- --content deleted-- IC Implementation and Control IT Information technology
JIT Just In Time
MSA Measurement System Analysis
NC Non-conformance
NCR Non-conformance report NDI Non-destructive Investigation
R-Sq R-Squared
Stork Stork Fokker Aerospace TQM Total Quality Management VOC Voice of Customer
WCP@2010 World Class Performance at 2010
Table of content
1 Introduction ... 7
1.1 Stork Fokker Aerospace ... 7
1.2 Background information ... 7
1.3 Scope and deliverables ... 8
2 Research design ... 10
2.1 Conceptual Framework ... 10
2.2 Research questions ... 11
2.3 Research structure ... 11
3 Literature review ... 13
3.1 Quality improvement methods ... 13
3.2 Lean manufacturing ... 14
3.3 Six Sigma ... 16
3.3.1 Critical success factors of Six Sigma ... 16
3.4 Lean manufacturing and Six Sigma combined ... 21
3.5 Summary ... 22
4 Current state diagnosis ... 23
4.1 Non-quality costs ... 23
4.1.1 The --content deleted-- line ... 23
4.1.2 Stakeholders analysis ... 23
4.2 Projects’ success ... 24
4.2.1 Initiation and implementation ... 24
4.2.2 Interview ... 25
5 Project Facelift ... 27
5.1 Facelift yield ... 27
5.2 NCRs root cause analysis ... 28
5.3 Variation in skins ... 29
5.3.1 Hypothesis tests ... 29
5.3.2 Significant factors ... 29
5.4 Summary of results ... 36
6 Six Sigma @ Stork ... 37
6.1 CSFs to focus on ... 37
6.2 Alignment to CSFs ... 40
6.2.1 Leadership creativity and quality strategy ... 40
6.2.2 Process control planning ... 40
6.2.3 Management leadership ... 43
6.2.4 Variability reduction & effort from the entire organization ... 43
6.3 Assessing future Six Sigma projects ... 43
7 Conclusions and Recommendations ... 45
List of figures
Figure 1: The Airbus --CONTENT DELETED-- ... 7
Figure 2: Conceptual framework ... 10
Figure 3: Research structure and activities... 12
Figure 4: Lean implementation steps ... 15
Table 1: Critical success factors ... 17
Table 2: Profiles and roles in Six Sigma ... 19
Figure 5: The DMAIC process (McClusky, 2000) ... 19
Figure 6: Integration of Lean and Six Sigma ... 22
Figure 8: Project initiation process ... 24
Figure 9: Implementation process ... 25
Figure 10: Yield skins 2009 ... 27
Figure 11: Main causes of NCRs ... 28
Figure 12: A wrinkle/buckle ... 28
Figure 13: Skin shape ... 30
Figure 14: Pareto amount of NCRs at lower edge and corresponding fitted line plot ... 31
Figure 15: Pareto amount of NCRs at nose and corresponding fitted line plot. ... 32
Figure 18a & 18b: Fibre disturbance ... 33
Figure 19: Response information ... 34
Figure 20: Pareto diagram of loombatch ... 34
Figure 22: Pareto of NCRs per supplied batch ... 34
Table 3: Supplier’s material test results ... 34
Figure 21: Remarkable batches ... 34
Figure 23: Pareto of different laminators ... 35
Figure 24: Response information test between Rob and Natasja ... 35
Figure 25: Response information test between Rob and Rebecca ... 36
Figure 26: Survey results role of management leadership and the quality policy ... 37
Figure 27: Boxplot of survey results ... 38
Figure 28: Priority distinction ... 39
Figure 29: Assessing critical success factors ... 44
1 Introduction
In 2004 Stork Fokker BV (from now on Stork) introduced the use of thermoplastic materials for the production of certain airplane components. It is believed that the use of this new technology will reinforce their market position and competitive advantage through knowledge and production excellence. However, the use of thermoplastic materials lead to high non-quality costs in the production processes. A conducted Six Sigma project proved that most non-quality costs are caused by non-conformance (NC)1
While the technical issues in a Six Sigma project’s at Stork can be challenging, embedding the implementation/change and control features in the organization are of major concern. According to Hayes et al. (2005), only about one-third of the western companies meet the expectations of a quality program like Six Sigma and the successful initiatives are due to good completion of the ‘soft’ elements like top management leadership and employee involvement.
within the production of the --content deleted-- and sub-spars of the Airbus --CONTENT DELETED--. The so called --content deleted-- production process. Stork wants to reduce these costs in a Six Sigma project called Facelift, in order to sustain their competitive advantage.
This thesis focuses on the reduction of the non-quality costs made in the production process of the --content deleted-- of the --CONTENT DELETED-- and on the success of Six Sigma project initiatives at Stork.
1.1 Stork Fokker Aerospace
Stork Aerospace designs, develops and produces advances structures and electrical systems for the aerospace and defence industry and supplies maintenance services and products to aircraft owners and operators. Customers are companies as Airbus, Boeing, Dassault, Gulfstream and Lockhead Martin. The mission of the Aerospace group is to be partner of choice for the mentioned activities, at a competitive price through advanced technologies and process excellence. The Aerospace group carries out these activities with approximately 3,700 employees. The activities and services are carried out for Fokker aircrafts as well as other types of aircrafts.
The centre of excellence and technology for composites and machining is located in Hoogeveen. Here, by using composite techniques high-strength, low-weight items are produced and machined for the aerospace industry. The main focus of the composites group is on aerospace applications like control surfaces, tails, doors and fairings for several aircraft manufacturers. An example of a successful produced structure is the Airbus --CONTENT DELETED--/--CONTENT DELETED-- fixed wing leading edge, made of fibre reinforced thermoplastic.
For the reduction of the non-quality costs the focus will be on the production of --CONTENT DELETED-- DELETED--DELETED--content deletedDELETED--DELETED--. Figure 1 below shows the Airbus DELETED--DELETED--CONTENT DELETEDDELETED--DELETED-- and the wing skin.
--content deleted--.
Figure 1: The Airbus --CONTENT DELETED--
1.2 Background information
In order to sustain their competitive advantage by focusing on reliability, product quality and costs, Stork initiated in 2006 an ambitious program called World Class Performance (WCP@2010). The aim of the program was to be a world class performing company by the year 2010. According to Hayes et al. (2005), the introduction of a WCP strategy can be considered as an attempt to impose a radically new approach by learning of others practices. A WCP initiative is characterized by learning before doing with the expectation of breakthrough improvement of performance. Based on the reliability and
quality strategy, Stork chose to introduce Lean Six Sigma to the company as the way of attaining structural improvements. The focus of lean manufacturing is per definition on reliability, speed and flexibility (Hayes et al., 2005) and Six Sigma’s focus is primarily on the quality (Anthony et al., 2003). Currently different initiatives like a Kanban - production authorization system by cards – and Six Sigma quality improvement projects can be found in the company.
As a result of the WCP@2010 program, Stork initiated projects to reduce the main cost places. One project exposed that tremendous savings could be realized by reducing the non-quality costs caused by non-conformances in different production lines. And so project --content deleted-- (--CONTENT DELETED--) was born to reduce these non-conformances and with that the non-quality costs by increasing the first time yield of production processes. Project --CONTENT DELETED-- resulted in savings of approximately 25% of all non-quality costs in the Aerospace plants of Stork. Management and the project team where satisfied with the results except for the composites part of the plant in Hoogeveen, and in particular with the Airbus --CONTENT DELETED-- & --CONTENT DELETED-- production line. The costs in these lines were still too high. In 2008 a total of 866 --content deleted-- and --content deleted-- were made. For 48.5% (420) of these products a non-conformance report (NCR) was written. The NCR were written due to different reasons like wrinkles, buckles, deviations in thickness and other visual deviations. The costs of the NCRs plus the cost of scrap were about 310K Euro in 2008. Besides these made costs, not eliminating the non-conformances in production might result in an increase of throughput times and threat the planned delivery of 30 ship sets per year. Consequently, this will result in non-delivery fines from the customer and endanger future businesses. Especially if demand for thermoplastic products keeps rising. Stork decided to bring down the non-quality costs in the --content deleted-- production line by introducing another Six Sigma project, called Facelift. A team consisting of ten persons with different specialties and backgrounds was assembled for this task. The first focus of this paper contributes to the reduction of the cost made in the production of the --content deleted--.
As mentioned previously only about one-third of the western companies meet the expectations of a quality program like Six Sigma. Six Sigma is an extension of TQM and the adoption of TQM asks for effort from the entire organization and devotion to the soft elements. Moreover, the focus of the quality program should be on the long term. Changing attitudes and practices takes a long time, as does integrating all the programs that attack different sources of quality problems (including work practices, equipment, product designs and suppliers) into a cohesive whole. The quality program will be most successful if the entire organization supports the initiative by adopting the soft elements. The second focus of this thesis will be on the alignment to these soft elements.
1.3 Scope and deliverables
This project is exclusively executed for Stork Fokker B.V. and in particular for the Fokker EASP division in Hoogeveen. The solutions of how to reduce the non-quality costs in the --content deleted-- line and how to improve the Six Sigma projects’ success will focus on the situation ad hoc. All statements in this report are based on available literature and within the boundaries of the available data gathered at the company. Literature concerning the relevant topics will be obtained by using the databases: EBSCOhost, Elsevier Sciencedirect, SpringerLink, Emerald, Google Scholar and literature used during the education program.
There are several databases in the company from which data is gathered. These are BaaN and KIS2
The deliverables of this thesis are:
. Moreover, data from observations, interviews and from the supplier of the materials are used for analysis. Finally, all statistical analysis are performed with Minitab 15 or Microsoft Excel, and tested with a significant relevance of 95% as discussed and agreed with the managers in the company.
- An analysis of the variables causing variance in the --content deleted-- wings production process;
- An analysis of the main causes for non-conformance reports and the low yield of the --content deleted-- process;
- Advice about how to reduce the non-quality costs made in the --content deleted-- production process;
- Suggestions for further investigation concerning the reduction of the non-quality costs; - An analysis of literature about all critical success factors for Six Sigma implementation; - Analysis of the critical success factors presented and controlled at Stork;
- Advice about how to align to the critical success factors for more Six Sigma success;
2 Research design
The objective of this research is to give advice about how to increase the production line yield of the --content deleted-- by eliminating sources of variation from the production and to give advice about the alignment of the company to the soft elements of Six Sigma. Eliminating the sources of variation from the production process and increasing the line yield should result in less non-conformance cost. Aligning the company to the soft elements, also called the critical success factors, will result in an overall improvement of Six Sigma projects’ success.
2.1 Conceptual Framework
To achieve the objective a conceptual framework of the problem area is developed. The conceptual framework aims to show the connections between the points of interest that will be focussed on in this thesis. The conceptual framework describes the global view which is fundamental in this research. The connections between the variables are based on the diagnosis and assumed to exist and to be causal. See figure 2 below.
Eliminate sources
of variation Tuning NCR assesment
Increase line yield
Reduce non-quality costs Reduce NCR's
Increase Six Sigma projects’ success Reassess and align
CSFs for Six Sigma projects
Six Sigma projects
Facelift
The CSFs (soft elements) of Six Sigma are reassessed and aligned to the company to obtain the second objective of this research. That is to give advice on which CSFs to focus for more Six Sigma projects’ success. The reassessment is based on the findings from literature review and interviews.
2.2 Research questions
The main research question in this thesis is defined as:
How can the non-quality costs made in the production process of the --content deleted-- and sub spars of the Airbus --CONTENT DELETED-- be reduced, and what critical success factors does the company have to control in order to improve the success of Six Sigma project?
The following sub questions are formulated to answer the main research question: Non-quality costs
1. How can the yield in the of the --content deleted-- production process be increased? 2. What are the main causes for the written non-conformance reports and how can these
causes be eliminated?
3. What are the main causes of variation in the --content deleted-- production process and how can these variations be eliminated?
Six Sigma success
4. Which of the CSF for Six Sigma implementation deserve more attention at Stork? 5. How can the missing CSFs be obtained by the organization?
The first three questions concern the reduction of the non quality costs. As could be seen in the conceptual model, by eliminating the sources of variation the line yield will be increased and the costs for producing non quality products will be reduced. The last two questions address the improvement of Six Sigma projects success. A better alignment of the company to the CSFs that deserve additional attention will increase the success of Six Sigma projects.
Each sub question will be addressed in a specific section in chapter three, the field research.
2.3 Research structure
The method used to structure the research activities in order to solve the problem is the DOV method of De Leeuw, (2002). DOV stands for Diagnosis (Diagnosticeren), Design (Ontwerpen) and Change (Veranderen). The DOV-method can be characterized as a functionalist system approach (De Leeuw, 2000). Figure 3 shows the content in each step of the method.
3 Literature review
Six Sigma or in case of Stork combined with Lean Manufacturing, Lean Six Sigma is one of the possible methods/ philosophies that can be used in quality management. Because Lean Six sigma is used by Stork to improve their processes it will be elaborated in detail in this chapter. Also the critical success factors for Six Sigma projects are reviewed and discussed in this chapter.
3.1 Quality improvement methods
The concept of quality improvement was developed in the manufacturing industry to provide customers with products that confirmed to pre-defined specifications. Today, many managers in service and manufacturing business recognize the profitable impact that quality improvement methods have on competitiveness and customer satisfaction. This has fuelled a rapid growing interest in the subject. The term quality (from Latin qualitas) is a term that often is interpreted differently by people. In this report quality is defined as the total conformance to requirements. These requirements can either be the total customer requirement (e.g. the product/service compared to competitors in the marketplace) or the producer’s requirements (e.g. the degree to which the product/service was produced correctly).
Quality improvement methods strive to improve operational performance in order to enhance customer satisfaction with a company’s product or service. In other words they all strive to improve the quality. Some popular other improvement methods that have gained lots of attention in literature and business as popular quality improvement methods over the last two decades are Business Process Re-engineering (BPR) and International Organization for Standardization 9000 (ISO 9000). ISO 9000 however, is somewhat discussable as a quality improvement method as it not really establishes quality improvement; it is more meant for ensuring the constant product or service quality.
Differences between the improvement methods can be found in the organization and sectors where the different methods are used. For example since TQM is about the total quality throughout the whole organization it is not specified in which sector it is used (Prajogo, 1999) while Lean Six Sigma is mostly applied in production industry and product and process development (IBIS UvA, 2009). Another difference can be found in the focus and the application guidance of the methods. For example, Lean Six Sigma focuses on a specific problem of a flow and uses the DMAIC cycle to solve the problem while BPR focuses on select but broad business processes at which the processes changes are envisioned and initiated, redesigned, reconstructed and monitored.
One of the major developments in management practice of the last two decades was the emergence of TQM. TQM was introduced in the USA in the 1980s as a response to the quality improvement of Japanese companies (Prajogo, 2006). From then TQM was recognized as a possible source for competitive advantage and widely promoted around the world (Cheng, 2009). The method refers to a broad set of management and control processes designed to focus on the entire organization by providing tools to satisfy the customer. TQM focuses on continuous improvement at which the following elements stand central (Li et al., 2008): customer satisfaction, continuous improvement, leadership, change management, employee involvement and empowerment and process involvement. Six Sigma can be seen as an extension of TQM efforts by using detailed metrics at which it is not required to abandon TQM activities (Cheng, 2009). Naslund (2008) emphasises that it would be difficult to identify differences between TQM and Six Sigma if statistical process control would be included in TQM.
The main differences between Six Sigma and TQM are:
- Six Sigma requires devotion to a whole management philosophy rather than the deployments of quality management tools and techniques (Dale et al., 2000).
The main similarities are:
- According to Revere and Black (2003), the integrating of the Six Sigma metrics with TQM provides a measure of comparability that can be used in facilitating process improvements. - According to Goeke and Offendile (2005), the Six Sigma management philosophy is analogous
to TQM’s management philosophy. Both are about specifically designed processes to achieve measurable goals like efficiency and productivity increase and product and process enhancement.
Lean Six-Sigma is preferred by Stork because of its sequential and analytic approach. At Stork and in other companies, Six-Sigma is often the leading method with the Lean methods adapted (Arnheiter and Maleyeff, 2005). Due to this reason the Six-Sigma part is more extensive elaborated then the Lean part in the following paragraph.
3.2 Lean manufacturing
Since the nineties Lean manufacturing has gained lots of attention. The concept of Lean manufacturing (also called Lean thinking or simply Lean) can be traced to the Toyota production system (TPS), a manufacturing philosophy striving for continuous improvement of processes (Inman, 1999). The fast growth of Toyota from a small car manufacturer to the world’s largest car manufacturer has caused the growing interest in the ‘Lean’ philosophy (Reuters, 2008).
The lean philosophy is a way of thinking. In this way of thinking the main focus is to create value for the customer. Therefore the aim in Lean manufacturing is to eliminate every non-adding activity, or in other words to eliminate all waste. According to Taiichi Ohno, the founder of the TPS in the nineteen-fifties, there are seven types of wastes. Eliminating the wastes lead to the reduction of cycle times, a minimal stock inventory, minimization of changeover times, and a clear and convenient work environment.
Womack & Jones (1996) specify five main organizational principles for creating a Lean production system. These are:
1. Specify value by product;
2. Identify the value stream for each product; 3. Make value flow without interruptions;
4. Only produce what is pulled by the customer; and
5. Strive for perfection by continually removing successive layers of waste as they are uncovered.
Even thought the philosophy might sound simple, implementing Lean to the company can be a tough challenge as many difficulty and obstruction can be faced (Scherrer-Rathje, Boyle & Deflorin, 2009). The biggest barrier can be found in people’s attitude and commitment to the philosophy. Change is considered as one of the hardest obstacle in al companies. This, not only for the production department, but also for sales, engineering and supporting departments as Lean can also be implemented here (Arnheiter & Maleyeff, 2005). Some emerging ambiguities concerning the implementation are where to start with Lean and who to involve.
Hines and Taylor (2000) have built a simple step by step introductory guide to ‘Lean’. Figure 4 gives an overview of the implementation steps3
3 See Hines & Taylor for more information about the implementation of Lean.
1. Understanding waste
2. Setting the direction
3. Understanding the big picture
4. Detailed mapping 5. Getting suppliers & customers involved 6. Checking the plan fits the
direction & ensuring buy-in
1 2 3 4 5 6
Lean Thinking
Figure 4: Lean implementation steps The first step in implementation is to know what kinds of wastes there are and to make a distinction between the activities that are value added, non-value added and necessary but non-value added. The second step is to set the direction. An often seen difficulty in applying Lean is the lack of direction, a lack of planning and a lack of adequate project sequencing of senior management. This steps aims to set senior management forethought.
The aim of the third step is to develop an overview of the key features of the entire process, to understand the big picture. This contains, among others, visualizing the flows, seeing wastes and showing the relationships between information and physical flows. Having this done pictures what really happens in the company and not what is supposed to happen.
At the fourth step the senior management team should have a pretty good idea of the direction and possible areas that could be addressed. However, Lean change will not happen unless the wider workforce is involved, commitment and support of all employees is needed for success. Those who are actually involved in the day to day information and physical flows should be involved in this step. In this fifth step suppliers and customers are to be involved. There are two challenges in this step. The first is supply chain co-ordination, where the challenge is to eliminate inefficiencies and wastes between companies. And, supply chain development at which inefficiency inside certain companies within the supply chain are assessed. Tools for supply chain coordination are used in this step.
The sixth and final step is to go back to the original target setting and to review if the means for improvement are going to meet the set targets. In other words, to make a workable plan with supplied resources over a sensible time frame.
There are several advantages as well as disadvantages of Lean manufacturing that can be found in the literature. Some of these are described below (Arnheiter & Maleyeff, 2005) (Howell & Ballard, 1998) (Nicholas, 1998) (Womack & Jones, 1996).
Advantages: Disadvantages:
- Less work in process which increases working capital;
- Focus on total value chain rather than only production;
- Only value-added processes which allow a competitively price;
- Fool proof process increase efficiency and safety for employees;
- Increased deliver reliability and
competitiveness due to lead time reduction; - Encouragement of employees to think about
process quality;
- Enhancement of safety, hygienic and ergonomics.
- Resistance of management to decentralize decision making process;
- Resistance to change on shop floor level; - Misconception of Lean. People think about
the lay-off op employees instead of non-value lay-off;
3.3 Six Sigma
Six sigma is introduced in 1986 by Motorola as a quality performance measurement. Today the Six Sigma methodology is used in organization as a project-driven management approach to improve the organization’s products, services and processes (Kwak & Anbari, 2006). Companies use Six Sigma to increase their profits by making data driven decision and continually reducing defects in the organization.
Six Sigma can be perceived at different levels (Kwak & Anbari, 2006):
• Statistical viewpoint: Hahn et al. (1999), Hoerl & Snee (2002) and Montgomery (2001) discuss the Six Sigma method from a statistical, probabilistic and quantitative point of view. From this view Six Sigma means having less than 3.4 defects per million opportunities, or a success rate of 99.9997%. Many organizations use Six Sigma as a mind set to strive for and perform at a three sigma level which conforms a success rate of 93% (McClusky, 2000). An opportunity is every step where something may go wrong.
• Business viewpoint: From this viewpoint Six Sigma is defined as a ‘business strategy used to improve business performance profitability, to improve the effectiveness and efficiency of all operations to meet or exceed customer’s needs and expectations’ (Antony & Banuelas, 2001). As organizations realized the benefits, this approach has expanded to different functional areas as marketing, engineering, purchasing and administrative support (Kwak & Anbari, 2006).
Many organizations have announced major benefits as a result of Six Sigma implementation (Kwak & Anbari, 2008). General Electrics is one of the most successful companies of the last decade in implementing six sigma projects (Banuelas & Antony, 2002). GE 1999 annual report stated:
…the six sigma initiative is in its fifth year- its fifth trip through the operating system. From a standing start in 1996, with no financial benefit to the company, it has flourished to the point where it produced more than $2 billion in benefits in 1999, with much more to come this decade.
Also Motorola claims to have tremendous benefits. As a result of the implementation of Six Sigma Motorola saved $2.2 billion in reducing costs of poor quality such as reduced scrap, rework and warranty costs (Banuelas & Antony, 2002).
Despite of the immense popularity and the prevalent adoption of Six Sigma, there is also an increasing concern across industries regarding the failure of Six Sigma programs (Feng and Manuel, 2007; Hindo, 2007; Zimmerman & Weiss, 2005). According to Wurtzel (2008) one reason for failure of the program is the lack of an implementation model that details the sequences of Six Sigma elements/ activities. Another reason is lack to control the ‘soft’ elements like management leadership and training (Hayes et al., 2005). Knowing this one should not question whether Six Sigma programs can provide value, but why do so many Six Sigma programs fail? In other words what are the critical success factors for Six Sigma implementation and how can the implementation of these programs be successfully guided?
3.3.1 Critical success factors of Six Sigma
summarized the CSF for TQM implementation of 37 studies. In this paper a distinction is made in ‘vital few’ groups of CSFs that account for 80 percent and ‘remaining’ accounting for 20 percent.
CSF for Six Sigma implementation CSF implication
Management commitment and
involvement Six Sigma requires top management dedication and contributions to resources and effort. A good example of this is General Electric’s former CEO Jack Welch who restructured the business and changed the attitudes of employees towards Six Sigma.
Understanding of Six Sigma
methodology, tools and techniques Projects have to be carefully selected, planned and reviewed in order to maximize the benefits. Track project constraints, mainly costs, schedule and scope.
The tool and techniques of Six Sigma are addressed below.
Linking Six Sigma to business
strategy Select projects that are organizationally and financially beneficial. Linking Six Sigma to customers There have to be clear measures and metrics to incorporate
customer requirements.
Project selection reviews and
tracking Review the project periodically to evaluate the status of the project as well as the performance of the tools and techniques. Capture the lessons learned.
Organizational infrastructure This comes down to well trained individuals that are ready for action. Moreover, commitment of resources, time, money and effort from the entire organization is needed.
Cultural change Change must be understood first. This requires having a clear communication plan and channels for motivating individuals to overcome resistance and educating senior managers, employees and customers on the benefit
Project management skills Announcing the results of the project including successes, obstacles and challenges will help to avoid making similar mistakes in the future and adopt only the very best practices.
Linking six Sigma to suppliers In order to reduce variability and to maintain reduced cost few suppliers with Six Sigma performance can be pursuit. A win-win situation should be considered.
Training Helps people to better understand and apply the fundamentals, tools and techniques. Training is also a part of the communication. The insides of training will be further elaborated below.
Linking Six Sigma to human
resources This, in order to internalise truly changing behaviour. Human resources-based actions should be put into effect to promote desired behaviour and results. Link rewards to business strategy.
Creativity and quality Creativity should be used to improve the quality and enhance the creation of new products, services, idea, procedure, or process by persons working together in a complex social system.
Table 1: Critical success factors
Management leadership
Van Solingen (2007) defined leadership as:
management leadership for more Six Sigma success, this comes down to the combination of both. So managers should not solely focus on the planning and controlling of problems, but also on motivating and inspiring their employees with the mission and vision in mind.
According to Den Hartog et al. (1997), there are two kinds of approaches in leadership:
transformational and charismatic. Both have leadership approaches have different important elements. In the transformational approach the leader is a person that is charismatic, inspiring, stimulates intellectuality and gives attention to every employee. Charisma is linked to making the employees feel pride about their job and giving them trust and respect. Inspiring is done by communicating the vision of the company and being the role model as a leader. Stimulating intellectuality of the employees is done by stimulating employees to think critically about themselves and the processes. And finally, individual attention is given by coaching, guiding and support.
In the charismatic approach the leader is an extraordinary and skilled person, seen as exceptional in the organization. The leader is often presented to resolve a crisis situation or social disorder by appealing his mission and vision and setting followers. The ‘quality’ of the leader and his policy is proved by repeated successes.
Stork is currently aware of the fact that more (management) leadership is desired and select new managers among others on their leadership skills in job interviews and acquisitions.
Creativity and quality
The quality strategy can be defined as: The overall intentions and direction of an organization as regards quality as formally expressed by top management4. The policy is the translation of the
objective of the strategy. Leadership creativity, however, is somewhat harder to define as it can be interpreted differently. Leadership creativity will be defined in this report as: the motivation to generate new ideas (Im and Workman Jr., 2004). The two definitions have a lot in common as they both suggest intention and direction (motivation) of the entire organization or the leader. So one can conclude that organizational creativity can (and should) be used to improve the quality and enhance the creation of valuable new products, service, idea, procedure, or process by persons working together in a complex social system5
The difference between a creative and a noncreative idea can be found in divergent and convergent thinking (Amabile, 2000). Divergent thinking is the intellectual ability to think of many original, diverse, and elaborate ideas. While convergent thinking is the intellectual ability to logically evaluate, critique and choose the best idea from a selection of ideas. (Sometimes combination of several ideas into one main idea.) According to Creative Expedition (2003) the root cause for not being creative is fear of change, fear of more work, fear of failure, fear of criticism, fear of rejection and fear of losing control. Moreover, engineers tend to be left-brain thinkers relying, among others, on: control; reason; collectivity; conformity; structure; incrementalism and business as usual. While right-brain thinkers, also called imaginers, can be associated with creativity; madness; individuality; deviance chaos; radical break-point and new business. Because Stork is a company with eminently engineers (especially in Six Sigma project teams), this might be a reason for lack in leadership creativity.
.
Profiles and roles
As discussed, training is one of the crucial factors in the successful implementation. The Six Sigma method uses a belt system that must be applied through the company starting with top management (i.e. the champions) and should be cascaded down through the organizational hierarchy. Banuelas & Antony (2002); Harry et al. (2000), describe the following roles and profiles in the belt system, see table 2.
Profile Role
Executive leadership CEO or member of top management.
Responsible for setting up a vision for Six Sigma implementation.
Empowerment of other role holders with the freedom and resources to explore new ideas and breakthrough improvements.
Champion Member of senior management respected leader and mentor of business issues.
Responsible for providing resources and strong leadership for projects and inspiring a shared vision. Converts gains into €
Master black belt Identified by champion. Acts as in-house coach on Six Sigma.
Responsible for assisting champions and guiding black belts and green belts. Responsible for consistent application of Six Sigma across various functions and departments.
Black belt Technical degree, master of basic and advanced tools and respected by peers and management.
Responsible Six Sigma project execution. Also considered as change agent. Moreover, teaches and mentors cross-functional team members.
Green belt Proficiency in basic and advanced tools and responsible for Six Sigma implementations.
Member of the process improvement team and responsible for the assistants of the black belts.
Table 2: Profiles and roles in Six Sigma
Six Sigma’s tools, techniques and principles
The fundamental principle of Six Sigma is to ‘take an organization to an improved level of sigma capability through the rigorous application of statistical tools and techniques’ (Antony et al., 2003). Six Sigma is a systematic, data-driven approach using the define, measure, analyze, improve and control (DMAIC) process. Figure 5 defines the five key steps Six Sigma DMAIC process.
Define
The objects in this phase are the completion of a process map, a team charter and the identification of the aspects that are critical to the customer’s (CTQ). In order to identify the CTQ’s, the customer’s opinions, also called the voice of the customer (VOC), is collected by interviews and analyzed. Then, the CTQ’s are translated into customers’ needs.
The aim of the team charter is to clarify what is expected from the team, to keep focus and ensure the alignment with organizational priorities. Finally the process map is made which contains an identification of key inputs and requirements and a time path for the project.
Measure
In this phase the measurable CTQ’s that will be improved are selected, the performance standards are determined and the measurement system to measure the output signal (Y) is verified.
Identifying the CTQ characteristics can be difficult as the relationship between what the customer desires and the way to achieve it can be complex. Once the measurable CTQ’s are identified, the performance standard can be set. This gives established and conformed specification limits for the output signal.
In order to verify the measurements, tools as a Measurement System Analysis and Gage R&R are used.
Analysis
The variation sources influencing the process are to be revealed and the current process baseline and project goals are to be set in this phase. In order to determine the baseline process data is gathered and statistically analyzed. This can lead to a thorough understanding of the current process capability. The phase should be completed with a list of statistically significant factors influencing the process based on historical data. This list gives the input for to determine which factors will be further investigated in the next phase of the DMAIC process.
A common complication of this phase is that there is no historical data available. Improve
In this phase the factors influencing the process are narrowed further down to determine the critical factors that can cause changes in the output. Once these relationships are determined the tolerances for the process can be specified in order to optimize the process.
The objective in this phase is to develop a solution and to plan and execute the full-scale implementation. The critical factors and their relationship to the process output can be identified by using statistical tools.
Control
The control phase is perhaps the most important phase in the Six Sigma project. The objective in this phase is to make sure that the changes will last even after the project is closed. The questions that raise here are how can be made sure that the process will be performed according the specified criteria until these are changed.
Moreover, any deviations of the process are to be detected as soon as possible. To be able to detect these deviations a Measurement System Analysis (MSA) can be performed on the control factors. The MSA allows checking whether the new and improved process goals are met.
Advantages and disadvantages
Exceptional focus on the customers; The stages allow to evaluate the gains of the project and to abort the project; The outcomes are required to be in financial terms; Decisions making is based on facts. The most important disadvantages are: Successful implementation can be difficult or even not succeed; A successful implementing of Six Sigma asks a significant amount of resources (time and money) that may be discouraging; Employees, management and shop floor level, can be resistant to change. (Van Daalen, 2009; Antony, 2004; Pyzdek, 2003; Goh, 2002).
3.4 Lean manufacturing and Six Sigma combined
Womack and Jones (1994) have defined Lean as: “the systematic removal of waste by all members of the organization from all areas of the value stream”. The purpose of Six Sigma is to reduce costs by reducing the variability in the processes which leads to decreased defects. Now, while lean focuses to increase the production flow speed by eliminating non-value added processes, Six Sigma focuses on product quality by identifying and dealing with variances in production processes.
In Lean Six Sigma the customer takes a central point. Companies should know their customers’ demands and keep them satisfied with speed and quality. This can be achieved by process improvement. In Lean Six Sigma customer and supplier should work together to achieve maximal results. Lean Six sigma is about looking beyond your own processes.
Six Sigma projects are mainly executed in operating fields as production industry, product and process development and increasingly more in healthcare, financial services and Business services. In the Six Sigma projects the DMAIC process is the leading used tool. The Lean tools like 5S, value stream mapping and pull production should be implemented throughout the whole organization to encourage ‘Lean thinking’.
According to Arnheiter & Maleyeff (2005), some companies might find that they eventually reach a point of diminishing returns. Having solved al major problems and key inefficiencies no further improvements are easy generated. An organization should capitalize on the strengths of both Lean management and Six Sigma to obtain more customer value and less production costs. Or, in other words, to increase overall company performance. The following three primary tenets of Lean manufacturing can be recognized in every Lean Six Sigma organization (Arnheiter & Maleyeff, 2005):
- The organization will incorporate an overriding philosophy that seeks to maximize the value-added content of all operations;
- The organization would constantly evaluate all incentive systems in place to ensure that they result in global optimization instead of local optimization;
- The organization would incorporate a management decision-making process that bases every decision on its relative impact on the customer.
And the following Six Sigma tenets can be found:
- The organization would stress data-driven methodologies in all decision making, so that changes are based on scientific rather than ad hoc studies.
- The organization would promote methodologies that strive to minimize variation of quality characteristics.
- The organization would design and implement a company-wide and highly structured education and training regime.
Figure 6: Integration of Lean and Six Sigma
3.5 Summary
4 Current state diagnosis
The focus in this chapter is on the characteristics of Stork concerning the non-quality costs and Six Sigma projects. To have a better understanding of what the exact problem areas are, first the --content deleted-- line will be addressed to depict where the variation in the line can arise. Each step of the line is elaborated in detail. Also the results of a stakeholder’s analysis are described. The stakeholders’ analysis aims to diagnose the problem area of the non-quality costs from different viewpoints to have a better understanding of the ambiguity of the problem and to incorporate the voice of each ‘customer’ when designing a solution.
To increase Six Sigma projects’ success, the projects initiation and implementation process are described. This description allows having better understanding about when and how projects are initiated and how the outcomes are implemented. Also the results of interviews with employees of different functions will be given. The aim of these interviews is to underpin why the success of projects at the company are marginal.
4.1 Non-quality costs
The initiator and supervisor of project Facelift is Patrick Bleiji. Mr. Bleiji is Lean Six-Sigma Master Black belt and employee at the change office department. Roel Hessen and Renze Kuiken are the problem owners. They are directly responsible for the results of the operations department in Hoogeveen and the Airbus program respectively. A description of the --content deleted-- line is given first, followed by a stakeholder’s analysis.
4.1.1 The --content deleted-- line
--Content deleted--.
4.1.2 Stakeholders analysis
Diagnosing the organization or the relevant issues from different viewpoints allows understanding the ambiguous reality, which is necessary for having good insights on the problem area. Interviews with different stakeholders of the --content deleted-- line are held to understand and incorporate their point of view about the situation. The interviews are elaborated in a stakeholder’s analysis which can be found in appendix I.
The results of the stakeholders’ analysis are summarized into instrumental complaints concerning the production process. The instrumental problems have a causal relation to the functional problem that has to be solved (De Leeuw, 2002). The instrumental complains are:
--content deleted--.
According to the instrumental problems of the stakeholders, the focus in the production process should be on the stability, the production outlines, the ambiguity, training/ instructions of operators, the (first time) yield and on standardization. This leads to the following statement of the functional problem:
There is no predictable and reliable --content deleted-- production process, which is executed within simple and clear production outlines. Moreover, there are no clear instructions for writing NCRs.
4.2 Projects’ success
The initiation and implementation process of projects is discussed first, followed by a summary of the interviews. Finally the outcomes will be summarized into instrumental and functional complaints.
4.2.1 Initiation and implementation
An improvement project is practically always initiated by a production manager of a certain department who sees that a process needs and can be improved. The production manager proposes to start the improvement to the management team that judges the priority and the costs and gains of the project. If the project has no priority and the potential gains are uncertain, no resources will be provided to initiate an improvement. Then again the production chef can decide whether he wants to initiate the improvement with his own budget.
If the problem is found to be important and it is relatively easy to solve because it only concern the department of the production chef, resources will be made free and the production chef can count on support of the management. The production chef is responsible for communicating the results to the management team.
If however the complexity of the problem is high and the problem is important for the business the management team can decide to assemble a project team concerning people from different disciplines to tackle the problem. In a Six Sigma project the project team will be asked to hold review presentation at every tollgate. Figure 8 illustrated the project initiation process.
Figure 8: Implementation process The project initiation process showed that projects are primarily initiated by the production chef and the management team decides whether resources will be made available or not. For ‘easy’ improvements this might work, but according to the CSFs for Six Sigma project not. For successful Six Sigma projects the whole organization should be involved and the team should be encouraged. The team is currently too much independent. Also, employees at shop floor level are expected to carry out the instruction they receive and are involved as one of the last in the improvement. Perhaps more success can be gained when this group is involved earlier as they are the ones performing the ‘job’.
4.2.2 Interview
The question why is it complicated to implement and control projects outcomes here at Stork Hoogeveen stood central during this analysis. Different employees, from floor operators to process managers, are interviewed to gather information. The interviews were structured according to predefined questions that can be found in appendix II. The outcomes are summarized below.
- Six Sigma project outcomes are quite new to the organization. Since the introduction of Six Sigma in 2006 only one major project outcome is implemented. However, some principles of Lean manufacturing like 5S and a Kanban system are implemented. The same applies for other (small) project outcomes.
- Implementation of technological changes can be challenging but are generally good manageable. However, controlling the made changes are of major concern and a bigger challenge. People tend to fall in old habits and continue the same way as they were doing. People resist the change modification and this is strange since all interviewees agree that they are necessary.
- People seem not to take responsibility for the desired change, only if it is in their benefit. If a change jeopardizes someone’s position are can mean more work nobody will be willing to carry out the change.
- The implementation and control of technological change is a problem to the organization, at which control more than implementation.
- The effectiveness of a change is based on the quality of the changes times the acceptance. E= QxA. At Stork the quality is not the problem but the acceptance. Stork recently started a program named PITCH that among other goals aims to increase the acceptance and commitment of employees during changes. In this program employees directly involved in the processes will have the possibility to make choices that they think are best. In short time since the project initiation an increase in productivity could be noticed.
- People from different disciplines claim to be willing to change, but that it is not possible because they are dependant of other departments.
- The question how important is it according to you to have technological changes now and then, has a loaded character. All interviewees agree that having changes can be desired. But, when probing this question the interviewees blame other colleagues only to change if it is in their favour. If a change jeopardizes someone’s position in the organization nobody will be willing to carry out the change according the interviewees.
- All interviewees recognize that the implementation and control of technological changes is a problem to the organization. Control is more a problem than implementation but perhaps a consequence of the implementation. The impacts of the changes in the organization are judged as being poor.
The introduction of the Lean philosophy in 2003/ 2004 was the result of the increased competition at the time. Stork was not able to charge the desired profit margin on their production price and was forced to cut costs in order to retain the desired profit margin. The introduction of Lean manufacturing was found a good way to reduce production costs and improvement the process at the same time. From the interviews can be derived that most people in the company are not ‘ready’ for Lean manufacturing and Six Sigma, the new production philosophy. Most people do not exactly know what the possibilities Lean can provide and what the effect it will have on their work. Moreover, approximately 65%6 of the employees work more than 10 years for the company and still preserve
the ‘old organizational culture’. The old culture can be found back in the ways employees assess things, in the way employees preserve how to perform their job. Although there are several note able Lean (and Six Sigma) initiatives introduced in the company, most of the interviewees think it will take time (5 – 10 years) before the manufacturing philosophy will pay off. One should question whether this is not too late.
Functional vs. instrumental complaints
The instrumental complaints that play a role according the interviewees in the IC process are: • There is less experience in implementing and controlling Lean Six Sigma initiatives
• Employees, in all layers, tend to fall back to old habits and continue their (old) familiar way of working.
• People resist change.
• Employees choices are rather individualistic than collective. • Employees preserve the old organizational culture.
• People are dependent on actions of others for change. • Change initiatives are not enough accepted.
The functional problem that rose from the interviews conforms that aligning the CSFs to the company is necessary:
What should be done to make the Lean Six Sigma initiatives at Stork a success?
5 Project Facelift
The aim of this chapter is to answer the first three sub questions that reciprocally will lead to the reduction of the non-quality costs in the --content deleted-- production line. The three sub questions that will be discussed in this chapter are respectively: How can the yield in the of the --content deleted-- production process be increased? What are the main causes for the written non-conformance reports and how can these causes be eliminated? And what are the main causes of variation in the --content deleted-- production process and how can these variations be eliminated?
5.1 Facelift yield
The yield is defined as the amount of produced products that satisfy the predefined product specifications at once. Whether an NCR has to be written or not is decided after the autoclave (AC) process because products can visually be judged from this moment. Therefore the yield calculation is based on the products that pass this step at once. The formula below will be used to calculate the amount of yield:
# Orders hit by NC
Yield = 100 – ((---) x 100) % # Orders made in AC
Products hit by an NC are sent to stress engineers who investigate and decide whether the products can be used as they are, need adjustments or have to be considered as scrap/ waste. Their judgement is based on the tensile properties of the product.
In 2008 the average skins yield of the --content deleted-- line has been far from the desired 95%. The skins yielded not more than an approximate of –content deleted--. Also in 2009 there were no improvements of the yield. In contrary, it has become even worse as the employees were aware of the problems and have started to judge the products more critically which has resulted in more NCRs. Figure 10 shows the yield of the skins until week 44 in 2009. Also the amount of skins produced is displayed. Remarkably is that the yield had made good improvements from week 21 till 31. The process operators noted that the material felt different (harder and less moldable) then previously. This suggests variation in the supplied materials which will be addressed in section 5.3.2.
Figure 9: Yield skins 2009 There are basically two ways to increase the yield of the --content deleted-- line:
1. Find the root causes for NCRs.
The aim of this manner is to reveal what exactly causes the deviations that lead to an NCR. This question will be addressed in the next section at which the second sub question will be answered.
2. Reassess when an NCR gets written.
5.2 NCRs root cause analysis
NCRs are written when products do not conform the specified requirements. Figure 11 shows the reasons of NCRs in the production of Skins. The focus in eliminating the sources of variation will be due to awareness of data and time limitations only on the wrinkles and buckles. All reasons taken together, the wrinkles and buckles account for approximately --content deleted--. By eliminating the wrinkles and buckles the mass of the variation sources are eliminated. A cost calculation proofed that every skin not hit by an NCR saves 919 Euro. This comes down to total cost reduction of € 234K that could be realized in 20097.
--content deleted--.
Figure 10: Main causes of NCRs For a better understanding, figure 12 shows what a wrinkle/ buckle is.
--content deleted--.
Figure 11: A wrinkle/buckle Now the main causes of the NCRs are known, the question raises what causes them? This will be analysed in a root cause analysis for the skins by the use of fishbone diagrams. The analysis is the third step of Six Sigma’s DMAIC process.
Fishbone diagrams are made based on the yield figures and the pareto diagrams of what caused the NCRs. Their aim is to hunt down the root causes of the low yield. In addition, the diagrams give good insight on the variability’s that possibly disturb the process. Knowing these variables will allow eliminating the sources of variation which will be discussed in section 5.3. The question what causes the wrinkles/ buckles in the skins stood central when the fishbone diagram was made. The diagram is made with employees from different disciplines that are related to the process. See appendix III for the fishbone diagram.
The fishbone diagram has resulted in a few outstanding problem areas in each discipline. There are found to be seven problem areas. The first one is ‘Human’. In this area possibly lack in experience, lack in training and a too short introduction period are expected to have significant influence on the output, the low yield rate.
The second area is ‘Communication and documentation’. Topics for further investigation here are: no problem owner, no documentation of how to execute the job and no knowledge transfer of the job. The third area of concern is ‘The moulds’. The topics that desire attention here are: preventive maintenance, vacuum transport not equally distributed on product, leakages in connection, ratchets and cover plates and multiple cover plates/ design.
The fourth area of concern is ‘Environment’. In this area the topics unattended transport in process and disturbances as draught, dust and humidity stand central.
The fifth area of concern is ‘Material’. In this area the topics of concern are pre-cut process, no guidelines for older material, workability/ drape ability of the material and unclear how material expands during Autoclave process.
The sixth area of concern is ‘Method’ .In this area the topics of concern are: the packaging method, no adjustment of specifications and no clear guidelines for among others spot-welds and laminating. The seventh and final area of concern is ‘Resources’. The topics here are: specifications do not provide clear border, old specs are not adjusted to new findings, drawing are hard to adjust and no clear documentation of what is acceptable.
The fishbone diagram revealed the possible sources that cause the variation in the --content deleted-- production line. The next section will give an analysis of collected data to expose and distinguish the real causes of variation.
5.3 Variation in skins
The data analysis relies on statistics. The aim of the data analysis is to find one or more ‘x-sis’ that cause the low yield rate. The ‘x-sis’ are believed to be the factors that cause the variation in the process. Finding these factors will allow eliminating the variation. Let us find a ‘button’ to push in order to increase the yield!
5.3.1 Hypothesis tests
From all processes in the production process, the most variability can be found in the laminating process. The laminating process consists of approximately 10 steps. Although the predefined sequence for laminating, each step can be performed differently by every operator. The process is totally dependent on human factors and due to this reason main disturbance factors are expected here. Based on this knowledge the process is stripped to its sequences and hypotheses are formulated to cover every step. The hypothesis can be found in appendix IV.
To analyze the hypothesis, data is gathered from different sources. These sources are: - NCRs of the year 2008 till now;
- Data about the ingredients and features of the supplied materials from the supplier; - Routing documents that have to be filled in by the operators; and
- Organizational data from ERP systems.
Based on the available data every hypothesis is assessed and the most proper analysis method and technique is chosen. Because most data was discrete and discontinue (as well as the ‘input’ and ‘output’ data), the technique Binary Logistic Regression suited best and is used most. All hypotheses are tested with a significant reliability of 95%. This percentage is discussed in the project team and chosen because it was used in previous projects. The results of each hypothesis test can be found in appendix V. In the next section a summary of the most important results, the ones that have a significant influence, will be given. The significant sources of variation cause the low line yield.
5.3.2 Significant factors
The hypothesis test resulted in three significant factors that have influence on the yield rate of the skins in the --content deleted-- line. These three factors are elaborated.
1) Factor of influence: The 3D shape
The first factor of influence is 2a: Shape of the mould. The tested hypothesis was: The 3D shape of the skins has no influence on the commencement of NCRs.
In figure 13 a skin is illustrated as well as the radii positions that are tested to have influence. On the entire wing four imaginary lines are drawn to decide the radii on each position (e.g. lower edge, nose TCO etc.) of each skin section. The radii are determined based on the length of the line en the largest distance of the line to the surface of the skin, which is mostly in the middle of the skin section/ line. Subsequently this is done for all eight skins that a wing consists off.
determined. The yield is determined as the amount of good skins per radius divided by the total amount of skins per radius.
Figure 12: Skin shape On the yield per radius, a regression analysis is made in a fitted line plot. This plot visually shows the relation between the radius and the yield and the p-value of the regression. P-values less than 0.05 (reliability of 95%) are considered to have significant influence.
The Minitab results of the five different radiuses lower edge, nose, susbspar, upper edge and nose track cut-out (TCO) are shown below respectively.
Note that the bigger the radius (in meters) the flatter the product is. Also note that the first two diagrams in the pareto are from the --CONTENT DELETED-- and are not further analysed in the fitted line plot. Unfortunately, it was not possible to leave the results of the --CONTENT DELETED-- out. In the pareto diagrams of figure 14 the orange bar (characterized by 0) represents a product hit by NCR and a green bar (characterized by 1) represents a product not hit by a NCR. This accounts for each proceeding diagram in the other figures.
The ‘S’ in the legend of the fitted line plot represents the standard error of the estimate. The smaller the value of ‘S’, the stronger the linear relationship. The R-Squared (R-Sq) has to be interpreted as the percentage of variability between the variables under consideration. The value represents how much of the variation in the process output is accounted for by the model. The closer to 100% the better. Because the R-Sq tends to over estimate the strength of the association, especially when more than one independent variable is included in the model, the R-sq is adjusted (R-Sq(adj)).
40 20 0 1 0 0 1 1 0 40 20 0 1 0 40 20 0 R Lower Edge (m) = 44 Count NC Co un t
R Lower Edge (m) = 136 R Lower Edge (m) = 275 R Lower Edge (m) = 381
R Lower Edge (m) = 399 R Lower Edge (m) = 698 R Lower Edge (m) = 1188 R Lower Edge (m) = 1365
R Lower Edge (m) = 1470 R Lower Edge (m) = 1893
0 1 Count NC 8 17 21 14 20 30 17 28 23 21 17 29 20 28 8 34 16 30 18 23
Pareto Chart of Count NC by R Lower Edge (m)
2000 1500 1000 500 0,85 0,80 0,75 0,70 0,65 0,60 0,55 0,50 R Lower Edge Y ie ld S 0,0974382 R-Sq 9,0% R-Sq(adj) 0,0%
Fitted Line Plot All NC - R Lower Edge
Yield = 0,5718 + 0,000047 R Lower Edge
40 20 0 1 0 0 1 1 0 40 20 0 1 0 40 20 0 R Nose (m) = 58 Count NC Co un t
R Nose (m) = 164 R Nose (m) = 1577 R Nose (m) = 4386 R Nose (m) = 4844 R Nose (m) = 6181 R Nose (m) = 12650 R Nose (m) = 18093 R Nose (m) = 22444 R Nose (m) = 23909 0 1 Count NC 8 17 21 14 20 30 17 28 23 21 17 29 20 28 8 34 16 30 18 23
Pareto Chart of Count NC by R Nose (m)
25000 20000 15000 10000 5000 0 0,85 0,80 0,75 0,70 0,65 0,60 0,55 0,50 R Nose Y ie ld _1 S 0,0964643 R-Sq 10,8% R-Sq(adj) 0,0%
Fitted Line Plot All NC - R Nose
Yield_1 = 0,5754 + 0,000004 R Nose
Figure 14: Pareto amount of NCRs at nose and corresponding fitted line plot. The p-value of the subspar radius regression analysis is 0.028. Because 0.028 is obviously less than 0.05 the hypothesis has to be rejected. So, the radius at the subspar has significant influence on the NCRs. In this case the R-Sq is 51,3%, which means that the results are satisfying reliable. See figure 16. 40 20 0 1 0 0 1 1 0 40 20 0 1 0 40 20 0 R Subspar (m) = 147 Count NC Co un t
R Subspar (m) = 167 R Subspar (m) = 442 R Subspar (m) = 486 R Subspar (m) = 665 R Subspar (m) = 883 R Subspar (m) = 1047 R Subspar (m) = 2081 R Subspar (m) = 2442 R Subspar (m) = 3456 0 1 Count NC 8 17 21 14 20 30 17 28 20 28 18 23 21 23 17 29 16 30 8 34
Pareto Chart of Count NC by R Subspar (m)
3500 3000 2500 2000 1500 1000 500 0 0,85 0,80 0,75 0,70 0,65 0,60 0,55 0,50 R Subspar Y ie ld _2 S 0,0659782 R-Sq 58,3% R-Sq(adj) 51,3%
Fitted Line Plot All NC - R Subspar
Yield_2 = 0,5229 + 0,000066 R Subspar
