Quality Improvement Strategy in a Dynamic Aerospace Manufacturing Environment by Orion T. English B.S. Civil Engineering, North Carolina State University, 2007 Master of Civil Engineering, North Carolina State University, 2010 Submitted to the MIT Sloan School of Management and the Mechanical Engineering Department in Partial Fulfillment of the Requirements for the Degrees of Master of Business Administration and Master of Science in Mechanical Engineering In conjunction with the Leaders for Global Operations Program at the Massachusetts Institute of Technology June 2014 C MASSACHUSETTS INSTIft E. OF TECHNOLOGY JUN 18 201 BRARIES 2014 Orion T. English. All rights reserved. The author hereby grants to MIT permission to reproduce and to distribute publicly paper and electronic copies of this thesis document in whole or in part in any medium now known or hereafter. Signature of Author Signature redacted MIT Sloan School of Management, MIT Department of Mechanical Engineering May 9, 2014 Signature redacted Certified by Senior Research Sc ist, Emerit ,n V, hniel Whitney, Thesis Supervisor neering Systems Division and Mechanical Engineering Signature redacted Certified Charles Fine, Thesis Supervisor Professor of Management and Engineering Systems, MIT S nSchool of Management Signature redacted Accepted by A Accepted by David E. Hardt, Chair Mechanical Engineering Committee on Graduate Students Signature redacted Maura Herson, Director of MIT Sloan MBA Program MIT Sloan School of Management 1 This page intentionally left blank. 2 Quality Improvement Strategy in a Dynamic Aerospace Manufacturing Environment by Orion T. English Submitted to the MIT Sloan School of Management and the MIT Department of Mechanical Engineering on May 9, 2014 in Partial Fulfillment of the Requirements for the Degrees of Master of Business Administration and Master of Science in Mechanical Engineering Abstract In the manufacturing of any complex product it is a generally accepted phenomenon that defects will occur at various stages in the process. In aircraft modification and repair facilities, the low levels of automation and high degree of manual labor results in a significant increase in the rate of errors and defects caused throughout the production cycle. This results in a significant amount of unplanned rework that is scheduled and executed along with the previously planned work. This thesis presents a project carried out during an internship at Boeing focused on developing and implementing a quality management strategy targeting improvement projects to reduce rework and the occurrence of defects. This includes both the development of analysis and communication tools for identifying the most common causes of rework and working with teams to develop improvement projects to reduce their occurrence. The modification facility where the project took place was still in its early stage of operation, having only been in operation for a short period of time prior to the start of the internship. This created a very dynamic work environment that was constantly evolving and improving at every level of the organization. Previous quality initiatives had been started in the past but a lack of support and commitment from senior management inhibited their adoption. The quality initiative is focused on several key quality metrics that have been identified by the leadership team at the company. The quality management strategy is developed through a cross-functional team effort, bringing a data driven approach and aspects of several common continuous improvement methodologies. Following the framework established during the internship, some project examples are provided along with the methodology behind the root cause and corrective action steps taken. Thesis Supervisor: Daniel Whitney Title: Senior Research Scientist, Emeritus, Engineering Systems Division and Mechanical Engineering Thesis Supervisor: Charles Fine Title: Professor of Operations Management and Engineering Systems 3 This page intentionally left blank. 4 Acknowledgments I would like to thank everyone at Boeing who assisted with this project and contributed throughout its development. These individuals were always willing to share their knowledge and opinions and to help out when needed. One member of the project team was particularly active throughout the duration of the project and was instrumental in introducing me to the right people across the organization. The facility director also played a crucial role in the execution and communication of the project and provided much needed support during our time together on site. I would also like to thank my thesis advisors Daniel Whitney and Charlie Fine. Both provided insightful comments and guidance throughout the project and in the preparation of this thesis. 5 This page intentionally left blank. 6 Table of Contents Abstract ....................................................................................................................................................... 3 Acknow ledgm ents ....................................................................................................................................... 5 List of Figures .............................................................................................................................................. 9 Introduction ...................................................................................................................................... 10 1. 1.1. Com pany Background ............................................................................................................... 10 1.2. Site Background ........................................................................................................................ 10 1.3. Project Background ................................................................................................................... 11 1.4. Hypothesis................................................................................................................................ 12 Thesis Sum m ary ........................................................................................................................ 12 Literature Review .............................................................................................................................. 13 . . 2. 3. 4. 2.1. Quality M anagem ent in the Aerospace Industry ................................................................. 13 2.2. M anaging Quality Initiatives and Their Adoption ................................................................. 15 2.3. Culture and Quality M anagem ent ......................................................................................... 18 2.4. The Cost of Poor Quality ........................................................................................................... 20 2.5. Chapter Sum mary ..................................................................................................................... 22 M odification Center Organizational Assessm ent.......................................................................... 23 3.1. Introduction .............................................................................................................................. 23 3.2. EM C Current State..................................................................................................................... 24 3.2.1. Quality Defects and Rework ........................................................................................... 24 3.2.2. Inspection .......................................................................................................................... 25 3.2.3. Training.............................................................................................................................. 27 3.2.4. Facility Layout and Production Team Structure ............................................................... 28 3.2.5. Product Grouping .............................................................................................................. 29 3.2.6. Cultural and Functional Enablers................................................................................... 30 3.2.7. Challenges for the Quality M anagem ent Initiative ...................................................... 31 3.3. Future State............................................................................................................................... 34 3.4. Chapter Sum m ary ..................................................................................................................... 36 EM C Quality Plan M ethodology ................................................................................................... 36 4.1. Benchm arking ........................................................................................................................... 37 4.2. Continuous Improvement and Quality Management Methodologies ................................ 38 4.2.1. Theory of Constraints Overview .................................................................................... 7 38 4.2.2. Lean Overview ................................................................................................................... 41 4.2.3. Six Sigm a Overview ........................................................................................................... 44 4.2.4. Total Quality Management and the EMC Combined Approach .................................. 45 4.3. Shop Floor Com m unication............................................................................................ 47 4.3.2. Vertical Com m unication ............................................................................................... 49 4.3.3. Periodic Skill Assessm ents ............................................................................................. 51 4.3.4. M eeting Cadence .............................................................................................................. 52 53 Functional Project Team s................................................................................................ 53 4.4.2. Targeted Project Team s .................................................................................................. 53 Chapter Sum m ary ..................................................................................................................... 54 Data Analysis Tools: M ethodology and Form ulation................................................................... 55 Data Collection & Sources......................................................................................................... 55 5.1.1. Existing Data Sources ........................................................................................................ 56 5.1.2. Data Limitations ................................................................................................................ 57 5.1.3. Centralized Database Solution....................................................................................... 57 5.1.4. Qualitative Data Collection ........................................................................................... 58 Data Analysis............................................................................................................................. 59 5.1. 5.2. 5.2.1. Prelim inary Data Analysis ............................................................................................. 59 5.2.2. Proactive Data Solution ................................................................................................. 63 5.3. Chapter Sum m ary ..................................................................................................................... 66 M edium Blow Test Project................................................................................................................ 67 6.1. Team Goal and Com position..................................................................................................... 67 6.2. Background................................................................................................................................ 68 6.3. Approach ................................................................................................................................... 68 6.4. Key Findings and Recom m endations..................................................................................... 70 6.4.1. Pre-inspection Process ...................................................................................................... 71 6.4.2. Poor Com m unication ........................................................................................................ 72 6.4.3. Lack of Accountability ....................................................................................................... 73 Discussion of Results................................................................................................................. 74 Recom m endations and Conclusions.............................................................................................. 75 6.5. 7. Team Form ation ........................................................................................................................ 4.4.1. 4.5. 6. 47 4.3.1. 4.4. 5. Com m unication Strategy .......................................................................................................... 8 7.1. Recommendations .................................................................................................................... 75 7.2. Conclusion and Hypothesis Assessment................................................................................. 77 W orks Cited............................................................................................................................................... 78 Appendices................................................................................................................................................ 81 Appendix A: EMC Quality Plan Document ....................................................................................... 81 List of Figures Figure 1: Undiscussible Dynamics of Poor Quality Management (Beer, 2003) ...................................... 16 Figure 2: Quality Initiative Adoption Model (Martinez-Jurado and Moyano-Fuentes, 2012).............. 17 Figure 3: Classification of COPQ (Thomasson and Wallin, 2013).............................................................. 20 Figure 4: Quality Cost Examples (Campanella, 1999)................................................................................. 21 Figure 5: The Iceberg of Visible and Invisible Costs ..................................................................................... 22 Figure 6: Report of Major Cracks By Each Inspector (Drury et al., 1997)................................................ 26 Figure 7: Challenges Identified ............................................................................................................................ 32 Figure 8: EMC Quality Strategy to Achieve Future State Goals................................................................... 35 Figure 9: TOC Five Focusing Steps (Goldratt and Cox, 2004) ..................................................................... 39 Figure 10: Thinking Process Tools and Their Roles (Rahman, 2002)....................................................... 40 Figure 11: Five Fundamental Steps of Lean (Akbulut-Bailey et al., 2012; Nave, 2002).......................... 42 Figure 12: Example of Lean Tools Available to Project Teams ................................................................... 43 Figure 13: Quality Methodology Summary .................................................................................................... 47 Figure 14: Defect Tracking Analysis Flow....................................................................................................... 49 Figure 15: Management Hierarchy and Communication Reports.............................................................. 50 Figure 16: Summary of Common Issues in Manufacturing .......................................................................... 59 Figure 17: Defects vs. Time ................................................................................................................................... 61 Figure 18: Top Five Defect Table ......................................................................................................................... 62 Figure 19: Second Issue Parts Report................................................................................................................. 66 Figure 20: DMAIC Project Outline for Medium Blow Test............................................................................... 69 9 Figure 21: Process Flow for Tw o Tests........................................................................................................... 70 Figure 22: Results from Root Cause and Corrective Action Analysis........................................................ 71 1. Introduction 1.1. Company Background The Boeing Company is the world's largest manufacturer of commercial jetliners and military aircraft combined (boeing.com). The company, founded in the early 20th century by William Boeing, has grown steadily through mergers of the Boeing Airplane Co., Douglas Aircraft Co., McDonnell Aircraft Corp., North American Aviation and Hughes Aircraft (boeing.com). The company has maintained a rich history of producing innovative products to meet the demands of both the commercial and the defense, space, and security industries. Today Boeing is the only manufacturer of large commercial aircraft in the US and competes in an aggressive duopoly market with Airbus. 1.2. Site Background The 787 Dreamliner program was launched in April of 2004 with the goal of designing the most fuel efficient and advanced commercial aircraft on the market. To meet this challenge, Boeing employed an outsourcing strategy for the majority of the design and integration of the new aircraft in an effort to reduce financial risk and expedite the design and development time (Rosenfield, 2009). However, the incorporation of several cutting edge technologies, paired with the global sourcing model, proved to be much more complicated during the initial execution. Unforeseen issues with suppliers across the globe resulted in many partially completed parts being shipped to Boeing for final assembly in their facility in Everett, WA, with a significant amount of work left to be completed (Lunsford, 2007). Despite the condition of the parts that were arriving at Boeing, the decision was 10 made to continue moving the production line to allow the company to move further down the experience curve. During the early production stages of the 787 program, several planes moved through the assembly process in the factory but left the facility in various stages of completion due to the compounding amount of extra work required of all the parts. As this work becomes much more difficult to complete outside of the facility, Boeing needed to set up a manufacturing operation aimed at completing the work on the airplanes. The Everett Modification Center (EMC) was set-up to address this growing number of aircraft that required additional work and modifications after leaving the factory line. Each aircraft that is brought to the EMC has a unique work statement in both scope and scale, which creates a very dynamic production environment. With many of these planes being far behind their original delivery schedule, the need to reduce the cycle time for manufacturing and assembly of each plane becomes even more critical to meeting deadlines. 1.3. Project Background At the Everett Modification Center (EMC) the non-standard work, dynamic build plans, and breadth of work statements add significant complexity to the facility operations and its employees. As each plane is unique in regards to the work required, an installer working in the EMC might have no previous experience with a particular part or job in their work assignment. Further, the long cycle times means that an installer may not work on the same or a similar job for several months. This added challenge results in installers not being able to develop the same level of expertise that is typical amongst employees working on a moving production line. This creates a dynamic work environment with significantly more opportunities for the generation of defects resulting in added rework time. 11 The primary objective of this project is to develop a quality management strategy focused on increasing first pass quality to reduce cycle time and rework. This project was inspired by the company's identification of defects and rework as a major factor of increasing direct costs of production. Even though the facility had made significant gains in operational efficiency and throughput since its inception, the rate of quality defects was not improving fast enough. Viewing this as a significant opportunity to reduce costs, the site leadership decided to partner with the author to develop and implement a formalized quality management strategy at the facility. The leadership team identified several key metrics for the focus of the quality management initiative. The goal is to develop a sustainable quality plan that has support from all levels of management and engages the employees on the shop floor to identify and execute quality improvement projects. By establishing a data monitoring and analysis system, a formalized quality project management flow, and a communication and accountability structure, the EMC will reduce the amount of caused defects and associated rework hours associated with them. 1.4. Hypothesis It is the author's hypothesis that a quality strategy utilizing several aspects of lean, six sigma, and theory of constraints can lead to a sustainable continuous reduction in rework and defects, resulting in improved costs and quality. 1.5. Thesis Summary This thesis details the process taken in the development of a quality management strategy for Boeing's EMC facility. It begins with a review of literature on the topics of quality management in the aerospace industry, the role of culture and its affect on initiative adoption, and the cost of quality. Both common pitfalls in management's implementation strategies as well as frameworks for successful adoption strategies are presented. 12 In Chapter 3, a current state of the operations and organization are presented and analyzed in their affect on the quality performance and management. A future state of the quality management initiative is presented that details the ultimate goals of the plan and addresses the challenges that are observed and documented in the current state analysis. In Chapter 4 several quality management methodologies are presented and their relevance and fit with the project is discussed in detail. Next, a further development of the methodology behind the quality plan document developed for the EMC and the challenges it seeks to address is presented. Chapter 5 describes the data tools that were developed to support the quality data analysis. Chapter 6 describes one of the projects that the author was involved in that followed the quality management process developed in this thesis. The project focuses on the execution process of a pressurization test performed on the aircraft and is a major milestone in the build cycle. Lastly, Chapter 7 contains a brief conclusion and recommendations for the company moving forward. 2. Literature Review This chapter begins with an overview of quality management practices in the aerospace industry, from governing boards to common improvement methodologies. Next, literature pertaining to the management of quality initiatives is presented. Research on both the common pitfalls in implementation strategies as well as successful frameworks are explored highlighting aspects that were both encountered and incorporated in the management strategy that is the focus of this thesis. The relationship between organizational culture and quality management is then explored through case studies. The end of the chapter presents literature on the costs of quality and develops the distinction between the cost of high quality and the cost of poor quality. 2.1. Quality Management in the Aerospace Industry 13 The aerospace industry has very stringent safety and reliability standards for the products delivered from manufacturers. This is because the result of an error can lead to severe damage and/or fatalities. In order to ensure that high standards are met throughout the industry, there is a focus on quality standards and systems across all aircraft manufacturers. For years, Original Equipment Manufacturers (OEMs) were very concerned about the proliferation of quality standards in the industry (Gordon, 2000). As customers would customize quality standards to their specific products, this resulted in OEMs being audited constantly to different criteria but for the same basic standards. As a focus on cost savings began to make its way through the industry, the removal of redundancy in quality systems and adherence to a single, universally accepted quality assurance system was identified as a key driver for cost savings. This led to the formation of the American Aerospace Quality Group (AAQG) in 1995 and later the International Aerospace Quality Group (IAQG) whose focus is on the development of quality standards for the aerospace industry (Gordon, 2000). Over the last several decades increasing rates of complexity, global competition, and ever changing technology has led companies to place a greater emphasis on operations management practices such as quality management. The focus on quality management systems as a source of competitive advantage to improve productivity and decrease costs has become an increasingly common theme in the aerospace manufacturing industry (Beer, 2003). The requirements for speed, flexibility, and adaptation have become critical components of quality management models in an effort to compete globally in a highly complex and technologically advanced field (Baldridge National Quality Program, 2005; EFQM, 2003). Several companies have realized significant improvements in performance resulting from the adoption of an improvement model. However it is a long-term process that requires constant attention in order to be successful (Hendricks & Singhal, 2001). 14 There is extensive research in the area of quality management for both manufacturing and service organizations (Dahlgaard et al., 2006; Martinez-Jurado and Moyano-Fuentes, 2012; Zarbo, 2012; Beer, 2003). Several improvement methodologies are commonly found in practice such as Lean, Six Sigma, and Theory of Constraints. Dahlgaard and Dahlgaard-Park suggest that these methodologies are largely the same and are all rooted in the Japanese Total Quality Management (TQM) practices. These methodologies will be reviewed in more detail in Chapter 4 of this thesis. 2.2. Managing Quality Initiatives and Their Adoption There has been extensive research in the area of quality management initiatives and the importance of managing change effectively in an organization. There is a wide spectrum of examples that detail both success stories as well as failures to implement quality improvement initiatives. Dooyoung, Kalinowski, & EI-Enien (1998) estimate that around 60-67% of all quality management program implementations fail. There are a variety of reasons that these initiatives fail which generally focus on a lack of understanding of the relationship of organizational culture and quality management practices. Beer states that it is the quality of management not the management of quality that determines whether or not quality management initiatives are adopted or fail to sustain in an organization (2003). His work identifies a gap that often exists between management's stated strategy and objectives (their rhetoric) and their actions. One common pitfall noted in the implementation of improvement initiatives is the paradox between management's lack of capacity to explore the gaps between their rhetoric and actions, which is the very process of inquiry, analysis, and action that forms the basis of any quality management program that causes them to fail (Beer, 2003). Through decades of research, Beer and Eisenstat have identified what they call the "silent killers" of strategy implementation (2000). These are depicted in Figure 1 which shows how the barriers relate to quality of direction, quality of learning, and the quality of implementation. 15 Ineffective Top Down or, 1ON Laissez-Mre TOP Team enlor Management Quality of Direction Sty* Unckoar Strotegy & Prioritws Quality of Poor Coordination Across Functions &Businesses C munCal Quy of Implementation Leaming 1nadequWt* Down the Line Leadership Skills & Development Figure1: UndiscussibleDynamics of PoorQuality Management (Beer,2003) Similar insights are stated and addressed in other works which seek to develop a general framework for managers seeking to implement a change initiative such as a quality management program (Martinez-Jurado and Moyano-Fuentes, 2012; Akbulut-Bailey et al., 2012; Zarbo, 2012). One common theme is the importance of communicating the need for and impact of adopting a quality management program in order to gain support and buy-in from all levels of the organization. Martinez-jurado and Moyano-Fuentes attempted to identify which factors explain why companies adopt a quality initiative and which prior factors are required to manage adoption successfully (2012). Their study analyzed five aerospace factories that were implementing a quality improvement change initiative for at least two years. While their study focuses on the adoption of the lean production model, the similar nature of the main improvement methodologies suggested by Dahlgaard and Dahlgaard-Park means that their model should be broadly applicable 16 for any quality management strategy. The result of the study is a model that maps the interrelationships between factors that trigger the adoption of such programs, that ensure successful adoption, and that companies have had to control during the adoption process. The model is depicted in Figure 2. TIGGER Comuitive Riva3!, FACORS within Agv: cSfttot Hargai ng Powerof External Manufacturing UESS Thn orNew Entr Factorm Plant Corporation Motivation FATOS upper104 ofaliyart management1andinstitutional spao rgani UnionisationLca Soptcis iitl OrPeples and Resistac Figure2: Quality InitiativeAdoption Model (Martinez-furado andMoyano-Fuentes,2012) Several of the factors in the model reinforce the claims in other literature. The trigger factors include things like increased customer demands for improved delivery times and reliability, market competition, and the threat of new competitors. The success factors also stress the importance of upper management and institutional support, but also point out that organizations with a deeprooted culture of quality and proper organizational support are more 17 likely to succeed in their efforts to adopt a quality management initiative. The control factors are of particular importance as many aerospace manufacturing facilities have a unionized labor force. It is important that top management achieve an agreement which includes union representatives on the adoption of the initiative before beginning the initiative (Martinez-Jurado and Moyano-Fuentes, 2012). The second factor which is addressed in several papers is the initial skepticism of employees (Beer, 2003; Naor et al., 2008). Addressing the skepticism of employees who have seen past initiativescome and go without success is criticalwhen beginning a new quality program. This can be overcome through upper management's commitment to the initiative, increased communication, transparency, and contact with the employees (Martinez-Jurado and Moyaon-Fuentes, 2012). 2.3. Culture and Quality Management While a significant amount of research explores the relationship between quality management and performance (Flynn et al., 1995; Kaynak, 2003), there is an increasing focus on the relationship between corporate culture and management strategies for creating and sustaining a quality centered organization. In a study by Naor et al (2008) an empirical relationship between organizational culture, quality infrastructure, core quality management, and manufacturing performance is explored by examining 189 manufacturing plants from 6 different countries. The basis of the study follows the assumption that culture is a measureable characteristic of organizations (Naor et al., 2008). While the word 'culture' is used to explain a vast array of phenomena, there is no universal definition (Rollinson and Broadfield, 2002). For the purpose of clarity, this thesis will use the definition as put forth by the Merriam-Webster Dictionary which defines culture as "a way of thinking, behaving, or working that exists in a place or organization" (2014). 18 The work by Naor et al. seeks to establish a relationship among four separate dimensions, quality infrastructure, core quality, culture, and manufacturing performance. Quality infrastructure practices refer to the social and behavioral characteristics of quality management such as top management support, workforce management, supplier involvement, and customer involvement (Flynn et al., 1995). Core quality is related to more of the technical aspects of quality management such as quality information, process management, and product design (Flynn et al., 1995). The study revealed that both organizational culture and infrastructure quality have a direct positive impact on manufacturing performance. Further, it was found that organizational culture has a stronger influence on infrastructure quality management practices than on core quality management practices (Naor et al., 2008). This suggests that the presence of specific cultural attributes can be linked to infrastructure quality practices. For example, a risk-taking, flexible organizational culture is significantly associated with quality improvement implementation (Shortell et al., 1995). Many authors stress the importance of assessing an organization's culture and aligning the quality management system and integration strategy with the current and desired organizational culture states. Hackman and Wageman note that organizations must delegate authority to lower-level cross-functional teams to implement the process changes identified using technical quality improvement methodologies (1995; Spector & Beer, 1994). In a strong hierarchicalculture, this shifts the balance of powerfrom the managerto his employees as he needs to function more as a facilitatorand coach ratherthan giving orders(Flynn et al., 1995). This is one example of a corporate culture that will require specialattention by management to facilitateemployee empowerment, involvement, and idea generation which are common themes in most quality improvement methodologies. 19 2.4. The Cost of Poor Quality In today's competitive market high quality is a critical component for a firm to sustain a competitive advantage. A previously accepted general belief was that high quality equals high costs. However, this paradigm shifted in the 1970's and 1980's to the belief that the cost of poor quality far outweighs to cost of high quality (Harrington, 1999). The cost of quality (COQ), according to the American Society for Quality (ASQ), is a term that is commonly used but largely misunderstood and states that "any cost that would not have been expended if quality were perfect contributes to the cost of quality." The term cost of poor quality (COPQ) is preferred over COQ as COQ can give the impression that high quality drives costs. COPQ is an effective tool to bridge the gap between upper management and the quality department as issues are much more likely to be addressed by upper management when they can see the financial impact (Gryna, 1999; Krishnan, 2006). COPQ Prevention losses ot AppraisalFiue losses Internal failure costs External faijure costs waste costs Figure 3: Classificationof COPQ (Thomasson and Wallin, 2013) Prevention Costs * New product review Appraisa( Costs 9 Quality planning * Supplier capability surveys "Process capability evaluations "Quality improvement team meetings and source inspection/test * In-process and final inspection/test e Product, process, or service audits * Calibration of measuring and test * Quality improvement projects SQuality education and training equipment a Associated supplies and materials *Incoming 20 itra a costs r - Scrap * Rework * Re-inspection e Re-testing - Material review - Downgrading tra alr Costs e Processing customer complaints * Customer returns * Warranty claims * Product recalls Figure4: Quality Cost Examples (Campanella,1999) COPQ is generally classified into three categories of costs: Prevention costs, appraisal costs, and failure costs (Harrington, 1999). There are several different classifications for COPQ that have been published throughout the last several decades (Gryna, 1999; Giakatis et al., 2001; Harrington, 1999; Thomasson and Wallin, 2013). A COPQ classification framework presented by Thomasson and Wallin is presented in Figure 3 (2013). A table is also included in Figure 4 that lists several examples of these costs as given by Campanella (1999). The list by Campanella contains several examples of quality costs which are unavoidable, upfront costs, but lead to an overall lower cost of quality. For example, the prevention and appraisal costs are necessary investments and will ultimately lead to lower instances of failure. These examples are more related to ensuring high quality and preventing an excess COPQ, which comes from the other two categories, Internal Failure and External Failure Costs. Some of the examples of failure costs have an obvious and direct link to their costs that are easier to quantify, whereas others are not so easily quantified. These costs are often referred to as visible costs and invisible costs (Gryna, 1999; Krishnan, 2006). The relationship between the visible and invisible costs is best understood using the iceberg analogy as shown in Figure 5. This represents the idea that only a small amount of the true cost of poor quality is visible to the organization, but a large amount of invisible costs are often unknown and can be significantly larger than what is easily visible. 21 Defects Rework Visible Overruns Scrap Inspection 4Product Failure in the Field Costs Rs evr Inappropriato Job T me Lost Due To Accidents Extra (Touch Equipmn Failures Operations Up, Trimming etc.) -----Specfcaos to imrove b UncsryPast procediuresRcivbe deacin Due qualityFitisftgemt a im e ni satviswee C eiseaes Severlquality,~~~ Caper doume ~~~ Sevralpaersd falimanm e Pit isti ufe mn Ph maPemn rn mmnato tratge Coset Lh ofne falure whieimany m framor throughot uchssfultyheite raite Qmln uality management Lanckmn of npItfstions man tay fcnbuseaui for 5:r Tffe Ira eer w iee esntiatin oha Visbe ased Insgiale e o impr maes Corts Cotatv s tfe edt psun gr nandfacknolgpe prtiue ogreas riary fheur forgmanufatin.gn implyeme prudli ts man amn roramsk thine sthergaa fil coman wt pruigt te nablers and thcmetewt tionos. On e mmn treden inhibitors that are present in the existing organizational culture when developing a quality management strategy. The goal of many quality management strategies is to develop a culture of quality throughout the organization, however this is a long process that requires incremental steps 22 to be taken to identify current state and long-term future state cultural changes that need to occur. In order for this to be successful, it requires continuous and unwavering support from the company leadership throughout the process. Another key insight gained from the literature review is the need to communicate the impact and importance for the organization to adopt a quality management program. This ensures that employees are not just adopting a quality management strategy because they are told to, but that they understand the benefits and importance of the successful implementation and also support the initiative themselves. One way this is easily understood and communicated at all levels is by looking at the cost of quality and, more importantly, the cost of poor quality. As much of the cost of poor quality falls into what was presented as the invisible costs, it is important for management to recognize and communicate that the impact of failures can extend far beyond what is typically quantified. 3. Modification Center Organizational Assessment 3.1. Introduction As discussed in chapter 1, the distinct operations of the modification facility require that a unique quality strategy be developed to suit the culture, operations, and team structures in the EMC. To this point, it is important to understand the unique work environment that is inherent in a modification facility for large commercial aircraft. This chapter will examine the most important and fundamental differences between the modification facility and a more traditional factory production line beginning with the defects and inspection process. Next, a brief explanation of some of the key operational differences is presented along with their observed effect on the quality performance of the facility. The key cultural and functional enablers to the adoption of the quality 23 initiative are then presented as well as the challenges to be addressed. Lastly a description of the ideal future state of the quality strategy is presented. 3.2. EMC Current State 3.2.1. Quality Defects and Rework Boeing splits their defects into two distinct categories which are based on the severity of the defect found. The first category, known as a pick up (PU), covers defects that can easily be repaired through standard procedures to bring the part or system back to within the proper specifications. These types of defects vary widely from minor surface scratches on parts, a missing nut or screw on a component assembly, or a misrouted wiring or tube assembly. While these types of defects are more common, they require less resources in their resolution. The quality department can address these defects by referencing a specification and they do not require review from the Materials Review Board (MRB), which is the governing authority for some types of defects. The second defect category is a non-conformance, which is often represented by the acronym NC. A non-conformance is a deviation from what is shown on the drawings. The NC defects are reviewed by either engineering or the M RB and are either fixed through the use of a standard process for more common defects or require unique solutions from engineering. NCs are much more expensive than PUs as they require more time and resources for their solutions. If a standard repair exists for an NC, a quality representative with MRB authority can evaluate and provide a fix for the NC, otherwise it must go to engineering for approval. For any given part, if a defect is found that falls in the NC category, the engineering or MRB department will evaluate the defect to find a solution that will result in a part of equal quality to the original and if not the part will be scrapped. In this instance, the parts fall into a category referred to as second issue parts. A third category of defect also exists which is known in the industry as escapes. Escapes are defects that are not found by 24 internal quality inspectors but rather found either during the customer inspection process before delivery or after the plane has been delivered. Prior to the start of the internship project, the leadership team at the modification facility identified both NCs and second issued parts as focus areas for significant cost savings. By reducing the occurrence of these defects, significant direct and indirect costs would be realized for the facility. While direct costs are much easier to quantify as they relate to the tangible and measurable costs, such as the cost of a part, man-hours for rework, or the man-hours used to develop an engineering solution, it is the indirect costs that often have a greater impact on the bottom line. Research on the costs of rework indicates that indirect costs can be as much as six times those of direct costs (Love, 2002). Indirect costs may include factors such as worker burnout (Owens et al, 2011) and decreased productivity (Moselhi etal, 2005). Other factors such as holdups due to long lead times for second issued parts also fall into the indirect costs of quality. 3.2.2. Inspection Throughout the construction process of complex products, like large commercial aircraft, there are tens of thousands of inspection points. There are two main types of inspection techniques used in civil aircraft inspection, Non-Destructive Inspection (NDI) and visual inspection. There are a variety of techniques for aircraft NDI such as X-ray, fluorescent particle, eddy current and ultrasound (Drury, 2001). However, over 80% of civil aircraft inspections are classified as visual inspections (Goranson and Rogers, 1987). Visual inspections involve the use of not only the eyes but shaking, listening, feeling, and even sometimes smelling the components being inspected (Drury et al., 1997). The reliance of visual inspections on more subjective measures has triggered studies on the performance and reliability of aircraft inspections. In a study by Drury et al. it was found that a group of 12 inspectors found only 68% of the opportunities when asked to inspect a 25 component that had 10 major cracks present (1997). The results of this study are illustrated in figure 6. The study by Drury et al. points out the MAJOP LW 1234 variabiliyacross the different - 2 inspectors with what they find during 4 the visual inspection process, even 5 6 . when they were able to focus their 8 9 Y Y Y Y defect type, which in this case was Y Y y Y Y Y Y Y Y 10 11 attention on one specific area and -~--~ Y --- 3 CKs 51617 Y Y Y y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y o _J__1 7 7 Y Y 7Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y _ __ w 9 5 10 8 5 9 5 4 S Y Y 6 Figure 6: Report of Major Cracks By Each Inspector (Druryet aL, 199 7) [i1 .I I ..1. .L.. ... .1.I.6.. .. . .. 1 ... .... crack detection (1997). In the EMC, the same challenge faced by the installers with low repeat, highly diverse tasks is also an issue with the inspection teams. The inspectors are assigned to specific airplanes within the EMC and support a large area of production, inspecting all phases of the build and associated tasks. Inspectors are often looking for hundreds of different types of defects that could be present in the different areas of the aircraft Because inspectors are unable to specialize in inspecting specific tasks or jobs and develop a deeper familiarity with the requirements, there is likely to be a decrease in the success rate of the inspectors with respect to the amount of defects found. A common issue identified amongst the installers in the EMC is that there is inconsistency within the inspection process and that different inspectors do not agree as to what qualifies as a discrepancy. This was especially apparent with jobs that were inspected by different inspectors that work different shifts. On many occasions, an inspector would begin to inspect tasks associated with a job during his shift If the job had tasks left to be completed but the shift was ending, the installers working in that same area of the airplane on the next shift would continue to work on the 26 / job until its completion. This meant that a new inspector would be approving the work as it is completed and submitted for inspection. It was common for tasks that were previously approved to be marked as having discrepancies by the new inspector simply because of a difference in their understanding of the requirements for the job. This same phenomenon was found in a study conducted by Drury and Sheehan when one of the subject inspectors rejected a significantly higher amount of good parts because he or she did not have a good understanding of the requirements (1968). 3.2.3. Training Each new hourly employee at Boeing undergoes an extensive, 3 month training program prior to their actual first day on the shop floor. The training covers all aspects of the job including the use of computer software, reading drawings, researching job instructions, and more hands on activities that pertain directly to the job such as drilling, wiring, and testing. The breadth and depth of information covered during the initial training period can often be overwhelming. There is also a long period of time between the classroom learning and on the job application of the skills. Many employees and managers would readily admit that they could not remember a large amount of what was covered once they arrived on the job. Typically, this is not an issue as the initial training is followed by more hands on learning in a particular job on the shop floor with a mentor. On a moving production line, a typical mechanic will repeat the same jobs many times over again, allowing both installers and inspectors to develop expertise in a relatively short period of time on the job. Conversely, at the modification center, the low amount of repeat, baseline work required each mechanic and inspector to be well versed in a wide variety of tasks that accompanied the diverse jobs they worked. Several installers commented on working on a job for several weeks and never being required to work that same job again. 27 To further develop employees new to the EMC, Boeing added a training center in the facility with full scale mock-ups of aircraft sections where installers could practice specific skills. Specialized training curriculums were also developed to address the needs of existing employees that would transfer in from other parts of the company where their work was much more standardized and repetitive. However, employees and managers were often unaware of these programs and did not utilize the training facility, sometimes leading to the creation of defects which could have been avoided with more practice. 3.2.4. Facility Layout and Production Team Structure The modification facility where this project takes place operates with what is commonly known as a fixed-position layout In order to maximize the amount of usable floor space in the facility, the planes were brought in and angled in order to fit the most aircraft in the production areas. This made it very difficult to move an airplane, and often required several other planes to move as well. A fixed-position layout is often used when a product is large and difficult to move. Because the planes remain in one position throughout the build cycle, all the resources must be brought to the aircraft. The high amount of manual labor and broad range of jobs being worked simultaneously make scheduling the production crews and sequencing jobs much more difficult than a more traditional linear factory line. To simplify the scheduling of production crews, they are separated into build teams and pervasive teams. A build team works on a single aircraft in one of the fixed-positions within the facility. They are assigned to a specific volume or space within an aircraft and perform all the scheduled work for that area until it is complete. This results in a very low amount of repeat work for the installers on the build teams. The pervasive teams are scheduled across all the airplanes in the various positions within the facility. They complete the same jobs on a much shorter cycle than the build teams, 28 operating more like a single position on a moving production line. The pervasive teams' work is more specialized than the build teams'. Tasks such as installing interiors, system tests, and work within the fuel tanks are assigned to the pervasive teams. The team structure has two major impacts on the quality performance. First, because the pervasive teams perform similar tasks across multiple airplanes, they are able to develop more expertise and saw greater learning curve benefits than the build teams. The build teams have very long cycle times between working on the same task or job. This inhibits an installer's ability to develop expertise and gain a higher level of efficiency in his or her work. This also results in a significantly higher rate of defects with the build teams as compared to the pervasive teams. Second, with the build teams all operating in different positions, knowledge sharing and communication is inhibited. Teams working in the same volumes and performing similar tasks but on a different airplane in a different position do not often interact with each other. This sometimes results in defects being generated from different teams performing the same work because the learning from one team is not passed on to the other. In one particular example, one of the teams had caused damage to a panel because a part was moved before a thorough inspection for foreign object debris (FOD) had occurred. Within a short period of time, another team repeated the exact same mistake, damaging the same panel on a different airplane. This situation could have easily been avoided through better, more frequent communication between the teams. 3.2.5. Product Grouping Every aircraft that comes into the facility has a unique statement of work and plan for completion. To address these differences, the planes scheduled to be worked at the EMC were divided into four categories based on the total anticipated flow or cycle time that they would require for completion. As each of the planes in the respective categories had similar requirements in their statements of 29 work, this plan allowed for easier scheduling of work and resource allocation in order to complete the build and deliver the planes to the customers on time. While this structure improved operational efficiency in the EMC, it created additional silos within the work teams. It was common to hear from both installers and managers that crews that were working on one category of aircraft were unaware of the difficulties that crews on another would encounter. This was due, in large part, to the unique requirement of the aircraft and their scheduled build packages. While this held true for some aspects of the build, there was a significant amount of overlap in the jobs that crews would work from one aircraft to another. This created a communication barrier as managers were less likely to seek help from a crew working on a different category of aircraft because of a perceived lack of experience and familiarity. 3.2.6. Cultural and Functional Enablers There are a number of positive cultural and functional enablers for a quality management change initiative to succeed. From a cultural standpoint, there is a strong sense of pride and passion for the products that radiated throughout the employees. Several of the employees were frustrated by inefficiencies that were present in their work, leading them to pursue improvements and remove roadblocks that caused lost time and rework. There also existed a deep-rooted company culture of continuous improvement based on the lean methodology, which was identified by Martinez-Jurado and Moyano-Fuentes as a success factor for adoption of a quality initiative. Several of the employees that work in the EMC have been a part of these efforts and bring previous knowledge into the facility as well. The company's focus on lean and waste elimination has also resulted in several functional enablers to implementing a quality program. For example, each team in the facility, regardless of function, 30 would meet regularly to discuss opportunities for improvement within their work areas and were encouraged to pursue solutions to these issues. For the production teams, these were not only focused on quality but also included productivity, safety, and compliance as well. This program encouraged idea generation and employee empowerment at the lowest levels of the organization. There was also a lean organizational structure at the EMC. The facility had a dedicated lean team focused on identifying improvement projects and performing lean assessments of the facility. In addition to the lean teams, support functions such as industrial engineering also played an active role in continuous improvement projects. 3.2.7. Challenges for the Quality Management Initiative In any organization that seeks to implement change there will be many challenges present that must be addressed moving forward and developing an implementation strategy. As the project team began to map out the current state of the quality management practices within the EMC, we decided it was equally important to identify the challenges that existed and how they would be addressed moving forward with the project. The following challenges were identified by the team throughout several of the early meetings and are summarized in figure 7 below. 31 Poor Communication -Poor knowledge share of best practices/expertise between groups working in the same volumes on the aircraft 'No discussion of defects as they occur Ornprovements only carried out within a single group, not facility wide solution Lack of Visibility 'No separation between caused defects vs inherited defects of Defects oManagers are not aware of chronic quality defects 'Employee improvement teams do not have access to quality defect data *Too much focus on schedule Too Much Firefighting -Reactive environment vs proactive environment leads to more rework and schedule issues eimprovement projects set aside to deal with more pressing issues -Quality believes manufacturing does not check their own work Department Prejudice .Manufacturing believes quality inspections and standards of quality are inconsistent *Engineering believes installers do not read instructions Movement of Moemestand 'Teams are separated and moved too often which leaves no time to develop expertise within a group Resistance to Change 'Quality program is just the "flavor of the week" Employees and Managers Change W~anagers never "own" quality defects because they inherit them from past managers Many employees have seen efforts in the past fail and believe this wilk too esome employees do not want to engage in the change effort Figure 7: ChallengesIdentified The issue of communication in the EMC was one of the most important issues to be addressed moving forward with the quality management initiative. As previously mentioned in this chapter, the facility layout was one cause for existing communication barriers amongst the manufacturing managers on the shop floor. Meetings often did not include managers from all positions within the facility. If a manager wanted to seek expertise or discuss a challenge they faced in their work package they needed to seek help from others on their own. With the pressure of schedules always being an issue, other managers were often hesitant to let their subject matter experts leave to go help out another team as this may cause them to miss schedule deadlines. There were also several 32 improvement projects that had been developed in teams at the EMC. However, due to the lack of communication in the facility these were not adopted across the facility by other teams and the gains would stay within the team who developed the solution. Through many interviews with both shop floor employees and managers alike, it became very apparent that there was a lack of visibility of the defects that were caused at the facility. The aircraft that were being worked at the EMC had gone through iterations of construction and deconstruction. Every time something is installed or taken off of an aircraft there is a chance that the installer will damage that part or something nearby which will lead to a defect. Because of this, many of the defects on the aircraft in work at the EMC were not caused by the installers working in the facility. However, this information was not being captured, leaving many installers and managers to claim that all their defects were found and not caused by the team. Installers, as well as their managers, also did not have access to their quality performance and defect data. Most managers and their teams could not identify what their most chronic cause of rework was. Many department prejudices exist in the EMC that also serve as obstacles to the successful implementation of a quality management system. Many of these stemmed from a misunderstanding or lack of empathy, but there usually was an element of truth as well. For example, the engineers believed that the installers did not read the notes for jobs or tasks that they worked on. While on site, the author noted that there was a constant pressure to keep the installers on the aircraft working on tasks. However, the computers that housed the drawings and installation notes were located off the aircraft The installers were discouraged from spending too much time at the computers and would often move quickly through the installation notes so they could get back on the aircraft and begin working. Further, many of the jobs contained long, complicated work instructions that were difficult to understand. 33 The issue of department prejudice was even more evident between the manufacturing and quality departments. Early on in the project during the initial current state assessment phase, the quality department made the claim that there is too much "hand holding" between the quality and the manufacturing departments. The quality teams felt that manufacturing did not check over their own work before putting the jobs up for inspection or, in some cases, just didn't know their work was incomplete. However, some of this behavior was a result of the pressure of managers to meet strict schedules and therefore encouraging employees to put jobs up for inspection before the work was actually completed in order to have a higher count of jobs competed by the end of a shift. The unique nature of each of the aircraft that was worked within the EMC also drove frequent shifts in both management and teams. This could be driven by special jobs required on one aircraft vs. another or unplanned events that cause a particular aircraft to require more expertise and employees to meet delivery schedules. This created a challenge with the quality initiative as a newly appointed manager in charge of an area was less inclined to work on an improvement project for defects that were not caused during his/her supervision. The same held true for employees who were shifted on to different teams. If a team was already working on a project that they had identified as an area of inefficiency, the new team member had little buy-in and commitment as they were not a part of the idea generation process. The sense of pride noted in the Section 3.2.6 also played a role in this as installers typically wanted to take ownership of their own work. 3.3. Future State The team decided to create a strategy moving forward that would focus on building quality into people by strengthening their involvement and engagement and encouraging cross-functional collaboration to reduce rework and defects. A major target goal outlined by the leadership team was to get manufacturing to "own" their quality, which ideally would remove the need to have a 34 quality department. To accomplish this, the team would focus on utilizing the current functional enablers in place such as the employee involvement program, lean team, peer-to-peer training, and support teams. The team would focus on establishing roles and responsibilities at all levels of the organization, increasing communication and visibility of quality defects, empowering employees to drive improvements, and formalizing the management of improvement projects throughout the facility. This would all be supported by a data driven approach to identifying and analyzing the most impactful causes of rework in the EMC. Figure 8 depicts the strategy vision developed by the project team for the future state of the quality management initiative in the EMC. Communication & Knowledge Sharing Employee Engagement Increase Visibility Quality Strategy Figure 8: EMC Quality Strategy to Achieve FutureState Goals 35 Many of the challenges that were discussed by the project team during the early development of the quality strategy mirrored those outlined in the literature review. The project team believed that focusing on an open communication strategy that addressed issues in the both the vertical and horizontal communication structures would help to overcome many of the challenges met by previous initiatives. Specifically, the resistance and early skepticism was one factor that the team addressed early on with increased communication and commitment from all levels of management. The project team also believed that a greater focus on the analysis and communication of data would help employees understand the impact and goals of the initiative and thus raise their level of involvement and adoption. For this reason, the team decided to focus on metrics that are easy to relate to such as cost and time to make the data more personal and meaningful for all the employees. 3.4. Chapter Summary The current state analysis of the quality management and operations in the EMC revealed several challenges that the team needed to address moving forward with the quality management strategy. While the site leadership had identified several areas of focus for the team, such as reducing PUs, NCs, and second issue parts, several underlying causes of their high occurrence were also identified. This highlights the need for a strategy that not only focuses on the analysis of specific defects but also seeks to build a more formalized quality management structure. The future state of the EMC quality plan addresses many of the underlying issues and challenges that were found across the facility. Engaging the employees at all levels and increasing the communication and visibility of the defects are some of the "softer" aspects of the strategy that will enable its success. 4. EMC Quality Plan Methodology 36 The project team decided to formalize all aspects of the quality initiative into a process document to increase the communication of and familiarity with the initiative by all employees in the EMC. This document is provided in Appendix A of this thesis. The methodology for the quality strategy outlined in Chapter 3, along with several of the more common methodologies used in practice, are presented in this chapter. The first section is a summary of the benchmarking efforts performed by the project team across several facilities and programs within the company. Then a brief overview of common improvement methodologies is presented along with their applicability and inclusion in the EMC quality initiative. Finally a description of the communication strategy and team dynamics is described in detail. 4.1. Benchmarking In the early stages of the project, the author spent time benchmarking other programs and facilities within the company to gain insights into their quality management systems. Throughout the benchmarking phase, several different approaches were observed that had reached varying levels of success. All of the programs shared a data driven analysis approach rooted in the continuous improvement methodologies outlined in section 4.2. One program that was particularly successful was called the Employee Quality Empowerment (EQE) program. This was implemented in a part and small assemblies manufacturing facility that makes many different parts for the company's fleet. The EQE program is a quality program that included both a top-down and a bottom-up approach to solving quality issues and reducing deviances. This approach to quality management is also recommended by Dahlgaard et al (1998a, b). The program lays out five steps to achieve this: Inform, Set Targets, Empower, Measure, and Reward. The EQE program is similar to Plan-DoCheck-Act (PDCA) continuous improvement cycle in its use of root cause analysis and development of corrective action plans. The EQE focused on a cross-functional approach to problem solving, ensuring the shop floor employees can relate to their quality performance data, comparing 37 manufacturing cell units, and developing an incentive plan to motivate employees. Other programs focused more on the use of tools to identify systemic and chronic quality escapes and forming action plans aimed at reducing and mitigating defects. In all cases where the company was successful with implementing and sustaining quality management programs there was engagement, support, and communication at all levels of the organization. The goals were well understood at the individual and team level, and a vision was shared at the organization level. Each of the programs required a continued and sustained effort from the management team. In many cases it was noted that it took significant time and continued effort to implement and sustain the initiatives. 4.2. Continuous Improvement and Quality Management Methodologies 4.2.1. Theory of Constraints Overview Theory of Constraints (TOC) is a methodology used to identify factors that limit or constrain a process from achieving its most efficient throughput This concept was introduced by Dr. Eliyahu Goldratt in his popular 1984 novel, "The Goal." The fundamental idea is that every process has a constraint or bottleneck and by focusing on improving the constraint a company will realize the fastest and most effective path towards greater profitability (Leanproduction.com). TOC specifically focuses on process improvement and developed a systematic, 5-step procedure known as the Five Focusing Steps (Goldratt and Cox, 2004) to identify and optimize an existing bottleneck within a given process (Kasemset 2011). The Five Focusing Steps are depicted in the Figure 9 below. 38 1. Identify the system's constraint(s) 5. If a constraint has been broken, repeat steps but do not allow 2. Decide how to exploit the system's constraint(s) inertia to cause a constraint 3. Subordinate eve rything else to the constraint(s) 4. Elevate the system's constraint(s) Figure 9: TOC Five Focusing Steps (Goldrattand Cox, 2004) The TOC methodology also employs a set of basic questions organized into what is called the Thinking Process (Goldratt and Cox, 2004) along with a set of tools for each step that can be used to help identify and remove constraints and bottlenecks within a system. Following a structured thought process helps to ensure that the efforts to remove the process constraints are more focused. The questions, along with their associated tools and objectives, are shown in Figure 10. 39 Question Objective Thinking Process Tool What to change? To identify the key] problem Current Reality Tree (CRT) What to change to? To develop simple practical solutions Evaporative Cloud (EC) Future Reality Tree (FRT) How to cause the change? To implement solutions Prerequisite Tree (PRT) Transition Tree (TT) Figure 10: Thinking Process Tools and Their Roles (Rahman,2002) In the aerospace industry the increasing pressure to meet delivery time and reliability commitments to customers emphasizes the need to maximize throughput and efficiency. Missing delivery dates results in a loss of revenue service for the customer and could potentially lead to a loss of customer loyalty on the part of the manufacturer. Adopting the TOC model for process improvement is a common practice on many assembly lines (Kasemset, 2011). TOC can help to reduce the overall cycle time in complex manufacturing operations and is very effective for moving production lines like those often found in the aerospace industry. In the EMC facility, the fixed-position layout increases the difficulty of employing a TOC model because several different teams will perform the same process across the facility as opposed to the same team repeating the same process as is typical in a factory line. However, bottlenecks are still 40 present that constrain the building of a single aircraft. In order to identify a bottleneck or constraint in a process, each manager who is working on a particular section of the airplane and who performs the same processes need to agree that the constraint indeed exists. Further, the dynamic nature of the build plans for each aircraft leads to a higher degree of difficulty with scheduling and standardizing processes than a more standard production line. Because each work package is scheduled separately and independently, some constraints due to scheduling may be relieved on one plane but still exist on another. The product grouping strategy that was developed becomes a key factor in the effective use of the TOC model. For example, the aircraft that had the longest flow times required specific wiring jobs that were not required in the build packages of the planes with shorter flow times. The wiring jobs were found to be a constraint in the build process for the longer flow aircraft, which triggered the formation of several teams tasked with re-engineering the process to make it more efficient. It was only the managers who worked on the longer flow aircraft that could identify these jobs as a process bottleneck. Looking at the aircraft that were grouped into similar flow times allowed for easier identification of the constraints in the build cycle using TOC. There are also processes that are required for every aircraft before final delivery that had been identified as bottlenecks using TOC. One example of these is the Medium Blow test, which is a test of how well the fuselage maintains a constant pressure. This process is performed on every aircraft in the facility, which enabled the constraints to be identified easier as every position experienced the same problem. A process improvement project focused on the medium blow process is explained in greater detail in Chapter 6 of this thesis. 4.2.2. Lean Overview 41 The concept of lean manufacturing was developed and introduced by Toyota as a part of the Toyota Production System (TPS). Similar to TOC, lean was popularized by the best-selling book by James P. Womack in 1990 titled The Machine That Changed the World. The fundamental concept of lean is the relentless removal of waste or non-value added activities from the manufacturing process (Womack et al. 1990). Waste can be defined as anything that is not necessary to produce the product or service (Nave, 2002). There are five steps that are essential in lean that are illustrated in Figure 11. K V V Value Identification Value Stream Analysis Improve Flow Customer Pull Perfection " Lean concept that focuses on determining I which features customers perceive as value in products and services The process of identifying which activities add value to the process and eliminating those identified as non-value adding " A focus on making sure that the products or services flow uninterrupted throughout the value stream process ] The effort of producing the product or service only when the customerwants it * The continuous improvement process to constantly strive for perfection and remove nonvalue added activities Figure11: Five FundamentalSteps of Lean (Akbulut-Bailey et aL, 2012; Nave, 2002) While lean also focuses on removing waste and improving process flow, there are also a number of side benefits that result as well. Quality, for example, is often improved with the implementation of lean manufacturing methodology. As a product spends less time in process, there is less chance for damage (Nave, 2002). There will also be less variation as processes are simplified throughout the 42 production cycle. While lean is not directly focused on quality improvement, it is an essential component to any quality management program. Boeing adopted the lean methodology over a decade ago as the primary focus of their manufacturing improvement strategy. Since that time the company has made tremendous gains in production efficiency and is now an industry leader in aerospace lean manufacturing. The company utilizes a vast number of tools that have been developed in association with the lean methodology. The project team identified several tools that are especially useful in a quality management program, such as Process Flow Charts, 5-Whys analysis, Cause and Effect diagrams, Fishbone diagrams, Histograms and Pareto Analysis. These tools were organized into a project management template that could be used by teams to help guide their analysis and management of improvement projects. They also serve as valuable communication tools that are easily understood throughout the organization and management Fishbone Diagram Cause Equipment Process Effect People Problem Seconday cause cause Materials Environment Management Figure 12: Example of Lean Tools Available to Project Teams 43 4.2.3. Six Sigma Overview Six Sigma is a methodology that takes a quantitative, data-driven approach to eliminating defects and improving quality in any process (isixsigma.com). It specifically refers to the capability of a process to deliver units that are within a set criteria of quality units (Klefsjo et al., 2006). Statistical analysis is a primary component of process monitoring in the Six Sigma methodology. In order for a company or process to achieve Six Sigma, it must produce less than 3.4 defects per million opportunities within the product, process, or service. The idea is to systematically eliminate the defects that are linked to a given process to move towards perfection in a process or system. The primary methodology employed in a Six Sigma management system is best known by the acronym DMAIC. DMAIC is the process that defines, measures, analyzes, improves, and controls an existing process. The five phases of quality control are defined below (Anderson et al., 2006; www.asq.org): * Define: This phase defines the process or product that requires improvement. The project goals, improvement activity, and customer (internal or external) are also defined and a map of the process to be improved is created. * Measure: This phase involves identifying specific factors of the process or product that have the most influence and making measurements. - Analyze: In this phase, an analysis is performed to determine the root cause of the problem. * Improve: This phase includes the design and implementation of solutions that will eliminate the root causes of the problems. * Control: This phase should verify that the implementation was successful and is monitored through statistical process control methods to ensure the improvements are sustained. Aspects of the six sigma methodology are very useful for analyzing and improving some of the more chronic causes of rework in the EMC. From a project management standpoint, the six sigma DMAIC 44 structure is very easy to understand and follow making it very effective for projects specifically focused on mitigating common quality escapes. However, as many of the jobs that are completed at the facility had never been completed at the company before, finding enough data to perform a thorough statistical analysis was sometimes difficult. This highlights the need for a quality management strategy that utilizes several known improvement methodologies rather than focusing on a single methodology. 4.2.4. Total Quality Management and the EMC Combined Approach The previous three methodologies are comprised of quality management and manufacturing philosophies and concepts which all have the same roots as the Total Quality Management (TQM) methodology (Dahlgaard and Dahlgaard-Park, 2006). TQM can be defined as the continuous improvement of work processes to enhance the organization's ability to deliver high-quality products or services in a cost-effective manner (Spector and Beer, 1994). The TQM methodology involves a multiple stakeholder philosophy that emphasizes the value of teamwork and collaboration (Beer, 2003). It is clear that to implement a TQM policy would be to include several aspects of the other methodologies described earlier in this chapter. TQM's aim is to shift the company culture from a passive and defensive culture to a proactive and open culture with total employee involvement (Dahlgaard and Dahlgaard-Park, 2006). TQM has been found to have a significant impact on the performance and culture of an organization when properly implemented. The project team viewed the quality strategy as falling primarily into this category of allencompassing quality management The team sought to take a proactive approach to quality improvement by taking steps to identify issues that were commonly found on the airplane (inherited) so that production could better plan and optimize around their solutions. There is an equal focus to identifying root causes of the defects generated at the EMC and developing solutions 45 and action plans using aspects of lean, six sigma, and TOC. The team agreed that no single methodology would fit the dynamic nature of the operations at the EMC, and that our strategy was more aligned with that of a TQM model. Each of the quality management methodologies reviewed in this chapter have their strengths and weaknesses. The unique challenges and operations of the EMC lend themselves to a combined approach that does not focus on any one particular methodology but rather takes aspects from each and applies them in the most effective manner. This helps to eliminate the limitations that are inherent in the specific methodologies and results in a more encompassing and widely applicable quality management strategy. The table in Figure 13 summarizes the methodologies reviewed in this chapter along with their inherent weaknesses as standalone management practices. The table also provides specific challenges in the EMC and how the various management philosophies will be used to address them. Methodology Six Sigma Lean TOC Theory Reducing variation Removing waste Managing constraints Focus Problem focusing Flow focusing System constraints Weakness * Ignores system interaction * Processes improved independently * Identify most impactful and chronic defects e Correlate chronic defects for the same volumes in different positions for build teams e Proactively address chronic inherited defects through data Use in EMC Quality Plan Ignores statistical or system analysis Root cause analysis tools for caused defects e Process mapping to identify additional causes/area of rework - Increase visibility of defects through lean tools e Eliminate rework e 46 * Ignores data analysis * Minimal worker input * Identify process constraints across product groups * Identify common process bottlenecks for build teams * Use to guide data analysis to further evaluate impact of system level constraints tools (waste) Figure13: Quality Methodology Summary 4.3. Communication Strategy It is an essential component of any quality management strategy to ensure that open and effective communication lines exist at all levels of the organization and incorporating them into the quality management strategy. The project team addressed the communication gaps in the EMC by mapping out how communication will flow between management and employees to facilitate knowledge sharing and help generate ideas and support for improvement projects. Several levels of communication were addressed by the team as a part of the quality management strategy that is detailed in this section. 4.3.1. Shop Floor Communication The first level of the communication plan involved the communication between the installers and their managers. The goals were to decrease the time between the occurrence of a defect and the root cause analysis and to increase the communication between the shop floor managers and the installers. Previously, as defects were generated and tags were written by inspectors there would be little to no communication between the inspector and the installer. There was also no communication between the managers and the inspectors after the tag was written explaining the findings by the inspector. Sometimes defects could go weeks before they would be addressed, making it much more difficult to figure out who was working on the job when the tag was written. This also meant that tags were not reviewed for accuracy and approval by the manufacturing department 47 To bridge the gap in the communication structure, the team pushed for the adoption of a tool called the Workcell Action Tracker (WAT). The purpose of the WAT was twofold: to track whether a defect was caused in the EMC or created elsewhere in the company, and to ensure communication between the shop floor manager and the installers each time a defect was created. Tracking an additional data point on whether the defect was caused at the EMC or inherited would allow the team to focus on mitigating only the chronic causes of rework that are caused at the EMC. Further, as trends are identified with recurring defects that are inherited, the production and planning teams can address them with greater efficiency and become more proactive. The tool also included a text box that allowed the manager to fill in details on the root cause and possible corrective action after it is discussed with the installer. Addressing the defects as they occur on a daily basis will not only improve the quality of the root cause analysis but it also increases the engagement of the installers who caused the defects. One manager spoke about having installers discuss in team meetings a recurring defect that they had caused, why it happens, what they had learned, and how to prevent it in the future. This particular team immediately began seeing a decrease in the amount of defects generated. To help managers in the use of the WAT tool a flow chart was created and is displayed in Figure 14. The flow chart was used to assist managers with the basic process of tracking the root cause of defects and specific actions to take with employees while addressing defects on a daily basis. To ensure that these conversations were happening between the managers and their employees, a report was created and reviewed daily showing all of the defects that were created, which manager owned the area and which ones had completed root cause analysis and filled in the tool. 48 VLF Defect Reduction Flow Q-aly rie prior day's NCs Wasthe NC crated by Assign to manuong Y Basin .NCRe Cn Id Y macaw No No Qualty to aagntagto appoprae organization (Suppler, Partner, etc) CondctRCA CAP - Corrective Acti n Plan SEC - Skis Enhance ment Center WAT - Workcell Action Tracker E .ru.nty ENo Develop CAP based an Wasthe Root Cause people r Enleretaintodeled den aet3&~~~D RCA-ause An olysis RCA ndegsltoane owth ReviewCAP apprate to onect pioces CIE, ME, PitrewdtddtbseW rocass- processralated Entere oCAPt Bosa, Tooling, etc) People Fdlowmeprops e Yes: Mnerli sctlanpeaess perPRO-19 Noll < Occurence 9 Develop CAP based on RCA findings to include appropriate aew wide adion to mitigate potential reoccorence ueddett 1i CAP eath rewNeodtbs(I action to mitigate be kientited ye: training a ade Have a discussion vAti employee about defect and , Devetop CAP to vark vAh SEC to and apprpriate tor rewrie acio Wi defed database (al) Enrit eployee in bainingand remindcrewto utlize th Identify training needed byemployee Re Enter resut of CAP into Ente resufts o CAP Into potential reoccunence ccurrence Was sktils Reew CAP apDe ce widei Can eidploee wCAP crew Enter remits of CAP Into defe database CMa No DevelopCAP based on RCAfnidingsto include having a disaisson With P empipe to be doamented via emailFde!lt sitaewdlditibaeIAT) and or appiopriate crewviie acion to milate potential reoccunence Enlerresta ofCAP into Figure 14: Defect TrackingAnalysis Flow 4.3.2. Vertical Communication The next level of communication that the team addressed was how quality defect information would be communicated through the different layers of management Poor vertical communication will reduce an organization's quality of learning and will inhibit the efficacy and adoption of quality management program (Beer, 2003). A lack of proper communication and details at the highest levels of the organization can also lead to the creation of more defects. For example, in the high level status meetings where managers are looking at aircraft status in terms of schedule, managers are often pressured to "sell" more jobs to get the aircraft status back on schedule. This can lead managers to complete jobs out of sequence so that their status looks better which may lead to the 49 creation of more defects. One key piece of information that is often left out of these meetings is the amount of rework the teams are completing which causes schedules to get behind (rework versus baseline hours). Bringing more detail into these meetings will help the leadership team to make more informed decisions for job scheduling and sequencing. At the leadership and senior manager level, it was decided that information on the greatest causes of rework as well as what actions are in place to mitigate those should be discussed. The goal is to ensure that projects are being prioritized and any help needed is being addressed and expedited at the highest levels of the organization. Figure 15 represents the management hierarchy within the EMC and the quality information communicated to each level weekly. " Director level dashboard shows high level quality performance of facility including top 5 defects, top 5 teams, and total rework time split out by each aircraft currently in work at the facility - Positional and pervasive dashboards with performance metrics including first pass quality (FPQ) %, total defects, top 5 teams and defects, total rework hours and 2 nd issue parts by aircraft - Dashboard includes top 5 NCs and PUs, FPQ%, total defects generated, rework hours per defect, current improvement projects, project timeline with milestones, help needed * First line managers receive reports from industrial engineering for quality defects and reissued parts report based on schedule~dwork - Receive data and analysis information from industrial engineers supporting the teams Figure15: ManagementHierarchyand CommunicationReports For these meetings, the team created a management dashboard that was linked to the quality defect database that was created by the team for analysis which is discussed in Chapter 5. The dashboard is an automated, web-based tool that provides timeline performance charts and quality metrics including the top defects, top teams, and associated rework time for three different levels of management starting at the senior managers and going up to the facility director. Information is 50 presented both on a weekly basis as well as a three month rolling basis due to the long cycle times of the aircraft in the EMC. The dashboards are also linked to projects in work along with timelines that include milestones. The dashboard is used at meetings called Special Attention Meetings or SAM meetings that are specifically focused on quality issues and discussion. An example of the management dashboard is included in Appendix B. A separate report was also generated by the team for use in the daily status meetings to provide more detail on the defects generated. At the lower levels of management the industrial engineers supporting the teams and the managers take an active role in preparing and disseminating data and reports. The managers receive two types of informational reports from the industrial engineers. The first is a predictive report outlining scheduled jobs that contain parts that were reissued in the past for various reasons which will be discussed in Chapter 5. The other report was an analysis of quality performance metrics similar to what was shown on the dashboards for senior managers but with a more detailed breakdown of the defects and their descriptive texts. These are prepared by the industrial engineers and are developed collaboratively by both the industrial engineers and the manufacturing teams that they support 4.3.3. Periodic Skill Assessments The installers at the EMC require a broad skill set to complete their work. The project team found that the root cause of a large amount of the defects specifically related to drilling and wire routing were due to a lack of training and expertise on the part of both the manager and the installer. Even with the addition of the mock training center for hands on practice noted in Section 3.2.3, many defects resulted from installers not having performed a specific task for a long period of time. Some of the installers commented that they did not have the proper skills or confidence to complete certain jobs without making a mistake which could lead to the generation of a defect 51 To address this disconnect in available training and defects caused by inexperience, the project team incorporated a tracking system for skill assessments and participation in essential training programs that specifically related to some of the top quality defects identified by the team. Periodic skill assessments would ensure that even employees without significant experience possessed the required skills to complete a job and achieve first pass quality. The assessments would be required at the different phases of the build cycle when the types of tasks worked began to shift. With a high degree of manual labor, a greater emphasis was placed on more specialized training for employees in all the job functions across the EMC. Several of the projects that were in work as a part of the quality initiative when the author left the project site were focused on developing new training curriculums, assessments, and assessment schedules. At the time the author left the project site, these curriculums were still in development. 4.3.4. Meeting Cadence The project team believed it to be critical to establish a weekly cadence for meetings described in this section that included agendas and attendees. In an environment where firefighting and schedule changes were the norm, the team focused on establishing a meeting rhythm to increase participation by both management and employees. The meetings consisted of a cross-functional group of employees representing all stakeholders including quality, engineering, manufacturing, industrial engineering, and support functions. The established meeting cadence would ensure that projects were moving forward and meeting milestones and progressing as needed. To encourage participation managers were asked to discuss their area's defect data with the other meeting attendees. The teams with the highest rate of defects were asked to describe the root cause of the defects and possible corrective actions with the other groups. This was an effort to 52 solicit feedback and collaboration across the different positions in the facility where teams were working on different categories of aircraft as described in Section 3.2.5. Teams with similar issues may have expertise to lend or experience with improvement efforts that could be shared. 4.4. Team Formation 4.4.1. Functional Project Teams Project teams for the quality initiative would be formed both on a functional basis as well as a target basis. Functionally, all the installers are on a specific team with a manager in the EMC whom they work with on a daily basis. Each team currently participates in the employee involvement program and has weekly scheduled meetings to meet and discuss improvement project ideas. These teams were expected to work one project related to quality each fiscal year that would be guided by the data and management structure from quality management initiative. The industrial engineers would work with the teams to support the analysis with Pareto charts, histograms, and pivot tables to help the team identify their top defects and greatest causes of rework. This ensured that the teams would have direction for their projects and remain focused on an issue until a project plan has been carried out. 4.4.2. Targeted Project Teams Target teams would also be formed for more systemic issues that were common across the EMC. These were issues that required more expertise and had a greater impact to the overall build and process flow in the facility. This team structure was more focused on providing cross-functional expertise to address challenges that were identified through analysis of the quality data. Including all the departments in generating a solution gave it a much greater chance for site-wide adoption as each department has a stake in the project. These projects were sponsored by multiple managers 53 with more senior ranking who would ensure successful completion and implementation of the projects. 4.5. Chapter Summary This Chapter outlines the methodology behind the EMC Quality Plan document that is included in Appendix A of this document. The project team decided the best approach to developing a quality improvement strategy for the EMC was to combine several of the well-known quality management methodologies and incorporate the most applicable aspects of each. As each methodology has a different focus and approach, there was not a single one that would fully address the challenges present at the EMC. A much more effective strategy is to utilize tools from the various methodologies to address different aspects of a quality management strategy. The quality plan document was created to outline a formalized management structure and process that detailed roles and responsibilities for all parties involved in the management and execution of quality improvement projects across the EMC. This was prepared both to help communicate the goals of the quality program to all levels of management as well as serve as a reference document. The team focused on the communication aspect of the plan as several of the challenges present, and sometimes causes of rework, stemmed from the lack of communication at all levels across the facility. The communication strategy addressed both the vertical and horizontal layers of communication with a goal of removing roadblocks for the execution of improvement projects by providing more visibility at the top of the organization. The team structure for the quality strategy was also discussed to address the various levels of quality management within the facility. The functional teams, of which all employees worked under the same manager, were already in place and were expected to work on improvement projects 54 throughout the year. The project team sought to utilize this existing team structure and provide them with the tools and support necessary to make significant gains on quality improvement related projects. The higher level, more systemic quality issues required a more focused targeted team to work on solutions that would require input and support from a broader range of employees. These teams would require employees to be selected based on their expertise or other criteria to help expedite the analysis and solution implementation. 5. Data Analysis Tools: Methodology and Formulation A big component of the quality strategy is to employ a data driven approach to identifying and solving high impact defects and causes of rework. This chapter describes the team's effort of analyzing the current data sources and what is needed to support the quality initiative. This is followed by a brief description of a centralized database that was created to support the analysis of quality defect. Then a tool is presented that was created to link scheduled jobs to past discrepancies to initiate proactive investigation of commonly damaged parts and reduce the impact of order lead time on production schedules. 5.1. Data Collection & Sources A key aspect to most quality management programs is the collection and dissemination of data. Due to strict requirements and regulations set by the Federal Aviation Administration, many aerospace manufacturers collect a vast amount of data. This expanse of data can often lead to a data overload for analysts that are seeking to isolate issues and identify the more important factors relating to a problem. Often an analysis can be slowed down by too much data instead of simply possessing the right data. This section will detail the data tools that were used and developed to help support the quality initiative at the EMC and the methodology behind their formulation. 55 5.1.1. Existing Data Sources In general, the aerospace industry is very data rich. Throughout the manufacturing process, every task within a given job or work package must be signed off by the mechanic, and sometimes inspector, so that each task can be tied back to the time, place, and person involved with the completion of a job. Every quality defect found throughout the production process is documented in much the same way. Quality inspectors record information pertaining to the defect such as: - Multi-digit, multi-tiered defect code e Location on the aircraft and the aircraft number * Job or task the defect was derived from * Name of the inspector who found the defect * Date and time the defect was found * Detailed description of the defect The IT department maintains all the data across the company in a data warehouse on an internal server that can be accessed only by analysts with permissions. While this data is accessible, it requires thorough knowledge of SQL coding and database architecture in order to create useful data pulls. Also, the data stored on the main server often times has field names that are difficult to understand for anybody who is not familiar with the underlying data. The company developed an in-house data software that contained quality information as well. This program allowed the user to view the defect data for a specific area of the aircraft. However, the user has no ability to perform any analysis of the data that is presented. All of the tables and charts are coded into the program and do not allow the user to manipulate the data. If the user wanted to make comparisons with another aircraft, or across many aircraft, the user would have to pull up those pages individually. The program also does not allow the user to look at data for past aircraft, only the aircraft that are currently in the manufacturing process are available to view by the user. 56 Further, the program looks at only one week and three week time windows. As this is ideal for a continuous or pulse production line, this was not useful for the EMC. 5.1.2. Data Limitations There are several limitations that have historically made it very difficult to analyze the data that is stored on the main server. The first, and most prominent, is that a large amount of the defects that are recorded on the inspection sheets are not caused by the mechanic who is working the job. Some are created either by a supplier or by the team that prepared the aircraft for the modification work by removing parts as necessary. This makes it difficult to track performance over time with certain tasks as improvements may be masked by the defects that found on the parts being worked. Another limitation in the data is the inaccuracy of the defect codes recorded by inspectors. It has been shown that when faced with too many options many people go through what is known as "choice overload," and become demotivated when making selections. With such an exhaustive list of defect codes, inspectors are more likely to use more common or generic defect codes even when they do not fit the defect that they are recording. This is further agitated by the fact that parent child relationships can be made with the digital inspection report sheets. Instead of writing a new defect report, an inspector has the option to link the report to another report that has been written for a different defect. This leads to inadequate information on the code sheet and sometimes inaccurate defect codes as well. Because defect codes are the easiest way to group defects when analyzing a set of data, this significantly reduces the reliability of the data. Another limitation in the data is that defects may be found later in the build during system tests. This results in the defects being coded to the test rather than to the job where they were actually created. 5.1.3. Centralized Database Solution 57 To address the challenges of data collection and access, the team decided that a dynamic, centralized database with a user friendly interface was necessary to support the quality management initiative. This would allow the team to have continuous access to the data necessary to identify the most impactful and chronic defects that are created at the EMC. The team decided that Microsoft Access was powerful enough to support the needs of the EMC and also provided the most user friendly platform. This also allowed the team to pull only the data fields from the main data warehouse that would be useful in the analysis of the defect data. As a part of this, the team decided to focus on a new metric which was the amount of rework time that resulted from each of the defects that were caused. Looking at not only the occurrence, but also the rework time would ensure the team focused on the most impactful discrepancies. The database was linked directly to the main data warehouse and pulled fields from several different departments using relational queries written in SQL. The information would be updated daily and was specifically set up to pull information on the aircraft that were manufactured in the EMC. The project team was also able to incorporate the data that was collected from the Workcell Action Tracker into the database. 5.1.4. Qualitative Data Collection At the beginning of the project the team decided to collect qualitative data from both employees and managers across the various departments in the EMC. The project team believed that pairing the more quantitative defect data with observations and experiences on the shop floor would be more effective in identifying root causes for the more systemic issues that exist within the EMC. It was also important that the team use this information objectively and not bring a bias into the initial analysis of the quality data. This information could also help the team to identify some of the less obvious issues and inefficiencies. Through both employee interviews and a survey sent to 58 about 200 people the team collected information on the top issues that cause delay and rework in the EMC. A summary of the results are presented in Figure 16. Issue Leads to Explanation Improperjob The work instructions and drawings are confusing or missing information Incorrect number or missing components from the internal or external supplier Job cannot be started because it is waiting for sequencing another Inexperienced labor Non-standard complex tasks, low amount of repeat work inhibits building expertise Inadequate documentation Incorrect number of parts Defect generation, delays due to research Order new parts, schedule delays, sequencing issues, lost time Schedule delays and sequencing issues one to be completed High defect rate, learning on the job External supplier quality Damaged incoming parts/components Incorrectly constructed components received by the external supplier Removal and rework Damage and/or scratches occurring on incoming components from internal or external suppliers Removal, rework, reissued parts Lead time for parts Long lead time for component orders due to low volume/low priority Schedule delays and sequencing issues Damaged parts/components Long time to complete jobs Damage found on the airplane during inspection or Rework, removal, reissued parts manufacturing High costs perjob, multiple installers Improper estimation of time to complete working the same job tasks/jobs Incorrect part The component from the internal supplier or previously installed component is incorrect Removal, engineering approval Unnecessary meetings Meetings not generating any actions Less time for value-add activities and effective management Incorrect error reporting Incorrect reporting of deviations on the quality report from the inspector Lose visibility of performance, misleading data Problems caused by the manufacturing and the mloeestesle deme department and the employees themselves A re-design of the product is required which requires a change order. Assembly error Change orders Rework, schedule delays, lost time Lost time, reissued parts, removal, rework Unplanned disruptions Unplanned disruptions such as helping a colleague, Lost time L _sttime Unp__nneddisruptons safety issues, and unplanned operations Figure16: Summary of Common Issues in Manufacturing 5.2. Data Analysis 5.2.1. Preliminary Data Analysis The first issue the team decided to investigate was the most common type of defect for each volume of the aircraft that was supervised by an individual manager. This was a logical breakdown for the data as each section of the aircraft would experience different types of defects as a result of the 59 work performed in that area. The team also decided to focus specifically on the defects that were associated with manufacturing. Other defects that were linked with previous component removals or engineering related issues were outside the scope of this project. Through the use of pivot tables that linked into the central EMC quality data base, one can easily create a Pareto of the top defects for each area over a specified period of time. One concern by the site leadership was to identify potential repetitive defects that occur on every aircraft built in the EMC. Establishing that a particular defect occurs multiple times is actually quite difficult. First, as noted in Section 3.2.5, the aircraft in work at the EMC were grouped into several categories based on the amount of work and anticipated cycle time to complete the work. The aircraft in the different categories have vastly different work statements, and even aircraft that were listed within the same category may not have the same jobs in their work statement. This means that an aircraft could have a large concentration of defects in a particular area when no other aircraft experienced the same behavior. Additionally, different inspectors may observe and record the same defect on different aircraft but their reports will not match up, which makes it more difficult to recognize the similarities in the defects. Lastly the majority of jobs may require the installation of several parts that are similar. If a defect is written for one of the parts it is not often known which part within the job the defect occurred on specifically. The fixed position facility layout also required a different approach be taken for the data analysis when comparing the quality performance of a specific volume to that of previous aircraft. In a typical factory line, one can analyze the date for a specific position on the line as the work performed will be largely the same and occur on a cyclical basis. However with a fixed position model, the entire build cycle is lumped into the data, thus making a comparison between the same volumes of multiple aircraft requires that the elements of time and schedule be added in as well. In 60 order to analyze how the quality performance of the same volume on different aircraft compared you would need to ensure the comparisons are made during the same phase of the build cycle. This is because different phases of the build lend themselves to different types of quality defects. This behavior is shown in Figure 17 which shows two of the most common defects over a period of time for a given volume of the aircraft. To see if the quality performance of a volume of the aircraft was improving in a specific defect category, it was important to make sure that the comparison was using the correct phase in the build cycle and that similar jobs are used in the sample set. Number of Defects vs. Time - Defect 1 - Defect 2 Figure 17: Defects vs. Time Due to the unique build packages and dynamic nature of the work performed at the EMC, the team decided that the best way to disseminate the data for the individual manufacturing teams was to leave it in its rawest form and focus on training the industrial engineers in data analysis methods. The training would focus on using techniques described in Chapter 4 including Pareto analysis, statistical analysis, histograms and pivot charts and tables. Each team on the shop floor was supported by an industrial engineer (IE) who was also responsible for scheduling and sequencing 61 the jobs for that volume. If an NC was written for a volume, the IE would be the one to schedule it into the existing build plan. Their familiarity with the jobs and scheduling would also help them be more effective with the data analysis. One of the drivers behind creating the database was to bring more freedom to analysts to manipulate the data and perform their own analysis. Previously this was not possible with the company's current data visualization program. This allows the teams to build stronger business cases for pursuing improvement projects by gathering additional data relevant to the project To assist the industrial engineers with some of the analysis, the project team created Excel templates that contained pivot tables and charts that could be easily manipulated to analyze different areas of the aircraft. The sheets were linked directly to the Access database and update automatically when the Excel file is opened. An example of one of the tables that is generated containing the top five defects is shown in Figure 18. The author led training seminars in the use of Microsoft Excel and Access, pivot tables, and sorting and filtering. These seminars also focused on the goals of the quality management initiative and the industrial engineers role in supporting the initiative. Defect Type Defect 1 Defect2 Defect 3 Defect4 Defect 5 Grand Total Top 5 OperatiSnIDefects (NCs) for Area XXXX Sum ofRework Hours %oftoat NC Count 1000 900 800 700 600 4000 25% 23% 20% 18% 15% 100% Ag Rework Hrs/Defect 10 10 10 10 10 50 100 90 80 70 60 80 Figure 18: Top Five Defect Table The project team also viewed the industrial engineer's support of the data analysis for the manufacturing teams as a way to strengthen the relationship between the manufacturing managers 62 and the engineers supporting their areas. The industrial engineers are responsible for the sequencing and scheduling of the jobs for the area that they support A common complaint amongst both managers and installers was that they were constrained from completing their work because of poor sequencing of the work. As defects are generated, the industrial engineers also schedule the rework into the build package, often with little knowledge of the severity or impact of the defect and its fix in regards to rework time. As the industrial engineers become more familiar with the average rework time for different types of defects they can more accurately schedule them into the work plans. One effort to open up communication between the industrial engineers and the manufacturing teams was to have them work together at a table on the shop floor for a portion of each day. The project team believed that having the manufacturing teams and the industrial engineers collaborate on improvement projects would continue to strengthen this relationship and encourage more open communication and collaboration. 5.2.2. Proactive Data Solution One of the goals of the quality management initiatives was to take a proactive approach to the defects and rework at the EMC. In an effort to stay ahead of the defects, the project team developed a predictive report that focused on part defects at the EMC. The report focused specifically on second issue parts which are parts that had defects that could not be repaired or were lost and needed to be reissued. Second issue parts at the EMC alone accounted for a loss of nearly $5M annually in part costs alone. This problem was further aggravated by the lead times required for many of the new parts to arrive at the facility. Because the facility performed modification work, it did not order many parts on a regular basis or with any significant volume. Also, across the company, if parts were needed on other newer aircraft they were given priority over the parts at the EMC. Thus creating a report that showed the most common parts that are reissued could help 63 both to raise awareness on the shop floor of previous damage and to anticipate a potential need earlier in the build cycle so that long lead time parts can be ordered and do not hold up production. The team decided to build this tool in Microsoft Access due to the familiarity with the program and its ease of use. The goal was to develop a tool that would look at scheduled work for a particular aircraft volume or area manager over a specified period of time and if the jobs scheduled had parts that were reissued on other aircraft it would show up in a report for the manager. This report would be looked at by both the manufacturing managers as well as the inventory management group to separate items by lead time. For reissued parts with particularly long lead times, the installers or inventory groups could check the parts ahead of time for damage and then place orders as necessary months before the job is actually going to be worked. Several challenges had to be overcome throughout the development of the tool. First was the issue of how to compare previous jobs to future planned jobs. This problem was rooted in the fact that each job worked in the EMC had its own unique job number. Even if the exact same job or task was scheduled and performed on multiple airplanes, they would have different job numbers, making it nearly impossible to connect the two. This was further complicated by the quality defect data collection. The second issue parts database was managed by the inventory group and was not a part of the quality data. The only reissued parts that were in the quality database were those that had an NC tag written for them because of damage, which accounted for only a small percentage of the total amount of second issue parts. Each time a tag was written for a defect, it also had a unique task number. Also, there existed a parent-child relationship for the defect data collected where an inspector, rather than writing a new tag, could reference a similar tag. Many of the defects would be a part of a long chain of defect tags that were referencing one another. 64 The team was able to get around these issues by creating a separate database for second issue parts that combined both the quality and inventory data warehouses. Then, by using the part numbers, which are consistent with the exception of part revisions, the team developed a series of relational queries that would relate past reissued parts with future scheduled work and automated them with the use of macros. As the industrial engineers build the production schedules they are loaded into a large data warehouse on the main server. Using the second issue parts database and the production schedule data, the macros would run relational queries that would compile data on all the jobs scheduled that contained parts that had been reissued in the past. This data was incorporated into a report that showed information on reissued parts including: e Which aircraft the part was reissued for * Part number - Part name/description e How many parts were reissued - The unit of measure for the part - Where on the aircraft the part was located - The reason the parts were reissued e When the part was reissued The manufacturing and inventory groups determined that the tool should be set up to look two months ahead in the schedule. The team then created a simple user interface in Microsoft Excel that would allow the user to select the aircraft they were working on and their respective work package. The user could also choose how the parts would be identified, either by part name, description, number, or a combination of the three. An example of the tool's interface with the actual data removed is shown in Figure 19. 65 Aircraft Number xxxxx <-- Select line number Supervisor xxxxx <-- Select Volume 2nd Issue Parts Predictive Report Planned Work 8/16/2013 - Total Job Number - Total WINDOWCLIP 8/16/2013 -Total Job Number -Total BRACKET WINDOW CLIP ASSEMBLY 8/16/2013 -Total Job Number -Total WIRE HARNESS 8/19/2013 -Total Job Number -Total SHIM YOKE WINDOW 8/20/2013 -Total Job Number -Total BRACKET PANEL Defect Reasons DPRT 3 3 3 7 7 7 Parts Parts Parts Parts 24 24 24 2 2 1 1 Parts MioQM Grand Lost Parts Revision NC Removal Total 5 1 9 5 1 9 5 1 9 22 4 10 1 44 22 4 10 1 44 22 4 8 1 42 1 1 1 1 100 4 128 100 4 128 100 4 128 1 3 1 3 1 1 1 1 4 1 5 4 1 5 3 3 1 1 2 Figure 19: Second Issue PartsReport 5.3. Chapter Summary To support the quality management initiative the team focused on providing the right tools and accessibility for efficient analysis and dissemination of data. As many of the company-wide software tools were not tailored to the specific needs of the EMC, these tools had to be developed in house and customized to increase their effectiveness. Despite the limitations noted in the data, there is a lot that can be learned by comparing similar volumes across multiple aircraft and 66 monitoring current performance. Creating a centralized database provides the needed access and ease of use for quality data to support the quality plan. Incorporating tools to collect additional information on caused vs inherited defects will also help to increase the reliability of the data and ensure proper communication between managers and their employees closer to the time of occurrence of the defects. The proactive tool focused on identifying scheduled jobs that have parts with high rates of replacement is especially important in the EMC as low volume and frequency of the orders add to the long lead times. Anticipating long lead time parts that may require replacement will help to reduce the overall flow time for the aircraft. Also identifying which parts that are scheduled to be installed are the most commonly damaged will raise the awareness from installers and caution them to take additional care during installation. In 2012, the EMC spent over five million dollars on second issue parts which does not take into account the overhead and down time associated with the parts. 6. Medium Blow Test Project 6.1. Team Goal and Composition A targeted, cross-functional team was formed to address an issue in the EMC that was identified using the Theory of Constraints and Lean frameworks as a bottleneck of the build cycle for an aircraft in the EMC. The goal was to analyze and improve process that would reduce the amount of medium blow tests performed on an aircraft. The team included representation from industrial engineering, manufacturing, manufacturing engineering, quality, and lean departments. The team used the tools developed for the quality management initiative and the combined improvement methodologies outlined in Section 4.2 of this thesis. 67 6.2. Background The medium blow test is designed to test how well the aircraft's fuselage can maintain cabin pressure. The test involves pressurizing the fuselage to a specified pressure and monitoring how long the cabin will maintain the pressure. If the inside pressure drops rapidly after the air pressure is no longer applied it indicates that there are leaks around the fuselage area. This is a common test in aircraft manufacturing and represents a major milestone in the build cycle. While the test is being performed, the fuselage must be free from employees and the testing team needs to have access to all areas surrounding the fuselage. Some leaks are concentrated and easily found as they grow very loud due to the air escaping at a very high velocity. Other leaks are very subtle and require the use of smoke or feathers to detect their existence. The more subtle leaks often go unnoticed in the first tests if there are other more significant leaks. This is because they will not see as much pressure because the air is leaking out of the larger leaks. In order to detect all the leaks, the test crew must survey the entire area of the fuselage. Due the access and personnel constraints associated with the test, the test is always scheduled on a late shift when the least amount of employees are present in the building. 6.3. Approach For efficient project management and organization, the team first outlined a project schedule following the DMAIC format which was included as a tool in the project management deck created for the quality management initiative. The details of this are included in Figure 20. The team first analyzed historical data from a large sample set of past blow tests to identify the most chronic parts and locations where leaks are found on the aircraft. The aircraft was divided into sections for the analysis and a tool was used to map the exact location for each discrepancy to show areas of concentration. The leaks were further categorized by keywords indicating which part, or between 68 which parts, was the source of the leak. The team was able to generate a report that isolated the primary parts and locations, as well as the percent of the aircraft sample, where the leaks were found. In addition to the quality data, the team observed the medium blow test in its entirety. Through interviews and observations, the team built a process flow map for the test. The process flow map is shown in figure 21. A larger version of this figure is also included in Appendix C for ease of reading. blow test is a system bottleneck " Goals: reduce blow count and total cycle time for aircraft build * Conduct manufacturing/quality interviews " Medium Define 2 j I Measure * Observe inspection process - Observe documentation/defect tracking process " Map process - identify inefficiencies and communication barriers Analyze " Analyze historical test results on similar aircraft - Identify chronic defects and locations " Perform root cause analysis on cause of test failure Improve " Create future state process flow map - Streamline documentation, defect tracking, and communication handoffs " Reduce cycle time for test Control * Create pre-inspection job to identify and fix known discrepancies prior to 15 blow test, constrain scheduling of test to job completion " Establish constraint relationships between leak repairs and test schedule " Formalize test status meetings, increase communication/accountability Figure 20: DMAIC Project Outline for Medium Blow Test 69 I PROCESS FLOWCKART Who ", Input -- .M M-air -TnTes Promee MwNber N Io Cr*SWom Dats; riO[MLNY for. M rIinr, TMPX TA-ul Procem. Owuir _ Revmkrn Date: [Mo., Tom le I output POW" FHCM ..... .... Output ---------(iii -----------) ........ IMM ... ------ 'am VOW !i- Oiv ------------- ------------ VF WN - --- --- -- - -- -- ---------- am - N ----------- ----------- ................ ---------- SNE54RWU --------------------- 1--l------------------...-------...---.. ---- ,-- ----------------- I- - ------- ----------------------------------------------- 1e FT "M Figure21: ProcessFlow for Two Tests 6.4. Key Findings and Recommendations Based on the data analysis and observations, the team went through a root cause and corrective action analysis for the test. The team then identified several causes that lead to a higher amount of blow tests required for each aircraft. Figure 22 highlights several of the key findings along with proposed solutions that the team believed would decrease the amount of test runs needed to pass the aircraft. 70 Issues Solutions - No pre-inspection process to check for common failures before first test Require pre-inspection utilizing checklist w Focus pre-inspection effort on most 'Preliminary cabin pressure leak checklist. a is not utilized - Poor communication between m EMC and test team - common failures Constrain test scheduling to preinspection job completion Formalize information handoff between test team and EMC UKeep pressurization test team included - Only communication between the test team and Mfg is the test report in all updates affecting status and issues - - Tests scheduled before discrepant leak areas are fixed - Lack of accountability to complete open jobs after each test - Set up constraint relationships and schedule status meetings prior to scheduling next test Create job linked to aircraft volumes for defect tags Figure22: Results from Root Cause and CorrectiveAction Analysis 6.4.1. Pre-inspection Process The first, and perhaps most important, issue that the team found was that there was no preinspection process for the medium blow test. Through the analysis of the quality data, the team was able to identify several locations and specific parts of the aircraft that were especially prone to leaking. After observing its first test, the team, along with the manufacturing leads and quality inspectors, visually inspected all the areas found to be leaking. Several of the leaks had already been identified by the team as the most common parts and locations likely to leak. The team was surprised to find that many of these potential leaks could be identifiable prior to the test through typical visual inspection techniques. For example, the kick plates at the doors often did not contain enough sealant to be airtight An inspector could just tap the plate with an object such as a quarter and hear whether or not the plate was properly sealed. If this was checked before the test the plate 71 could be removed and re-sealed prior to the test to eliminate the chance of leakage. In some instances the leaks were so severe that they were likely masking other leaks nearby which would not be found until the next time the aircraft was tested. Not addressing areas known to commonly leak leads the EMC to perform more tests than necessary as many leaks will not become detectable until the more severe ones are sealed. To solve this issue, the team recommended the creation of a pre-inspection job that contained a checklist which contained instructions for inspection of specific parts and locations that commonly leak during the test. The pre-inspection instructions would be written specifically to look for instances that would result in an air leak. It was also found to be much more economical to create jobs to remove, seal, and reinstall specific components of the aircraft prior to the test based on their probability of failure. The team also recommended that the test be constrained by the completion of these jobs before it could be scheduled to maximize the usefulness of the first test. 6.4.2. Poor Communication The team observed a large disconnect between the testing team and the manufacturing teams who would actually perform the sealing jobs. After a test was complete, the only information handoff was a test report that contained the test results and discrepancies and locations. This was provided to a manufacturing manager on the pervasive team in charge of scheduling the test. That manager would then disseminate this information verbally in a meeting to all the manufacturing managers with leaks in their areas. Each manager would give an estimated time for completion of the discrepancies and the next test would be scheduled based off the verbal commitment of the managers. 72 Not having any information from the test team about the leaks meant that the installers and manufacturing managers had to do their own investigation for the repair. The managers would commit to a schedule to complete the repairs before they had a sense of how long the jobs would take to complete. Sometimes jobs would be held up because of lead times of parts needed to complete the repair. Further, they had not yet coordinated with the industrial engineers who schedule the work for their areas to see when they had installers who could work on the repairs. This would often result in the next test being scheduled priorto the completion of all the defectsfound on the previous test This cycle would drive up the total number of tests requiredcausing significant schedule delays, wasted time and money. To address the issue of poor communication the team recommended that mandatory status meetings be scheduled and attended prior to the scheduling of the second blow test. This will ensure that everyone will meet the deadline and reduce the chance of pressurizing the airplane with discrepancies that have not yet been repaired. The team also recommended a similar constraint relationship be established for the defect tags written as a result of the test. This would prevent the test from being scheduled with open defects and help to reduce the total amount of tests required. 6.4.3. Lack of Accountability The last issue the team sought to address was that of accountability. The lack of accountability was found to be a result of the way the jobs were written and scheduled. After each blow test, all of the leaks found would be written in a single inspection report that referenced the test job. The report could contain tens of defects detailed in its text in any location in the fuselage, making it difficult to identify each one for the installer completing the repairs. Further, because the defects are all referencing the test job, they do not actually show up on the schedules of the area managers. 73 Managers are only held accountable for completing the scheduled jobs for their area, so if a repair job from the medium blow test is not on a manager's schedule they cannot be held accountable for completing the work in a timely manner. As many managers are busy dealing with the daily firefighting that is inherent in their jobs, it is easier for them to prioritize their own work for which they are accountable over that associated with the blow test. The lack of accountability was identified as the root cause of the managers not meeting their initial commitments and having to push back their estimated completion dates. The team's final recommendation was to create a rework job for each area manager in the fuselage section of the aircraft. This job would only show up on an area manager's schedule if there were defects found during the test in that area. This would significantly help the overall management of the medium blow test process. The repair work could also have constraint relationships with the test so that its schedule was dependent on their satisfactory completion and inspection. 6.5. Discussion of Results The personnel and space constraints, scheduling, and equipment mobilization all contribute to the impact the medium blow test has on the total aircraft build cycle. Reducing the amount of tests required to pass the airplane reduces a major bottleneck in the process. Not meeting delivery schedules has a far reaching effect on both the company and their customers which results in significant costs to both parties. The team believed that they could reduce the total time for the blow tests on all remaining aircraft for the EMC by 60% by following the recommendations and changes to the test management process. This would result in roughly $200,000 - $300,000 savings per aircraft. Incorporating the key stakeholders into the project team was imperative to achieving buy-in and acceptance of the 74 changes proposed. At the conclusion of the internship, the leadership team at the EMC had agreed with and approved all changes to the process, including the creation of the jobs to improve the overall process management. 7. Recommendations and Conclusions The successful implementation of a quality management strategy is an ongoing process that requires several years before significant gains are realized. When developing a quality management strategy for a manufacturing facility that faces the challenge of low repeat work, large breadth of work, and a low planning horizon, it is important to focus not just on a single improvement methodology, but rather incorporate several aspects of many. Limiting the focus to a single methodology can constrain the effectiveness of the management system as a whole. With any method chosen, it is imperative that management provide the proper support to enable greater involvement, engagement, and idea generation at the lowest levels of the organization as they are the ones closest to the product and the problems. Additionally, making data easily accessible, and training employees in analysis techniques is critical for identifying high impact defects and driving actions for improvement projects. 7.1. Recommendations In order to continue the implementation effort and build a culture of quality at the EMC, it is imperative that continuous support for the vision and goals of the initiative is present for a sustained period of time. While the author was on site three changes in site leadership occurred, which significantly disrupted the momentum and adoption of the quality management strategy. This recommendation is further supported in Chapter 2 by several case studies that identified management's role and support for change initiatives as one of the primary drivers for their adoption in an organization. While it may not be possible to maintain a consistent set of individuals 75 on site, there should be a greater focus on information handoffs and knowledge transfer during these times of transition to ensure that progress is not lost. There must also be a clear vision that is communicated frequently and effectively at all levels of the organization for the quality initiative, especially during the early development stage of the initiative. An effective communication strategy will engage all levels of management to bring them on board early rather than blindside them all at once with a new initiative. It is also important to communicate the impact of the programs adoption at all levels to help achieve more commitment and buy-in from lower levels of management This will ensure that the initiative is being carried out because managers know if and believe in the benefits of the program rather than just doing what they are told. Commitment is always better than compliance. The author also believes that the most effective implementation strategy for the quality management initiative would be to first focus on a few smaller teams of selected individuals who are highly engaged to implement improvement projects rather than attempting to engage all employees at once. This will ensure a greater chance of success and bring greater support to the teams and their projects (Beer, 2003). Once several projects have been carried out, the success of the teams can be exploited throughout the organization and the momentum from the teams can be carried forward into other teams as well. Not every individual will be willing to change and adopt a new strategy. By focusing first on the individuals who are more willing to adopt the initiative, the site management will have a greater chance for early success and execution of improvement projects. It is also recommended that an incentive structure be built in to a quality management initiative. It is well known that employees are, in part, motivated by incentives (Naor et al.2008). These can be 76 an effective driver for increased adoption rates across all employees. It should further be encouraged that managers actively engage in verbal encouragement as incentives as well. In the survey conducted by the project team, one of the top motivators identified by employees was just a verbal recognition that they had done a good job. This should not be overlooked when deriving an incentive strategy for a quality management strategy. 7.2. Conclusion and Hypothesis Assessment At the end of this project, the quality management initiative was gaining significant momentum in its adoption and success. This was due primarily to the new site leadership that was installed two months prior to the end of the project. The new leadership was not a part of the original strategy development for the project, but took a positive stance on the gains made by the project team and committed to continuing to push for the project's implementation. The importance of support and commitment from top management on the success of a change initiative such as the one detailed in this thesis became very clear during this transition. With constant changes in management it becomes increasingly more difficult to support an initiative such as a quality management program. The momentum gained towards the end of the project, paired with the effective use of the tools developed and the projects initiated support the hypothesis that a combined methodology for quality management is the most effective strategy. Several projects were identified and kicked off prior to the author leaving the site that were facilitated using the tools from lean, six sigma, and TOC. These projects all had targeted reductions in rework that totaled over 10,000 man-hours and significant cost savings for each aircraft still to be worked at the EMC. However, the sustainability of the plan was not certain at the end of the author's time on site. The results and adoption relied directly on the continued support from management. While several projects were identified and worked through the root cause and corrective action phases to develop action plans at the end of 77 the author's time on site, the actual implementation results were not observed thus no concrete evidence to support the hypothesis can be presented. Works Cited Akbulut-Bailey, A.Y., Motwani, J. and Smedley, E.M. (2012) 'When Lean and Six Sigma converge: a case study of a successful implementation of Lean Six Sigma at an aerospace company', In.J. Technology Management,Vol. 57, Nos. 1/2/3, pp.18-32. Andersson, R., Eriksson, H. and Torstensson, H. (2006)'The similarities and differences between TQM, Six Sigma and lean', TQM Magazine, Vol. 18, No. 3, pp.282-296. Baldrige National Quality Program (2005). CriteriaforPerformance Excellence. National Institute of Standards and Technology. Gaithersburg: Department of Commerce. Beer, M. (2003).Why total quality management programs do not persist: The role of management quality and implications for leading a TQM transformation. Decision Sciences, 34(4), 623-642. Beer, M., and Eisenstat, R. A. (2000). The silent killers of strategy implementation and learning. Sloan Management Review, 41(4), 29-40. Cameron, K. S., and Quinn, R. E. (1999). Diagnosing and changing organizationalculture: Based on the competing valuesframework. Reading, MA: Addison-Wesley. Campanella, J. (Ed.). (1999). Principlesof quality costs: Principles, implementation and use (3rd ed.). Milwaukee, WI: American Society for Quality, Quality Cost Committee. Dahlgaard, J. J., and Dahlgaard-Park, S. M., (2006). Lean production, six sigma quality, TQM and company culture. The TQM Magazine, Vol. 18, No. 3,263-281. Dahlgaard, J.J., Dahlgaard-Park, S.M. and Edgeman, R. (1998a). Core value deployment - the need for a new renaissance. Total Quality ManagemenL Vol. 9 No. 4. Dahlgaard, J.J., Dahlgaard-Park, S.M. and Edgeman, R. (1998b). Core values - the precondition for business excellence. Total Quality ManagemenL Vol. 9 No. 4. Dooyoung, S., Kalinowski, J. G., and& EI-Enien, G. (1998). Critical implementation issues in total quality management. SAM Advanced ManagementJournal,63(1), 10-14. Drury, C. G., and Sheehan, J.J. (1968). Ergonomic and economic factors in an industrial inspection task. InternationalJournalof ProductionResearch, 7(4), 33 3-341. Drury, C. G., Spencer, F. W., & Schurman, D. L. (1997). Measuring human detection performance in aircraft visual inspection. Proceedingsof the Human Factorsand ErgonomicsSociety Annual Meeting, 41(1), 304-308. 78 Drury, C.G. (2001). Human reliability in Civil Aircraft Inspection. Paperpresented at the RTO HFM Workshop on "The Human Factorin System Reliability - Is Human Performance Predictable?"held in Siena, Italy, 1-2 December 1999, and published in RTO MP-032. EFQM (2003). Excellence FundamentalConcepts. Brussels, European Foundation for Quality Management Flynn, B. B., Schroeder, R. G., & Sakakibara, S. (1995). The impact of quality management practices on performance and competitive advantage. Decision Sciences, 26(5), 659-691. Giakatis, G., Enkawa, T. and Washitani, K. (2001). Hidden quality costs and the distinction between quality cost and quality loss. Total Quality Management.Vol 12, No 2, pp. 179-190. Goldratt, E. M. and Cox, J., (2004). The Goal: A Process of Ongoing Improvement, Third Revised Edition, North River Press. Goranson, U. F. and Rogers, J.T. (1983). Elements of Damage Tolerance Verification, 12th Symposium of InternationalCommercialAeronautical Fatigue,Toulouse, France. Gordon, Dale K. "The Past, Present and Future Direction of Aerospace Quality Standards." Quality Progress.American Society For Quality, June 2000. Web. 21 Dec. 2013. <http://asq.org/qualityprogress/2 000/06/standards-outlook/the-past-present-and-future-direction-of-aerospacequality-standards.html>. Gryna, F. M. Quality and Costs. In: Juran, J. M. & Godfrey, A. B. (1999).Juran'sQuality Handbook.New York: McGraw-Hill. 5" Edition H. James Harrington, (1999). Performance improvement: a total poor-quality cost system. The TQM Magazine, Vol. 11 Iss: 4, pp. 2 2 1 - 230 Hackman, J. R., and Wageman, R. (1995). Total quality management: Empirical, conceptual and practical issues. Administrative Science Quarterly,40, 309-342. Hendricks, K. B., and& Singhal, V. (2001). The long-run stock price performance of firms with effective TQM programs. ManagementScience, 47, 359-368. Iyengar, S. S., and Lepper, M. R. (2000). When choice is demotivating: Can one desire too much of a good thing? JournalofPersonality and Social Psychology, 79(6), 995. Kasemset, C., (2011). A review on quality improvement and theory of constraints (TOC). 2011 IEEE InternationalConference on Quality and Reliability,ICQR 2011 Kaynak, H. (2003). The relationship between total quality management practices and their effects on firm performance. Journalof OperationsManagement,21(4), 405-435. Klefsjo, B., Bergquist, B. and Edgeman, R.L. (2006)'Six Sigma and total quality management: different day, same soup?', InternationalJournalof Six Sigma and Competitive Advantage, Vol. 2, No. 2, pp. 1 6 2 - 1 7 8 . 79 Krishnan, S. (2006). Increasing the visibility of hidden failure costs. MeasuringBusiness Excellence. Vol 10, Issue 4, 77-101. Love, P. E. D. (2002). Auditing the indirect consequences of rework in construction: A case based approach. ManagerialAuditing Journal,17(3), 138-146. Lunsfordj. (2007). Boeing Scrambles to Repair Problems with New Plane. The Wall StreetJournal, December 7,2007. Martinez-Jurado, P.J., and Moyano-Fuentes, J. (2012). Key determinants of lean production adoption: evidence from the aerospace sector. ProductionPlanning& Control.Vol. 25, No. 4,332345 Moselhi, 0., Assem, I.,and& EI-Rayes, K. (2005). Change orders impact on labor productivity. Journal of ConstructionEngineering& Management,131(3), 3 54-359. Naor, M., Goldstein, S., Linderman, K., and Schroeder, R. (2008). The role of culture as driver of quality management and performance: infrastructure versus core quality practices. Decision Sciences, 39(4), 671-702. Nave, D. (2002) 'How to compare Six Sigma, lean and the theory of constraints', Quality Progress, Vol. 35, No. 3, pp.73-78. Owens, B. D., Leveson, N. G., and& Hoffman, J. A. (2011). Procedure rework: A dynamic process with implications for the "rework cycle" and "disaster dynamics." System Dynamics Review, 27(3), 244269. Rahman, S., (2002) "The theory of constraints' thinking process approach to developing strategies in supply chains", InternationalJournalof Physical Distribution& Logistics Management,Vol. 32, No. 10,809-828. Rollinson, D., Broadfield, A., (2002). OrganizationalBehavior and Analysis - An IntegratedApproach, second ed. Person Education. Rosenfield, D., and Reavis, C. (2009). Boeing Commercial Airplanes' 787 Dreamliner. MIT Sloan School of Management Salmador, M. P., Bueno, E. and Maranhao, R. (2008). Total quality management: a critical analysis from a complexity approach. Total Quality Management Vol. 19, No. 5,513-533. Shortell, S. M., O'Brien, J. L., Carman, J. M., Foster, R. W., Hughes, E. F. X., Boerstler, H., et al. (1995). Assessing the impact of Continuous Quality Improvement/Total Quality Management: Concept versus implementation. Health Services Research,30(2), 377-401. Spector, B., and Beer, M. (1994). Beyond TQM programs.Journal of OrganizationalChange Management,7(2), 63-70. Thomasson, M., and Wallin, J. (2013). Cost of poor quality; definition and development of a processbased framework. Departmentof Technology Managementand Economics. Chalmers University of Technology 80 Womack, J.P., Jones, D.T. and Roos, D. (1990) The Machine that Changedthe World, Free Press/Simon & Schuster, Inc., New York, NY. Zarbo, R.J. (2012). Creating and sustaining a lean culture of continuous process improvement. American Journalof Clinical Pathology.No. 138, 321-326 Appendices Appendix A: EMC Quality Plan Document EMC Quality Strategy for 2013 Developed by the EMC Quality Strategy Team Updated 6/28/2013 81 Table of Contents Executive Sum mary ................................................................................................................................... 83 2013 Goals and M etrics ............................................................................................................................. 84 Quality Performance Dashboard ............................................................................................................... 84 Project M anagement Flow for EM C Quality Plan ...................................................................................... 86 Project monitoring ................................................................................................................................. 86 Communication ......................................................................................................................................... 86 Daily Com munication: ........................................................................................................................... 86 W eekly Com munication ........................................................................................................................ 87 M eetings ................................................................................................................................................ 87 Projects .................................................................................................................................................. 88 Training ...................................................................................................................................................... 89 Tools .......................................................................................................................................................... 89 W orkcell Action Tracker ........................................................................................................................ 89 Predictive Reports ................................................................................................................................. 90 Databases .............................................................................................................................................. 90 82 Executive Summary This user guide is an overview to heighten quality awareness and share the established process for the EMC Quality Initiatives that have been identified by the leadership team as focus areas for 2013. The Quality Initiatives directly tie in to the EMC vision with a focus on First Pass Quality. This document explains the process and target condition of how the EMC production and support system will work to lower the amount of defects that are generated here in the EMC. The goals of the EMC quality plan are twofold: 1. To address defects and capture/communicate learning of the root cause and corrective actions as soon as they occur during the build cycle. 2. Look at past impacful discrepancies throughout the build cycle to better plan, develop, and share mitigation strategies as we move forward with our work at the EMC. 83 2013 Goals and Metrics The EMC leadership has identified 5 focal areas for quality improvements which include, FOD/5S audits, tool control, NC/100 reduction, 2 nd issue parts reduction, and improved first pass quality for shake inspections. The 2013 goals are shown in Figure 1 below. The EMC Quality Plan will use cost code 4 (CC4) hours as well as defect generation (count) as metrics to help identify the high impact NCs that are caused in the EMC. Looking at rework hours associated with both the NC's and pick-ups (PU) will allow us to see how much time we are spending on specific issues across the EMC and to prioritize improvement projects based on the amount of rework time to be saved and impact to the build. Reducing time spend on reworking discrepancies will increase the amount of time spent on the build of the airplane. 787 EMC Quality / Compliance Situation Target (Measurable) Improve our Quality and Compliance from 2012 in the following areas: " NC's/100s - FOD / 5S " Tool Control 1. Achieve 10% improvement in NC's/100 2. 5S - Level 3.0 3. Maintain Tool Control monthly audit passing rate at 90% 4. Improve 1st pass quality for BAC and Customer Shakes to 95% 5. Reduce 2nd Issue parts for DPRT's and Manufacturing caused NUT tags by 10% " BAC/Customer Shakes " 2 "d Issues Parts Actions All managers have quality improvement plan *RCCA for alltMFG caused NCs____ - Work with workplace coaches - SEC training and Assessment F Inlate predictiveanalysis NCs/100 - Continue weekly FOD sweeps Participate in the EMC 5S plan Teol Control Continue review metrics weekly and have action plans in place - All El teams have a minimum of I Tool Control improvement project BAC/CustomerShakes - RCCA for all escapements - Develop shake defect checklists 2"4 Issues Parts W___________________W__"i"_I-_!-lope - - - . Fob - - k 0 - Perform RCCA for all shop charged detective parts Initiate predictive analysis Share lessons learned on all above initiatives Recognize teams for improvements and reduction "rmum 81Wmow" Disution Uimitad to Boeing 4 sw -. Aav Personel vid, a Demnstrated Need to Know Figure 1 Quality Performance Dashboard In an effort to increase visibility, accountability, and awareness of quality issues, the EMC will distribute performance dashboards to each volume across the EMC on a weekly basis. The performance dashboard will display information relating to the EMC quality initiatives. These dashboards will utilize both visual and quantitative data that reflects the quality performance of the volumes that they support. The quality performance dashboard will also be supported at the general management, airplane, and EMC levels. The general and airplane level metrics can be found at the following location: General and AP level dashboards, and the volume level here: volume level dashboards. 84 Figure 2 is an example of the dashboard to support the team for each volume of the airplane. The quality performance dashboard will be updated by the lEs supporting the volumes each week by COB Wednesday. The IEwill discuss data with the 1s line managers and aid them in identifying and managing their quality projects. Stoplight metrics for RCA on defects, shakes, and project status. 1 st 2 nd 90 Day 2nd Issue Parts Data Shift Team Status for 25- RCCA Status - NC's RCCA Status - 2"d Issues Shake Training Completed FPQ BAC/Customer Shakes Skill Team Assessment Skill Team Gap Quality Project Status Kaizen Paper Status issue parts each volume each week 10 80PRT OtmtParts P 2'd Shift Team Status 1. 2. 3. 4. RCCA Status - NC's RCCA Status - 2"d Issues Shake Training Completed FPQ BAC/Customer Shakes Skill Team Assessment Skill Team Gap Quality Project Status Kaizen Paper Status 787g EM* LNe 20 5. Top 5 DPRTs: Wiring Bracket - 17 DPRT, 22 Items Connector/backshell- 13 DPRT, 18 Items Insulation - 9 DPRT, 9 Items Blanket - 7 DPRT, 7 Items Drip Shield - 7 DPRT, 7 Items Top 5 Lost Parts: Marker - 16 LP, 27 items Wiring Bracket - 13 LP, 14 Items Wire Harness - 5 LP, 5 Items Bushing - 2 LP, 8 Items 5. Spanner Bar - 2 LP, 6 Items 1. 2. 3. 4. ult . - ei efrac ahor Volume level CC4 Data for Ops De.- Weekly PU/NC Generation Ops NC/PU Rework Hours by Week 10 0 - ONCOPS 1 PU L *OIrR.C Top NCs Over 90 Days Defect Part- Eng I rawing/Design OPSNC Part - Dama ged Hole/CS - h islocated Hole/CS - CIversize/Undersize Part - Ridin g Part - Mislo rated Wire/Conne ctor - Damaged Hole/CS - I issing/Extra Test - Failed Top PUs Over 90 Days Count Hrs 150 97 84 62 54 42 NPU Avg _ 7 10 6 3 7 4 21 10 14 21 8 10 39 4 10 39 9 4 36 21 2 1 18 21 Defect Fasten or - Missing/Extra Coatin g/Covering - Missing/Extra Hole/C 5- Missing/Extra Coatin /Covering - Mislocated Part - Incorrect Part - Missing/Extra Seal/Luube- Incomplete Placard s - Missing/Extra 85 Faster r - Opposite/Reversed Direction Duct/H ose/Tube - Damaged Hrs CountAvg 17 3 6 9 2 5 7 1 7 6 1 6 5 2 3 3 2 5 5 2 2 1 5 5 3 2 1 1 3 2 Figure 2 Project Management Flow for EMC Quality Plan Following the established process developed for the EMC of tracking, analyzing, and reporting data will drive out impactful projects to help increase build rate and reduce defect generation. Proper project management and communication will be required to ensure that improvement projects reach successful completion and the quality goals for the EMC are achieved. Participation from both manufacturing and quality leadership will be required to enable successful execution of project findings. Projects will be identified and supported through both data analysis utilizing the databases that have been developed for the EMC and throughout the build cycle by managers and their teams. Project teams will include cross-functional support throughout the problem definition, root cause analysis, solution development, implementation, and sustain phases. Tools from the Boeing Problem Solving Model (BPSM) shall be used and included in project reports. The following is a link to the BPSM site: http://leanplus.web.boeing.com/NavTool/ Project monitoring To ensure proper execution and sustainment of quality improvement projects, a robust monitoring process will be required. A weekly project status meeting will be established between the superintendents and the general managers as well as with site leadership. The projects will be monitored through completion which will be marked by the generation of a kaizan newspaper. Industrial engineers supporting the builds and scheduling the work will ensure implementation actions are sustained throughout the build cycle and carried between airplanes. Each project will result in the generation of a kaizan newspaper that details the project findings and solutions. These will be categorized by affected volumes, defect types, and build cycles for ease of reference. The kaizan newspapers will be shared with all managers affected by the solutions through a presentation during the regularly scheduled report outs. Industrial engineering analysts will be the primary role to create and maintain the standard quality performance dashboards that will be used to support the manufacturing defect and CC4 hour reduction plans. The lEs will also help the project teams gather data that is related to their Quality projects to help track results from successful implementation. Communication The following section details the general communication process to support the quality plan: Daily Communication: * * Each morning at the boardwalk meeting the airplane captain will run through the previous day's NC and 2nd issue data with all managers present and ready to speak to their volume's data. For each NC and 2nd issue part lt line managers will discuss with the MT working the SOI to gather/record root cause and corrective action (RCCA) data according to the procedure as detailed here: Defect reduction flow Here is a direct link to the Work cell Action Tracker where all RCCA information is to be captured for caused NCs: Predictive Analysis Mgmt Tool 86 Weekly Communication Volume level lEs supporting each volume will develop and distribute the quality performance dashboard for the first line managers whom they support. lEs will then discuss the data with the 1 st line manager when presenting the hard copy. The hard copy will be posted to the visibility board on the shop floor above the bar chart for each volume by COB each Wednesday. The dashboard template can be found here: Quality Plan 4SQ A guide to the stoplight metrics can be found here: Quality Plan - Stoplight Metrics 1t lines will discuss the data and findings with their crews on a weekly basis to look at running metrics and group performance. General managers will discuss projects weekly with project teams to update status and any help needed to continue project moving forward. * * e General/AP level Airplane captains will share root cause and corrective action findings bi-weekly at the status to plan meeting. Generals to discuss metrics with superintendent at weekly staff meetings and give high level quality plan overviews. * Meetings EMC/AP level meetings: The weekly EMC/airplane level quality meetings will have the following agenda: e Each project owner should briefly answer the following: What did you do last week? What do you plan to do this week? Is there any help needed? * Project owners should briefly discuss issues, findings, and next steps of their projects in work * Discuss prioritization of the Projects The EMC/AP level meeting attendees should have the following roles present * Manufacturing and Quality Superintendent * Manufacturing Generals e Industrial engineers supporting the projects * Initiatives manager As a follow on to this meeting, EMC superintendents should discuss AP level metrics with the Director in a weekly meeting along with an overview of micro/macro level project/action plans that are either currently being worked or identified. Any help needed by teams should also be presented to senior leadership to expedite solutions. The following figure details the meeting schedule to support the EMC quality plan: EMC Weekly Cadence Schedule Monday Tuesday Wednesday 5:30 87 Thursday Friday Previous day's NC Review at Previous day's NC Review at 6:00 Boardwalk 6:30 Boardwalk Previous day's NC Review at Boardwalk Previous day's NC Review at Previous day's NC Review at Boardwalk Boardwalk 7:00 7:30 Ship captains review NC data at STP 8:00 Ship captains review NC data at STP 8:30 9:00 9:30 -A 10:00 10:30 11:00 11:30 12:00 12:30 13:00 prepare and review quality _E's dashboard metrics with 1st lines by COB 13:30 14:00 14:30 15:00 Previous day's NC/PU Review at Boardwalk Previous day's NC/PU Review at Boardwalk Previous day's NC/PU Review at Boardwalk Previous day's NC/PU Review at Boardwalk Previous day's NC/PU Review at Boardwalk 15:30 16:00 Projects Projects will be identified at all levels of the organization and across any function supporting the build of the airplane. Collaboration on projects and implementation should be encouraged across multiple volumes/airplanes. All projects will be tracked on the Quality Projects Master Sheet found here: EMC Positional Quality Plans.xlsx. The Quality Council will be responsible for identifying teams to work macro level EMC quality improvement projects and it should be encouraged to utilize a crossfunctional team of employees across the EMC. 88 The projects will be coordinated with the Employee Involvement program leadership to ensure that we are not duplicating improvement efforts and that knowledge/learning is shared across the EMC. New projects that are identified and not being worked should be documented in the quality projects master sheet which links to projects action plans. A number of project management/RCCA tools are available to the teams and can be found here: PM Tools. Http://lean plus.web.boeing.com/productservice/employeeinvolvement.cfm Both the projects and communication will be initiated and worked with a Cadence. Following this cadence will ensure proper communication and distribution of data and ideas. If any additional meetings are needed, this can and should be arranged. Weekly review of projects at the General level will ensure that communication and support for implementation is there for each project. If a team needs help to execute an improvement project, the General will take it to the EMC/AP level quality meeting to discuss needs/roadblocks to implementation. Training The mechanics at the EMC require a broad skill set to complete their work. The lack of repeat work and movement of employees heightens the need for skill assessments and focused training as the volumes enter different stages of the build. Each team will be audited to track assessment training requirements and to identify areas and employees that will require additional training to reduce caused defects. Training assessments and schedules will be set up such that targeted practice training and peer to peer training occurs during times of the most relevance. For example, teams will be scheduled to go through the peer to peer shake training prior to beginning shakes for their volumes. Each first line manager will be responsible for ensuring that all of their employees attend the required training as determined by the skill assessments, build cycle and quality team. Tools The EMC quality team has developed several tools to assist in the execution of the EMC quality plan. The following section details the tools available. Workcell Action Tracker As the statement of our remaining work indicates that we will see an upward trend in defects that are not caused in the EMC, we must work to separate the data between caused and inherited defects. Reducing the additional "noise" in the data will allow us to show improvements in caused defects and to focus on future mitigation projects for these issues. In order to make this distinction of caused vs. inherited defects, the Workcell Action Tracker (WAT) was developed as a tool to assist in defect tracking. For every NC that is generated in the EMC, the manager associated with the NC will fill in the required fields in the WAT. The purpose of this new documentation process is both to isolate defects that are caused here in the EMC for future mitigation action plans and to increase awareness of these issues at all levels of the organization. We will address issues as close to their discovery as possible to capture all learning and mitigation strategies. 89 A flow chart was developed to assist managers in the process of daily NC tracking which is located here: Defect reduction flow. The flow chart also includes links to the WAT. Predictive Reports In an attempt to aid first line managers with identifying potential quality defects for their volumes, the quality team has developed a predictive report that identifies which scheduled jobs have had 2"d issue parts or NCs written against them in the past. The managers can sort and print by their respective volumes or by airplane line number for the reports. Databases In order to aid in project identification and evaluation, the quality team has established simple databases that house information on NCs, PUs, CC4 hours relating to both NC and PU data, and the 2"d issue database. CC4 data can be found in an Access database in the rawest form in the following location: CC4 Labor For EMC.accdb The table TMASTER_2/2 houses the data. This data is from terradata, and is different from finance data due to the nature of charging issues. Regardless, it is highly positively correlated with the finance data, and provides greater ability to drive action than finance. It is best used with knowledge of filtering and pivot tables. This data gets updated through the use of a macro embedded within the database that is called "Update." The table has the following layout: 90 Field Data Type Explanation Line Number Number Line Numberthe SO worked SO[ Number Text SOI Number Worked SHOP ORDER DESCRI PTION Text Description of SOI What Type of SO it is Baseline =CC3, Everything is =CC4 (CO=Check Order, Customer=Customer Pickups, ENG= EQC's,ENGNC=Eng Coded NC's, ENG_PU=PU SOls caused by working Eng NC's, ME=SRR's, OPSNC=Operations Coded NC's, PU=Pickups, Removals, Text SOITYPE SMNC's=Supplier Coded NC's, SupPU's= PU's caused by working Supplier Issues, Travelled= Baseline Supplier work travelled) SOIBucket Text SUPERINTENDENT GENERAL TEAM Text Text Text Number Text Text Text Text Text Text Text Date/Tim SOITIME WORK CENTER NC _EPD# SOIDEFECTFOUNON DEFECT S-Factory G-Factory Team Factory SOI Closed Date What Type of SOI it is Baseline =CC3, Everything is =CC4 (ENG= All Eng Caused CC4, OPS=All Operations Caused CC4, Removals, Supplier=All Supplier Caused CC4) IE's View, Please use S-Factory for Positional Superintendents IE's View, Please use G-Factory for Positional Generals IE's View, Please use Team Factory for Positional and volume 1( the time it took to work the SOI's, In hours WorkCenter SOI was worked in EPD Number if rework was a NC or PU SOI that the NC or PU was written on Defect that caused the NC or PU Positional Superintendent the work was performed Positional General the work was performed Positional Team the work was performed Date the SOI was worked Pick up and NC generation data can be found in the following location: NC And Pickup Data This data includes CORRS data associated with the NCs generated. 91 Appendix B - Management Dashboard Example (Data removed) Everett Modification Center Last Upded: 719MI3 " Quality Performance Dashboard - EMC Pervasive Rollup Metrics Breakdown Total nurmber of NCs Initiated Numver of Ops NCs Initiated Weekly EMC % Change ms By ype EMC Penimlue Ntfienerullom Qty of PUs Initiated Qty of ERs Initiated Qty of ERs Cancelled SOls completed SOls FPQ % EPD Routback % CC4 Hours CC3 Hours Hour Rollup CC4/CC3 Ratio efect NC Top 5 NCs By CC4 Hours R Su Bathtubs Sinks Dishes Floors Toilets ToP Cunt 100 Sum 1su 10 100 100 100 10 10 10 100 10 5 PUsBy CC4 Defect A 10 10 10 10 10 Hours Sc.ratch bent Mark Smear Count Av3 Hrs too 10 10 10D 10 10 100 10 10 1OD 10 10 100 10 10 Tomt 500 Dirt V V V V 5 Hours T 500 P1) EC Periwive 2nd Issue Parts ill o500 1000 Grand Ttotal Wornkehope: S V VV V Jul 92 I A -C I I I