Improvement of Final Demand Data Flow at Polestar Refining Final Demand Management in the Geometry Assur- ance Process Master’s thesis in Product Development GIZEM EKEN MEDINA HAMZA DEPARTMENT OF INDUSTRIAL AND MATERIALS SCIENCE CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2024 www.chalmers.se www.chalmers.se Master’s thesis 2024 Improvement of Final Demand Data Flow at Polestar Refining Final Demand Management in the Geometry Assurance Process GIZEM EKEN MEDINA HAMZA Department of Industrial and Materials Science Division of Product Development Chalmers University of Technology Gothenburg, Sweden 2024 Improvement of Final Demand Data Flow at Polestar Refining Final Demand Management in the Geometry Assurance Process GIZEM EKEN MEDINA HAMZA © GIZEM EKEN, 2024. © MEDINA HAMZA, 2024. Supervisor: Andreas Stenlund, PE Geometry Examiner: Kristina Wärmefjord, Department of Industrial and Materials Science Master’s Thesis 2024 Department of Industrial and Materials Science Division of Product Development Chalmers University of Technology SE-412 96 Gothenburg Telephone +46 31 772 1000 Cover: A sketch of a technical drawing of a Polestar vehicle [1]. Typeset in LATEX, template by Kyriaki Antoniadou-Plytaria Printed by Chalmers Reproservice Gothenburg, Sweden 2024 iv Abstract Geometric variations in components, despite modern manufacturing precision, can detract significantly from a vehicle’s functionality and perceived quality. Geome- try assurance is crucial for managing these variations by ensuring robustness and high product quality through assigning favorable tolerances. Moreover, efficient data management plays a significant role in geometry assurance, facilitating the necessary exchange of information across various stages of the product realization loop. This thesis focuses on optimizing the data flow and management of final demand data within the geometry assurance process at Polestar, a Swedish automotive company specializing in the production of electric vehicles. This master’s thesis aimed at mapping the current geometry assurance process at Polestar, and by identifying and mitigating inefficiencies, propose improvements for enhancing data quality and process performance. The scope of the study is limited to the geometry assurance process, focusing specifically on the transformation and management of final demand data throughout the distinct phases of the process. Through a literature study, interviews, a pilot study and the mapping of process activities, the current process was studied in detail. The findings reveal significant inefficiencies in the current data flow and process approach. Data flow challenges were predominantly associated with compromised data integrity, as a result of the manual steps involved in data transformation and the propagation of errors due to lack of integration between software tools. Further, the mapping of the current process outlined an approach that often lacks clarity and a solid ground. The proposed enhancements include the implementation of an API that enables automated data transfer, formatting and storing, eliminating errors associated with manual data exchange across departments and stages. It facilitates communication across three different software programs utilized in different phases of the geom- etry assurance process, ensuring that the demand data is linked. Through this, data flow efficiency is significantly enhanced and data quality is improved. Further, suggestions intended to optimize the overall performance of the process include im- plementing standardized work routines and optimizing time management practices. The suggested improvements not only address immediate inefficiencies but also lay a foundation for continuous improvement in data management practices within the company. Keywords: Geometry Assurance, Data Flow, Process Optimization, Product Real- ization Loop, Product Development, Final Demands, Robust Design, Tolerance. v Acknowledgements The completion of this master’s thesis owes its gratitude to several individuals and organizations. We extend our sincere appreciation to Polestar and PE Geometry for their belief in our capabilities and for entrusting us with this master’s thesis. We are especially grateful to our supervisors, Andreas Stenlund at PE Geometry and Charlie Berner at Polestar in Gothenburg for their invaluable guidance and support throughout the entire process. A special acknowledgment goes to Niklas Nylén from PE Geometry for generously offering his time and expertise to provide guidance when needed. We would also like to express our gratitude to Kai Ni and Qingdong Xu at Polestar in Chongqing for their assistance with measurements. Furthermore, we wish to convey our heartfelt thanks to our examiner, Kristina Wärmefjord, whose insightful feedback greatly contributed to the refinement of this thesis. Lastly, we are deeply appreciative of the unwavering support from our fami- lies and friends throughout this journey. Their encouragement has been instrumental in our success. Gizem Eken and Medina Hamza, Gothenburg, June 2024 vii List of Acronyms Below is the list of acronyms that have been used throughout this thesis listed in alphabetical order: 2D Two Dimensional 3D Three Dimensional CAD Computer-Aided Design CAT Computer-Aided Tolerancing CM4D Computer Measurement Machine Management Mechanism Data CMM Coordinate Measuring Machines DMIS Dimensional Measuring Interface Standard EX Exterior Demand FDA Final Demand Application FDJ Final Data Judgement GA Geometry Assurance Geo Geometry MBD Model-Based Definition ME Manufacturing Engineering MP Measurement Point PKI Process Control Instruction PLM Product Lifecycle Management PQ Perceived Quality PS Project Start QIF Quality Information Framework RCA Root Cause Analysis R&D Research & Development RD&T Robust Design & Tolerancing Software SIPOC Supplier, Inputs, Process, Outputs & Customers STEP Standard for the Exchange of Product model data VP Verification Production ix Contents List of Acronyms ix List of Figures xv List of Tables xvii 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.1 Company Background . . . . . . . . . . . . . . . . . . . . . . 2 1.1.1.1 Polestar . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.1.1.2 PE Geometry . . . . . . . . . . . . . . . . . . . . . . 3 1.2 Aims and Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3 Delimitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.4 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 Literature Study 5 2.1 Geometry Assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.1.1 Robust Design and Geometric Variation . . . . . . . . . . . . 5 2.1.2 Product Realization Loop . . . . . . . . . . . . . . . . . . . . 6 2.1.2.1 Concept Phase . . . . . . . . . . . . . . . . . . . . . 6 2.1.2.2 Verification Phase (pre-production) . . . . . . . . . . 8 2.1.2.3 Production Phase . . . . . . . . . . . . . . . . . . . . 9 2.1.3 RD&T . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2 Aesthetical Geometrical Requirements . . . . . . . . . . . . . . . . . 10 2.3 Data Management in the Geometry Assurance Processes . . . . . . . 11 2.3.1 Application Programming Interface (API) . . . . . . . . . . . 12 2.3.2 Industry Standards for Product Data Exchange . . . . . . . . 12 2.4 Benchmark of Data Management in Geometry Assurance Processes at Competing Companies . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.4.1 Volvo Cars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.4.2 Zeekr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 3 Methodology 15 3.1 Methodology Approach . . . . . . . . . . . . . . . . . . . . . . . . . . 15 3.2 Preparation Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.2.1 Literature Study . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.2.2 Preparatory Interviews . . . . . . . . . . . . . . . . . . . . . . 16 xi Contents 3.2.3 Interviews with Engineers from Competing Companies . . . . 17 3.3 Realization Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 3.3.1 Study of Current Process . . . . . . . . . . . . . . . . . . . . . 18 3.3.1.1 Mapping of Current Process . . . . . . . . . . . . . . 18 3.3.1.2 Process and Data Flow Diagram . . . . . . . . . . . 19 3.3.1.3 SIPOC Diagram . . . . . . . . . . . . . . . . . . . . 19 3.3.1.4 Pilot Study . . . . . . . . . . . . . . . . . . . . . . . 19 3.4 Analysis Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.4.1 Analysis of Current Process . . . . . . . . . . . . . . . . . . . 20 3.4.1.1 Content Analysis of Interviews . . . . . . . . . . . . 20 3.4.1.2 Pilot Study Analysis . . . . . . . . . . . . . . . . . . 20 3.4.2 Brainstorming of Solutions for the Data Flow . . . . . . . . . 21 3.4.2.1 Placement of Solutions in the Data Flow . . . . . . . 21 3.4.2.2 The 5 Hows Method . . . . . . . . . . . . . . . . . . 21 3.4.3 Suggestions for the Process Flow . . . . . . . . . . . . . . . . 22 4 Results 23 4.1 Current Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.1.1 Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.1.1.1 Preparatory Interviews . . . . . . . . . . . . . . . . . 23 4.1.1.2 In-Depth Interviews . . . . . . . . . . . . . . . . . . 24 4.1.2 Mapping of Current Process . . . . . . . . . . . . . . . . . . . 25 4.1.2.1 Concept Phase . . . . . . . . . . . . . . . . . . . . . 25 4.1.2.2 Verification and Production Phase . . . . . . . . . . 26 4.1.3 Significant Gates in the Current Process . . . . . . . . . . . . 28 4.1.4 Graphical Representation of the Current Process . . . . . . . 28 4.1.4.1 Process and Data Flow Diagram . . . . . . . . . . . 28 4.1.4.2 SIPOC Diagram . . . . . . . . . . . . . . . . . . . . 34 4.1.5 Pilot Process . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 4.1.5.1 Concept Phase . . . . . . . . . . . . . . . . . . . . . 35 4.1.5.2 Verification and Production Phase . . . . . . . . . . 40 4.2 Facets of the Current Process . . . . . . . . . . . . . . . . . . . . . . 46 4.2.1 Interviews within Polestar . . . . . . . . . . . . . . . . . . . . 46 4.2.2 Findings from the Pilot Study . . . . . . . . . . . . . . . . . . 49 4.3 Facets of Processes at Competing Companies . . . . . . . . . . . . . . 50 4.4 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.4.1 Identified Improvement Areas . . . . . . . . . . . . . . . . . . 53 4.4.1.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.4.1.2 Process . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.4.2 Formulated Solutions . . . . . . . . . . . . . . . . . . . . . . . 54 4.4.2.1 API . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.4.2.2 Possibility for Future Advancement . . . . . . . . . . 57 4.4.2.3 Process Optimization Suggestions . . . . . . . . . . . 59 4.5 Proposed Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 4.5.1 Assessment of Proposed Process . . . . . . . . . . . . . . . . . 67 5 Discussion 69 xii Contents 5.1 Advantages of Proposed Process . . . . . . . . . . . . . . . . . . . . . 69 5.1.1 API . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.1.2 Process Suggestions . . . . . . . . . . . . . . . . . . . . . . . . 70 5.2 Constraints of Proposed Process . . . . . . . . . . . . . . . . . . . . . 71 5.2.1 Technical Constraints . . . . . . . . . . . . . . . . . . . . . . . 71 5.2.2 Human Constraints . . . . . . . . . . . . . . . . . . . . . . . . 72 5.3 Future Development . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 5.4 Methodology of the Project . . . . . . . . . . . . . . . . . . . . . . . 74 5.4.1 Sources of Error . . . . . . . . . . . . . . . . . . . . . . . . . . 74 5.5 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5.5.1 RQ1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 5.5.2 RQ2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 5.5.3 RQ3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 6 Conclusion & Recommendations 79 References 81 A Appendix A. Interview Guide for the Preparatory Interviews I B Appendix B. Interview Guide for the Interviews with Engineers from Competing Companies III C Appendix C. Interview Guide for the In-depth Interview V D Appendix D. The Transcription of the Preparatory Interviews VII E Appendix E. The Transcription of the In-depth Interview XI F Appendix F. The Process and Data Flow Diagram XV G Appendix G. The Transcription of the Interviews with Engineers from Competing Companies XVII xiii Contents xiv List of Figures 1.1 The Polestar 1 coupé [10]. . . . . . . . . . . . . . . . . . . . . . . . . 2 2.1 P-diagram [14]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2 Product realization loop. Figure adapted from [13]. . . . . . . . . . . 6 2.3 Split-lines in a Polestar car. Adapted from [22]. . . . . . . . . . . . . 11 3.1 Methodology approach. . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.2 The 5 Hows. Adapted from [38]. . . . . . . . . . . . . . . . . . . . . 22 4.1 Initial mapping of process flow. . . . . . . . . . . . . . . . . . . . . . 24 4.2 Updated flow chart of the final demands management in the GA process. 25 4.3 Description of symbols and line styles used in the flow diagram. . . . 29 4.4 Initial segment of the flow diagram. . . . . . . . . . . . . . . . . . . . 30 4.5 The Verification Phase in the flow diagram. . . . . . . . . . . . . . . 32 4.6 The Production Phase in the flow diagram. . . . . . . . . . . . . . . . 34 4.7 SIPOC diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 4.8 Visual representation of the placement of demands EX076 and EX077 with imaginary values. . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.9 The gap and flush demands as measures in RD&T (EX076 to the left and EX077 to the right). . . . . . . . . . . . . . . . . . . . . . . . . . 37 4.10 Variation analysis results from RD&T for EX076 (Gap demand). . . . 37 4.11 Variation analysis results from RD&T for EX077 (Gap demand). . . . 38 4.12 Variation analysis results from RD&T for EX076 (Flush demand). . . 38 4.13 Variation analysis results from RD&T for EX077 (Flush demand). . . 39 4.14 Results for demand EX076. . . . . . . . . . . . . . . . . . . . . . . . 39 4.15 Results for demand EX077. . . . . . . . . . . . . . . . . . . . . . . . 40 4.16 MP section for demand EX076. . . . . . . . . . . . . . . . . . . . . . 41 4.17 MP drawing for the rear fender. . . . . . . . . . . . . . . . . . . . . . 41 4.18 MP drawing for the rear body lamp. . . . . . . . . . . . . . . . . . . 42 4.19 PKI drawing with gauge for EX076 gap demand. . . . . . . . . . . . 42 4.20 PKI drawing with gauge for EX076 flush demand. . . . . . . . . . . . 43 4.21 Measurement page in FDA. . . . . . . . . . . . . . . . . . . . . . . . 43 4.22 Results of measurements on the nominal buck. . . . . . . . . . . . . . 44 4.23 Filtering of the measurement results. . . . . . . . . . . . . . . . . . . 45 4.24 The results in CM4D where the confidential information is covered. . 46 4.25 Identified Improvement Areas in the Data Flow. . . . . . . . . . . . . 53 4.26 Semi-automated version of module ABCD. . . . . . . . . . . . . . . . 56 xv List of Figures 4.27 Semi-automated version of module E. . . . . . . . . . . . . . . . . . . 57 4.28 Fully automated version of module ABCD. . . . . . . . . . . . . . . . 58 4.29 Fully automated version of module E. . . . . . . . . . . . . . . . . . . 59 4.30 Data flow diagram of the proposed process. . . . . . . . . . . . . . . . 62 4.31 Mock-up of start-page of the user interface. . . . . . . . . . . . . . . . 63 4.32 Choice of session in user interface. . . . . . . . . . . . . . . . . . . . . 63 4.33 Required fields to prompt the simulation module. . . . . . . . . . . . 64 4.34 Additional required fields to prompt the simulation module. . . . . . 65 4.35 Concept Phase of the proposed process. . . . . . . . . . . . . . . . . . 66 4.36 Verification and Production Phase of the proposed process. . . . . . . 67 xvi List of Tables 2.1 Methods used during the Concept Phase. Information taken from [13] and [16]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2 Methods used during the Verification Phase. Information taken from [13] and [16]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.3 Methods used during the Production Phase. Information taken from [13] and [16]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 4.1 Demands EX076 and EX077. . . . . . . . . . . . . . . . . . . . . . . 36 4.2 Key takeaways from analysis of interview answers. . . . . . . . . . . . 49 4.3 Identified bottlenecks from the Pilot Study. . . . . . . . . . . . . . . . 50 4.4 Key takeaways from analysis of interview answers with engineers from competing companies. . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.5 Assessment of the proposed process’ feasibility [40]. . . . . . . . . . . 68 xvii List of Tables xviii 1 Introduction This chapter provides an overview of the background and factors initiating this master’s thesis, including a presentation of the company background. The aims and delimitations of the project work are discussed, followed by a description of the research questions intended to be addressed through the work. 1.1 Background Despite advancements in precision in modern manufacturing processes, geometric variations in physical artifacts occur and are caused by diverse process-related fac- tors, impacting functionality and perceived quality in the final product. A low perceived quality impairs the customer’s overall experience and view of the product, which directly correlates to the presence of deviation in the geometry [2]. Companies must address and manage these geometric variations throughout the entire product life cycle by specifying geometrical requirements for the components. Particularly gap and flush requirements are of great importance for the perceived quality of vehicles. Geometry assurance includes a set of activities aimed at managing and reducing the impact of geometric part variations on product quality [3]. It is a cru- cial part of the product realization process to ensure a design that is insensitive to variation and of high quality, fulfilling final demand requirements. The electric car manufacturing company Polestar needs to improve the final demand data flow in their geometry assurance process to improve the overall performance of the process, ensuring high standards of quality. Currently, geometry assurance pro- cesses for Polestar are carried out in collaboration with PE Geometry, a consultant firm that specializes in geometry assurance. Diverse activities are conducted during the development of a new Polestar car, including geometry assurance. Throughout the geometry assurance process, the engineers at Polestar observe the gradual refine- ment of the final demand data related to the car’s geometry over time, transforming it into other data types throughout the process. Improvement of this demand data management process is required by identifying bottlenecks in the workflow along with integrating potential improvements in the current approach [4]. Thus, this master’s thesis was proposed by Polestar together with PE Geometry and will be conducted by two mechanical engineering students from Chalmers University of Technology, during the spring semester of 2024. 1 1. Introduction 1.1.1 Company Background This section provides background information on the companies involved in the master’s thesis, intending to provide an understanding of their history and the work they undertake. 1.1.1.1 Polestar The company of Polestar is a Swedish automotive brand specializing in the produc- tion of high-performance electric vehicles, operating in nearly 30 markets worldwide. The company aims to enrich its customers’ lives with cars that integrate Scandina- vian minimalist design, high technology, and sustainability. In addition to meeting customer needs, the company aims to achieve “climate-neutral” production by 2030 [5]. This entails reaching zero net carbon footprint in car production, thereby min- imizing environmental impact. The original team of Polestar was initially a motor-sport team founded in 1996 by Jan Nilsson, known as Flash Engineering. It later rebranded to Polestar Racing when Christian Dahl acquired the team in 2005 [6], establishing an effective collab- oration with the Swedish car manufacturer Volvo Cars. Their partnership reached a new level in 2015 when it was sold to Volvo Cars. During this time, they agreed to produce cars branded with the Volvo logo but featuring Polestar modifications to enhance performance in terms of speed and power. This model designation became synonymous with their higher-performing cars for the Volvo Car Group [7]. By the end of 2017, the Polestar brand underwent revitalization and was sepa- rated from the Volvo brand, although it remained part of the corporation alongside its owner, Geely Holding. The company received significant support for initiation through a substantial investment, enabling the establishment of a production site in China [8], with its headquarters in Gothenburg, Sweden. Concurrently, they launched their first product, the Polestar 1, an exclusive 600-horsepower electric coupé [9], as seen in Figure 1.1. Figure 1.1: The Polestar 1 coupé [10]. Since then, they have been actively developing their brand to expand into several global markets by introducing and selling more models, with the Polestar 4 being 2 1. Introduction their latest launch, an SUV coupé with around 540 horsepower [11]. 1.1.1.2 PE Geometry PE Geometry is a Swedish acknowledged specialist company specializing in the field of geometry assurance, primarily within the automotive industry. They offer their expertise to the industry, such as Volvo Group and Scania Group, and universi- ties through consultancy, research and education. The company is partnered with NetGroup, a trusted consultant company for major companies in the automotive industry [12]. The company was established in 1999 in Gothenburg by Peter Edholm to further expand in 2013 to the American market, by establishing its subsidiary in North Carolina. Presently, PE Geometry has offices in Sweden, the US and India [12]. 1.2 Aims and Objectives The primary aim of this master’s thesis is to improve the data flow and demand data management in the current geometry assurance process at Polestar, by discerning bottlenecks and proposing strategic improvements to enhance its overall efficiency. Thus, the objectives of the thesis are to study the current process and identify areas of improvement. Additionally, the project objectives include a refinement of the process by integrating enhancements into the existing approach, as well as considering future practice. 1.3 Delimitations The scope of this master’s thesis is constrained by several delimitations to ensure a focused study. Firstly, a notable constraint is the time limit of 20 weeks imposed on the research. Secondly, the study will consider a limited segment of the entire geometry assurance process. Furthermore, the primary focus of this thesis remains on the identification of bottlenecks relating to final demand management within cur- rent geometry assurance process. The demand data treated in the thesis includes primarily gap and flush requirements. While the study aims to provide comprehen- sive insights within the defined time frame, certain aspects beyond these parameters may not be fully explored. The outcome of the project will focus on optimizations and improvements of the current process. Moreover, the master’s thesis will be undertaken by two students holding bachelor’s degrees in mechanical engineering. This implies that the scope of expertise may be limited to this field. However, external experts specializing in the relevant areas will be incorporated into the process. Although, there is a deficiency in departments at Polestar, as they lack a unit specializing in perceived quality. This could potentially impose a constraint on the project regarding retrieval of information and knowledge. 3 1. Introduction 1.4 Research Questions In the light of the aims and objectives of the project, three research questions that are to be answered through the thesis work were specified. These questions are formulated as follows: 1. What is the approach to current practices and documentation methods, includ- ing specification and management of demand data throughout the geometry assurance process at Polestar? 2. What bottlenecks are present in the demand data flow of the current geometry assurance process at Polestar? 3. How can the current practices employed in managing demand data in the geometry assurance process be improved, and how can these enhancements be integrated into the existing and future framework at Polestar? 4 2 Literature Study In the following chapter, the literature study providing the necessary theory for conducting the thesis work is presented. The theoretical framework covers the key subjects of the project, which are geometry assurance, GA, and its function in product design, geometrical requirements and perceived quality, as well as data management in GA processes. 2.1 Geometry Assurance This section introduces the concept of geometric variation in products, and how GA can be utilized to achieve robust design. It further provides insight into how GA activities are incorporated into the product realization loop. 2.1.1 Robust Design and Geometric Variation The definition of robust design is a design that is not affected by the presence of variation. The theories regarding geometric variation and robustness in design were originally initiated by Taguchi [13]. Factors that influence variation in a system are categorized as control factors and noise factors. The noise factors are difficult to control, and are external factors that amplify variation, whereas control factors are inputs that regulate variation. The concept of robustness in a system can be visualized using a P-diagram, as shown in Figure 2.1. It demonstrates how differ- ent factors affect a product or process, resulting in a response that measures the performance of the system [14]. Figure 2.1: P-diagram [14]. 5 2. Literature Study The presence of geometric variation and deviation from the nominal geometry in parts can be amplified as a result of additional deviations appearing from physical factors like wear, and thermal expansion. Additionally, variation in critical part manufacturing process parameters or assembly and joining processes can also cause part variation. Consequently, the quality of the product tends to decrease further during use. Hence, there is a demand for companies to control these geometric vari- ations throughout the entire product life cycle [3]. Tolerances serve as a means to specify the allowed variation. Variations on part- and product-level that are not within the given tolerances can impair the final product and fail to fulfill aesthetic and functional requirements [15]. To minimize the pres- ence and effect of variations in a product, a set of activities is conducted to control the geometry. This process is called GA and can be integrated into all phases of the product realization process. The main rule in the field of GA is to avoid introducing additional sources of variation throughout the realization process of the product [13]. 2.1.2 Product Realization Loop The steps of GA can be divided into different phases constituting the product real- ization loop, aiming to minimize geometrical variations in the designated product. It consists of three phases denoted as concept, verification and production, as vi- sualized in Figure 2.2. Each phase comprises several activities related to the GA process, intending to control the variations in the product’s geometry [16]. Figure 2.2: Product realization loop. Figure adapted from [13]. 2.1.2.1 Concept Phase In the first phase of the product realization loop, known as the Concept Phase, the product and production concepts are detailed. The developed concepts undergo 6 2. Literature Study evaluation to optimize them in terms of manufacturing variation, aiming to control their robustness through the placement of the locators [16]. This implies that no physical product is developed, instead, simulations and data-based testing are con- ducted to facilitate decisions regarding component separation and manufacturing processes. The tasks related to this phase are summarized as follows [13]: 1. Defining the split-lines to gain an understanding of which components will divide the product. 2. Defining the top-level requirements and their tolerances on the product. 3. Defining the positioning of the locators on the components to increase their robustness, enhancing the product’s overall robustness. 4. Allocating tolerances for each part of the product. The listed tasks are realized by a number of analysis methods, which are briefly presented in Table 2.1. 7 2. Literature Study Table 2.1: Methods used during the Concept Phase. Information taken from [13] and [16]. 2.1.2.2 Verification Phase (pre-production) The second phase of the product realization loop is the Verification Phase, encom- passing activities conducted in pre-production [16]. The objective of this phase is to test and verify the reduced number of concepts developed in the preceding stage, ensuring their practical viability at various points to gather as much information as possible [17]. This phase involves analyses of both the product and the production system, aiming for a seamless transition into full-scale production in the next phase. 8 2. Literature Study Consequently, all inspections, analyses, and tests are conducted to ensure a smooth start to production, including fixing locators and supporting the preparation pro- cess [13]. The methods used in this phase include inspection preparation and virtual trimming, further explained in Table 2.2. Table 2.2: Methods used during the Verification Phase. Information taken from [13] and [16]. Additionally, the variation simulation model can potentially be used as a digital twin, which is a virtual representation of a physical system that allows for real-time monitoring and analysis. By inputting part inspection data into the simulation model, process adjustments can be recommended to optimize geometric quality at the assembly level [18] 2.1.2.3 Production Phase The third and final phase in the product realization loop is the Production Phase. As the name implies, this phase involves the transition of the product and production concept into full-scale production, utilizing the variation simulation model to manage the production process and identify any occurring problems. Two methods employed during this phase are the Root Cause Analysis, RCA, and Six Sigma methodology [13][16], elaborated further in Figure 2.3. 9 2. Literature Study Table 2.3: Methods used during the Production Phase. Information taken from [13] and [16]. 2.1.3 RD&T A tool that is used in GA is RD&T, Robust Design & Tolerancing, which is a soft- ware for dimensional control, tolerance analysis and statistical simulation variation [19]. This refers to CAT, which stands for computer-aided tolerancing [2]. The tool- box available in the software enables the simulation and visualization of the impact of deviations. This makes it a valuable tool in all phases of the GA process, in advance of the production of physical prototypes. By implementing RD&T in a GA process, robustness and tolerances can be optimized, as well as assembly and joining sequences [19]. Simulating and analyzing the variation is a crucial step in the early development phases of the product realization loop [13]. RD&T allows for statistical analysis and simulation of variation using Monte Carlo simulation. Another common type of analysis that RD&T is used for is contribution analysis, which identifies the contribution of each tolerance that contributes to variation [19]. 2.2 Aesthetical Geometrical Requirements Geometric variation in products does not only impair the product quality and perfor- mance but also the customer’s experience and perception of the product. Perceived quality, PQ, is defined as the customer’s visual observation of the quality of the product, which for vehicles primarily relies on the relation between visible compo- nents and surfaces [2]. The surfaces visible to the users are called Class A surfaces, whereas the surfaces not visible are termed Class B surfaces [20]. The visual di- 10 2. Literature Study visions between components are called split-lines, which are marked in Figure 2.3. Generally, contact between components in split-lines is not desired. The space that is present in the split-lines are directly affected by the presence of geometric varia- tion in the components. To regulate the space, aesthetical geometrical requirements are specified, indicating allowed maximum and minimum values for the gap and flush-size. Gap refers to the clearance value between two parts, while flush refers to the height difference between the upper surfaces of these parts [21]. If these require- ments, also called final demand requirements, are not met, remarkably wide gaps, large flush, asymmetry and non-parallelism may occur in the split-lines, impairing the PQ of the vehicle [2]. Figure 2.3: Split-lines in a Polestar car. Adapted from [22]. 2.3 Data Management in the Geometry Assur- ance Processes As digitalized production systems and their links across value chains evolve, enhanc- ing efficiency in manufacturing becomes crucial. Additionally, adaptability to con- stantly changing customer and market demands should be implemented. Thus, data exchange between people, machines, and companies within complex information sys- tems is necessary. This evolution inducing digitalization in product development and manufacturing also generates a change within GA processes and the transfers of de- mand data. In a GA process, strong links between departments within a company, such as product design, engineering and manufacturing, are crucial. Thus, effective data and information management systems are essential [3]. GA relies on the exchange of data, especially demand data, between different actors in the process. Enhanced capabilities of data storage and transfer facilitate the utilization of demand data in GA. This means that the significance of challenges 11 2. Literature Study related to data quality, data validity, and data security becomes more pronounced. It is important to ensure that the correct data is used to secure the data quality and validity. Thus, an efficient and feasible product data management system that handles all models and data throughout all process activities is of great importance [3]. 2.3.1 Application Programming Interface (API) An application programming interface (API) serves as an intricate coding mecha- nism facilitating communication between different applications. An application is a form of software designed to enable users to perform specific tasks. APIs provide programs, software, and application developers with precise control over access to application interfaces, without the need for complete shutdown, enabling applica- tions to exchange information in a regulated manner [23]. 2.3.2 Industry Standards for Product Data Exchange As the digitalization trend emerges within manufacturing, a set of industry standards for data exchange between programs and equipment can be utilized to facilitate ef- ficiency. Some of these are the following: 1. DMIS, Dimensional Measuring Interface Standard, is a standardized language used for communication of data between computer systems and inspection equipment. It was developed with the aim to transfer files from CAD sys- tems without altering the data through translation or restructuring. DMIS provides a common interface allowing programs to be run on different CMMs, coordinate measuring machines, which ensures significant flexibility for man- ufacturers [24]. 2. STEP AP242, Standard for the Exchange of Product model data, is an inter- national standard that defines how to represent 3D mechanical product data in a neutral format, facilitating exchange between diverse CAD software plat- forms. It encompasses various aspects of product data, including geometry, assemblies, annotations, and metadata [25]. 3. QIF 3.0, Quality Information Framework, focuses on quality-related informa- tion associated with manufacturing processes and inspection activities. It provides a standardized approach to represent dimensional and geometric tol- erances, inspection plans, measurement results, and other quality-related data. For example, within QIF files, requirement fulfillment statuses for individual measurements can be stored and retrieved, directly linked to the respective features [25]. 4. MBD, Model-Based Definition, is a method that employs a shared 3D model of the product or production system throughout its development and life cy- cle. Through the implementation of MBD, manual labor and reliance on hu- 12 2. Literature Study mans can be reduced, enhancing efficiency while minimizing the risk of errors. This implies numerous advantages regarding both cost-effectiveness and oper- ational efficiency in manufacturing processes. For instance, the definition can be utilized to show that 2D drawings can be exchanged with 3D models or to automate data handling. [25]. 2.4 Benchmark of Data Management in Geome- try Assurance Processes at Competing Com- panies This section presents the data management methodologies in GA processes at Volvo Cars and Zeekr, formerly known as CEVT. By comparing and presenting these com- panies’ approaches, insights into strategies, tools and challenges can be identified. 2.4.1 Volvo Cars The GA process at Volvo Cars runs through four phases denoted as Concept Phase, Engineering Phase, followed by Industrialization Phase and finally Production Phase. There are several milestones and gateways present in the process, including design releases termed Digital Surface Model releases, or DMS-releases. These gateways are named DMS0, 2, 3, 4 and 5, which are iterations leading up to FDJ. FDJ stands for Final Data Judgement, which is a gateway indicating cross-functional confirmation that the final demands are fulfilled [26]. The process of defining FDR, final demand requirements, is initiated at the PQ department. RD&T is used to create PQ documents and SharePoint, which is a software part of the Microsoft package used to simplify collaboration and sharing of data [27], is utilized for tracking requirement data. The site in SharePoint where the requirement data is gathered is called the Geometry Scorecard, which supports the formation of the PQ documents. Additionally, another site called the Masterlist is available in SharePoint for storing old requirement data [26]. The reason behind stor- ing old data lies in the ongoing discussions that alter the requirements throughout the process, resulting in deviations from the original specifications. These changes are documented either in RD&T or in SharePoint [20]. RD&T is utilized for as- signing tolerances, calculations and managing requirements. Further, measurement preparation is carried out [26]. Volvo Cars has implemented a system called FLOW, used to document any issues identified and highlighted during summaries. The process involves creating, naming, and assigning tasks for PQ sections, with standardized templates and priority scales for the organization. It includes assigning tasks to responsible engineers, tracking progress with workflow statuses, and notifying interested parties. Problem Reports are created for issues, linked to tasks, and communication is facilitated through com- ments. Finally, during running production, GA is carried out in order to ensure that the FDR is fulfilled throughout the entire life cycle of the vehicle [26]. In this phase, 13 2. Literature Study the software CM4D is used, which is a dimensional analysis tool for manufacturing processes [20]. 2.4.2 Zeekr The GA process at Zeekr is initiated through the specification of PQ requirements. Benchmark of competing companies’ products is carried out to set a standard for the level of PQ. RD&T is utilized for GA activities. Further, PLM systems TeamCenter, and SystemWeaver, softwares for managing products or systems throughout their lifecycles, are utilized for managing part models and requirement data. Both Sys- temWeaver [28] and TeamCenter [29] are software used to manage the specification of software components and systems on a shared platform. A shared PowerPoint file is used by the PQ and GA team to track activity and changes [30]. Iteration and looping is a crucial part of the process at Zeekr. GA engineers are provided with feedback and perform calculations on the construction. Regular check- ins are arranged weekly with different teams [30]. 14 3 Methodology This chapter provides a description of the designated work process and the methods utilized, along with the motivation behind the chosen methodology. 3.1 Methodology Approach To improve the data flow, with a focus on final demands, within the GA process at Polestar, it is crucial to have understanding and support. These are derived from design research conducted using a specific methodology. To support the thesis work, the methodology will be based on the design research presented in [31] with adapta- tions, aiming to improve the design process, which aligns with the objective of this master’s thesis. Firstly, the design research methodology, DRM, from [31] initiates with a research clarification. This step aims to extract information and data from the literature study findings, aiding in addressing the research questions of the master’s thesis. Additionally, it provides clarity on the current situation of the organization, sup- porting the subsequent steps. Next, the second step involves initiating the first descriptive study. Despite col- lecting sufficient literature and specifying objectives, the research questions remain unanswered. Therefore, descriptive study 1 is employed in the methodology to broaden understanding and gather more information through interviews. This em- pirical data gathering is essential to enhance comprehension before progressing to problem identification and improving the current process. The third step presented in the DRM from [31] is the prescriptive study. In this phase, it is crucial to utilize the outcomes from the preceding steps to assess the current process and subsequently enhance it. This involves identifying problems within the current process to enable improvement and optimization. The fourth and final step outlined in the DRM from [31] is the second descriptive study. This step is where the improvements are implemented. However, as the implementation falls outside the scope of the master’s thesis’ objective, this phase will only analyze the outcomes from the prescriptive study and thereafter provide recommendations for implementation to the company and its future work. 15 3. Methodology The DRM will therefore be adapted and divided into three key phases, termed as Preparation, Realization and Analysis, each containing a set of activities. The adapted DRM is depicted in Figure 3.1 below. Figure 3.1: Methodology approach. 3.2 Preparation Phase The Preparation Phase includes a literature study and interviews with individuals currently engaged in GA processes, both within and outside Polestar. This was conducted in order to gain a thorough comprehension of the subject prior to the Realization Phase, laying a meticulous foundation for the planned work. 3.2.1 Literature Study The literature study was conducted with the purpose of gathering relevant and useful material serving as a theoretical framework for the thesis work. In order to ensure the best outcome of the project, a selection of the essential and most significant information was found and studied. The selection of subjects included in the study was done with the aims and objectives of the thesis as a basis, assuring that the theoretical framework necessary for conducting the thesis work was covered. The gathering of data and information was performed using online based search engines, preeminently the Chalmers Library database and Google Scholar. Academic papers and literature were the primary sources from where the information was obtained, which were found using relevant keywords related to the sought theory. 3.2.2 Preparatory Interviews Preparatory interviews were conducted to gather qualitative data and gain an under- standing of the current GA process at Polestar. Qualitative data collection is defined 16 3. Methodology as a research method that emphasizes understanding and interpreting phenomena through in-depth analysis of subjective experiences, perspectives, and social pro- cesses [32]. A qualitative research approach was selected to obtain a more in-depth, contextual understanding of the participant perspectives, and generate descriptive research data. The preparatory interviews were planned as semi-structured interviews, leading to the creation of an interview guide with a predefined set of relevant questions related to the work and its challenges, as outlined in Appendix A. Semi-structured inter- views are a qualitative research approach used to gather data and information from interviewees in a controlled yet conversational manner. While there is a predefined set of questions, there is room for room for follow-up questions and flexibility within the interview [33]. The preparatory interviews were conducted after obtaining in- formation from supervisors and professionals within the company, who assisted in selecting the relevant interviewees. Subsequently, the selected interviewees were con- tacted to schedule a meeting time that would suit their agendas. The interviews followed a chronological order, with the interviewee representing the Concept Phase being interviewed first. The interviews took place both digitally and on-site, depending on the availability of the interviewee. Since the interviews were semi-structured with a predefined set of questions, they were asked to the inter- viewee in a chronological manner. However, room was left for additional questions that could enhance the answers provided during the interview. To facilitate the transcription of the answers, the interview was recorded. The transcribed answers obtained from the interviews were gathered and structured in an Excel sheet. 3.2.3 Interviews with Engineers from Competing Compa- nies Additional interviews were conducted with engineers from competing companies to gain qualitative data on their GA process and compare it to the one at Polestar. These interviews provided a solid foundation for obtaining information on specific topics through a qualitative research approach, particularly their perspectives on how PQ impacts the GA process. The interviews were meticulously planned and executed in a semi-structured format. This approach involved a set of predefined questions, as outlined in Appendix B, with flexibility for addressing any additional questions that might arise during the interview. Moreover, the interviews were recorded to facilitate the transcription process. 17 3. Methodology 3.3 Realization Phase The Realization Phase involves a comprehensive examination of the current process employed at Polestar, encompassing all stages from the initial phase to the final one, with the results obtained from the preparatory phase in mind. This was partly done by conducting a simplified version of the actual process, which means that a much smaller amount of demands and correlating components were examined instead of the whole vehicle. Additionally, in-depth interviews were conducted with relevant engineers to delve deeper into the workflow and enhance the understanding of issues encountered during the examination. The objective was to map the process on a more detailed level, along with identifying both positive and negative aspects of the current process at Polestar, thereby facilitating results for the analysis phase. 3.3.1 Study of Current Process The current process was assessed by studying real documents from ongoing projects, conducting a simplified version of the process and conducting in-depth interviews with involved engineers within the company. 3.3.1.1 Mapping of Current Process To be able to comprehensively map the current process along with its phases and in- coming activities, the preparatory interviews, see Section 3.2.2, were complemented by conducting in-depth interviews with engineers within the company. Utilizing the insights obtained from these interviews, a flow chart was created to depict the current process, emphasizing both the activities and data flow. In-depth Interviews In-depth interviews were conducted to deepen the initial understanding of the cur- rent GA process and to fill in the gaps in the qualitative data previously gathered from the preparatory interviews. Similarly to the previous interviews, the in-depth interviews were also approached using a qualitative research approach to gain de- tailed insights into the participants’ perspectives. The in-depth interviews were individually tailored for each participant and struc- tured in both a semi-structured and an unstructured format. The semi-structured interview resulted in an interview guide containing a predefined set of questions, whilst there was no preparation for the unstructured interview. An unstructured interview aims to keep the interview open-ended to facilitate flexibility and nuanced answers from the participant by asking descriptive questions that do not disrupt their narration [34]. The interview guide for the semi-structured interview was customized based on the participant’s role and involvement in the product realization loop, as detailed in Appendix C. While the interview adhered to the chronological order of the prede- fined questions, there was flexibility to ask additional questions as they arose during the interview. Since the participants resided abroad, the interviews were conducted 18 3. Methodology digitally and recorded to facilitate transcription. 3.3.1.2 Process and Data Flow Diagram To graphically represent the process and data flow of the current process, a flow diagram was created using the tool Mural, a versatile platform utilized for gener- ating various types of diagrams. The flow diagram was developed iteratively and continuously updated throughout the study of the current process as more informa- tion was obtained. Various symbols and line styles were thoughtfully selected and incorporated to enhance readability and simplify comprehension for the reader. 3.3.1.3 SIPOC Diagram With the entire process and data flow charted, a SIPOC diagram was created in order to obtain an overview perspective of the process. SIPOC stands for Supplier, Inputs, Process, Outputs and Customers [35]. It is a tool used to map these elements of a process, which provides a comprehensive understanding of the entire system, ensuring no critical element is disregarded. The purpose of using a SIPOC diagram is to provide a holistic overview of a process and its most crucial parts. This helps identify bottlenecks and inefficiencies, highlighting areas in need of improvement. Additionally, by examining suppliers, inputs, processes, outputs, and customers, teams can identify potential bottlenecks or inefficiencies within the current process. 3.3.1.4 Pilot Study The initiation of the simplified version began with the acquisition of necessary doc- uments, files and software access. This process involved submitting an access re- quest through Polestar’s internal service portal and requesting access for utilizing components. While awaiting approval, an examination and analysis of the exterior demands’ requirements documentation took place. Two demands, including gap and flush requirements, were selected for use in the simplified version. Upon selecting the demands and their corresponding components, additional ap- proval was sought to be able to use them in the thesis work. Once granted, the Pilot Study was conducted according to the current designated work process at Polestar. The work was carried out in three phases, structured as Concept Phase, Verification Phase and Production Phase, as presented in Section 2.1 of the literature study. 3.4 Analysis Phase The Analysis Phase entails an examination of the current process, commencing with a comprehensive analysis of interviews and the Pilot Study. Through content anal- ysis of interviews and Pilot Study findings, the aim is to gain valuable insights into the facets of the current process, identifying both strengths and areas for improve- ment. Additionally, the phase involves an intensive brainstorming session aimed at 19 3. Methodology generating innovative solutions for optimizing the data flow. This includes brain- storming sessions dedicated to determining the placement of solutions within the data flow and employing the 5 Hows Method to delve deeper into solution details. The objective is to devise practical and effective strategies for enhancing the ef- ficiency and effectiveness of the final demands management in the GA process at Polestar. Furthermore, suggestions for refining the process flow are formulated based on the insights garnered from the analysis. 3.4.1 Analysis of Current Process This section encompasses a thorough examination of the existing processes, con- ducted through two main approaches: content analysis of conducted interviews and analysis of the Pilot Study. These analyses serve as a foundation for further analyses, providing crucial insights that inform subsequent stages of the methodology. 3.4.1.1 Content Analysis of Interviews The compiled sheet containing the transcription of all conducted interviews was used as a basis for examining the answers by utilizing content analysis, which is a tech- nique used with the aim to identify patterns and common themes within qualitative interview data [36]. Prior to analyzing the answers, a guiding scheme defining signif- icant themes and implications was defined to facilitate the identification of patterns among the answers. According to a content analysis-based method, a pre-defined scheme can be used to categorize responses into themes or concepts, which allows for systematic analysis and comparison [36]. The aim of the content analysis was to identify facets of the current process among the respondents’ answers, in order to obtain an understanding of what improvements are necessary. The key takeaways from the answers were summarized in a table. 3.4.1.2 Pilot Study Analysis The Pilot Study was conducted to gain firsthand experience with the execution of the GA process and to identify and document ineffective bottlenecks as well as other facets of the process. Throughout the Pilot Study, identified bottlenecks were documented for each step, concurrently with the initial flowchart being updated to reflect activities. Mapping the process flow facilitated the identification of bot- tlenecks, making it clearer where improvements were needed within the flow. The Pilot Study was iterated to enable more in-depth identification of bottlenecks with each iteration. For analyzing the interview answers, as detailed in Section 3.4.1.1, content analysis was utilized to examine the identified bottlenecks. The purpose of employing content analysis was to discern improvement areas and detect patterns among the process activities [36]. The key findings were synthesized and presented in a table. 20 3. Methodology 3.4.2 Brainstorming of Solutions for the Data Flow To generate solutions for the problems concerning the data flow, a brainstorming technique was utilized. Brainstorming is a method used to produce creative ideas for identified problems, either in smaller or larger groups. There are various types of brainstorming techniques, in this case, both traditional and nominal brainstorming were utilized to generate solutions ideas [37]. The brainstorming session was initiated with a nominal approach, where group members individually generated ideas and recorded them on Post-it notes. Facets from both the interviews and Pilot Study were used as a basis. Subsequently, a traditional brainstorming session commenced, during which members collectively discussed and generated qualitative ideas using a whiteboard. After identifying the solutions, their placements within the data flow were determined. Subsequently, the 5 Hows method, adapted from the 5 Whys method, was employed to elaborate on these solutions, as presented in the following sections. 3.4.2.1 Placement of Solutions in the Data Flow The first step of the solution formulation was to locate particularly critical areas in the process. This was done separately for the data flow and process-related improve- ment areas, as these two aspects are addressed with different solutions. For process related issues, the findings from the analysis of the interviews and Pilot Study were directly used as a basis. For the data flow, the flowchart was studied. The data transfer in each step of the process was observed and evaluated in terms of efficiency and preservation of data quality. The steps in the data flow that were assessed to be the most critical were marked with red, indicating that these areas are in need of a solution. Additionally, the placement of the solutions that were to be determined later in the process was specified in the flowchart. 3.4.2.2 The 5 Hows Method Once the solution areas were identified, determining how the solutions would func- tion was crucial. To address this, the 5 Whys method was utilized, a technique commonly used to identify root causes of problems [38]. In this case, the method was adapted to suit the outcomes of the brainstorming session, where the 5 Whys were employed as 5 Hows to determine how the solutions would operate in detail, as depicted in Figure 3.2. 21 3. Methodology Figure 3.2: The 5 Hows. Adapted from [38]. The solution was approached following the direction of the arrow depicted in the figure above. Initially, the solutions were defined to later uncover answers to ques- tions regarding how it works and how it happens, ultimately obtaining a sufficiently detailed solution. This solution was then visualized in the form of black boxes with inputs and outputs, using the tool Mural. The black boxes aim to provide a visual- ization of how the solution works by showing where the input and output go in the data flow. 3.4.3 Suggestions for the Process Flow To address process-related problems, the analysis of interview responses and insights gained from the Pilot Study served as the foundation for a content analysis. The objective of this analysis was to pinpoint common problem areas and patterns ob- tained from these activities, compiled in an Excel spreadsheet for easier reference. The outcome of the content analysis clarified areas in need of improvement, indicat- ing where enhancements were required. The results were presented in a list. To tackle these improvement areas and devise suitable suggestions for solutions, the 5 Whys method was adapted into the 5 Hows method, as elaborated in Section 3.4.2.2. The purpose of employing this method was to provide detailed insights into how each improvement area could be effectively addressed. Additionally, the identified strengths and potential improvements obtained from the interviews were used to generate the solutions. The key findings were summarized in a table. 22 4 Results This chapter presents the results generated from the various activities conducted in the Preparation, Realization and Analysis phases. It demonstrates the current process, mapped using the results from the interviews, Pilot Study and from study- ing the process activities. Finally, the facets identified within the process and the solutions addressing them are presented, along with a proposed process with the solutions integrated. 4.1 Current Process In this section, the results obtained from studying the current process are presented as a detailed explanation of the workflow. This includes the results from the inter- views, the mapping of the activities and data flow in the process, and a description of the execution of the Pilot Study. Further, a graphical representation of the process is provided. 4.1.1 Interviews The findings derived from the interviews are outlined in this section, presenting the outcomes of the preparatory and in-depth interviews. 4.1.1.1 Preparatory Interviews The interviewed engineers within the company are part of the R&D Geo Research and Development - Vehicle Integration and Geometry team and the ME Geo, Man- ufacturing Engineering - Geometry team. Both of these departments are involved within the GA process to a great extent, and manage the final demand data in some way. The engineers taking part in the preparatory interviews were the following: • GA Engineer at the R&D department in Coventry, the United Kingdom. • Geo Program Manager at the ME department in Gothenburg, Sweden. • Geo Engineer at the production facility in Chongqing, China. • Geo Equipment Engineer at the production facility in Chongqing, China. 23 4. Results Each conducted interview lasted approximately one hour, and the transcribed an- swers for each participant can be found in Appendix D. The results obtained from the content analysis of the interviews imply that there is a pattern present among the answers of the respondents. Based on the descriptions of the respondents’ work processes in the respective phases, an initial, brief mapping of the process flow could be done, which can be seen in Figure 4.1. A presentation and analysis of the key findings from the interviews are presented in Section 4.2.1. Figure 4.1: Initial mapping of process flow. The larger light green colored areas represent different phases in accordance with the product realization loop, see Section 2.1.2. Each row represents the responsible team for that phase. For instance, the Verification Phase extends over two rows meaning that the responsibility is split between the R&D Geo team and ME Geo team. 4.1.1.2 In-Depth Interviews The following engineers from the company participated in the in-depth interviews: • Senior Geo Engineer for Vehicle Integration and Geometry team at the R&D department in Coventry, the United Kingdom. 24 4. Results • Geo Engineer at the ME Geo team in Chongqing, China. The in-depth interviews each lasted approximately one hour and the transcribed responses for the semi-structured interview can be found in Appendix E, whilst the responses from the unstructured interview were not recorded and instead used to perform the Pilot Study. The content analysis of the results revealed a detailed understanding of how each phase within the GA process operates. This insight facilitated further enhancement and updating of the process flow mapping, as depicted in Figure 4.2. Figure 4.2: Updated flow chart of the final demands management in the GA process. The flow chart realizes the process activities and data flow between different phases and departments. The data transfers are visualized through white parallelograms, while the activities within the distinct phases are visualized through light grey rect- angles. 4.1.2 Mapping of Current Process The phases within the product realization loop are also evident within the Polestar organization, with different departments working in various stages. However, find- ings from the study of the GA process, along with content analyses from the various interviews, imply that the phases within Polestar are not as clearly defined as they are in the literature study, which is presented in Section 2.1.2. The results of the mapping will be presented in the following subsections. 4.1.2.1 Concept Phase The study results and insights from various interviews indicate that the Concept Phase of the final demands management within the GA process is overseen by the Vehicle Integration and Geometry team, a component of the R&D department. The Geo Engineers within this team closely collaborate with the requirements and final 25 4. Results demands, engaging in calculations of the requirements and their tolerances. The Concept Phase of the current process commences upon the delivery of the ve- hicle design and engineering models with demand data from the Design Engineers in the R&D Engineering team. This delivery comprises CAD information, repre- senting the vehicle’s geometry, both A and B surfaces, along with material details for the vehicle parts. The CAD files are manually imported into the designated CAT software, RD&T. Concurrently, the R&D Geo team manually creates an Excel sheet known as the Gap Plan, which outlines the vehicle’s specifications, including dimensional attributes such as final demands including tolerances. Additionally, the team is tasked with analyzing the Gap Plan to determine the target values for visual appearance and functional requirements, as part of dimensional engineering. This activity is conducted iteratively. Once the CAD files are imported into the RD&T software, several in-software steps are initiated. Firstly, the model is prepared and built by positioning the CAD parts according to reference system and assembly process, while simultaneously updating the Gap Plan. Thus, the Geo Engineers also work on the vehicle’s aesthetics since there is no PQ department in Polestar. However, they do not have final ownership of the PQ demands. The assessment of tolerances affecting the vehicle’s appear- ance is either discussed with the team or determined based on individual expertise during the vehicle dimensioning process. Further, this allows for conducting var- ious simulations, such as variation or contribution analyses, based on the type of final demand and the desired output of the engineer. The subsequent steps in the software are managed by the same team but are part of the next phase of the GA process, which will be presented in the next section. The results of the analyses can be iterated based on outcomes, in such cases, the simulation is conducted once again with modifications and optimizations until the desired outcome is reached. In other cases, when optimization or justification is not possible, the demand or design can be balanced. A balancing is carried out to determine whether the demand or design needs modification. Simultaneously, another activity takes place, known as ad-hoc design assistance, which can be outside of the actual work process. This involves providing dimensional assistance to Design Engineers, from the R&D Engineering team, as needed. When Design Engineers require dimensional assistance, they often send reflections or CAD files via email, to get evaluations for their thoughts or to determine which performs best from a dimensional standpoint. In these cases, the CAD files or reflections are circulated internally and the analysis is conducted to provide feedback directly to the Design Engineer. 4.1.2.2 Verification and Production Phase The results obtained from studying the current process indicate that there is not a clear, distinct barrier between the Verification Phase and the Production Phase of the current process at Polestar. The transition between the phases is merged as the ME Geo team and R&D Geo team are involved in both and integrate the two 26 4. Results processes through their work. Thus, the results obtained from studying the phases will be presented jointly. The Verification Phase is initiated through the final task of the R&D Geo team, which is to conduct an inspection preparation in RD&T. Inspection preparation in- volves identifying the necessary measuring points, or MP:s, for validation of the specifications outlined in the product drawings. Defining MP:s at each level, com- ponent, sub-assembly, assembly, and final product is crucial for ensuring precise verification of the product quality [39]. Subsequently, PKI, process control instruc- tion, drawings are created to provide operators with instructions on how to conduct measurements of the gap and flush demands by specifying the nominal values and tolerances. This also includes a measurement tool, a digital gauge, that is imported to RD&T and placed appropriately. The outputs from RD&T are part drawings in the form of JPEG files with outlined MP:s and sections, and text files containing input data for the following step. With the inspection preparation completed, the inspection and measurement of the defined gap and flush demands can be carried out. The data obtained from the inspection preparation is handed over to a Geo Engineer in the ME Geo team. The inspection activities are defined in a control plan, which is used as a basis for the measuring process. Firstly, a project containing the PKI drawing and data from the inspection preparation is created in a software called FDA, Final Demand Applica- tion, used for part verification. The current project is assigned in the interface and the site where the measuring is carried out is specified, usually at the production facility in Chongqing, China. Using a calibrated gauge for measuring, the nominal gap and flush demands specified in the PKI are measured by an operator on a vehicle or nominal buck, depending on the project’s status. A nominal buck is an aluminum hardware that serves as a conversion of the CAD models of the parts. It is a tool used for analysis and verification of the part geometry with interchangeable parts to be able to analyze production parts in a nominal environment, utilized among other tools for decision-making on demand or construction changes. In some cases, the outcome of the part verification may indicate that new PKI:s are necessary. The reason for this might be that the specified digital gauge selected in the inspection preparation was not fit for the part verification. This feedback infor- mation is then sent back to the R&D Geo team who iterates the update of the PKI:s. The results from measuring the demands on the vehicle or nominal buck are doc- umented by a Geo Engineer in FDA. The field will be displayed in different colors depending on whether the deviation of the measured value falls within the speci- fied tolerances. The results are automatically exported to CM4D, Computer Mea- surement Machine Management Mechanism Data, a database used for storing and analyzing data, from FDA. Once the data reaches CM4D, it is stored and utilized for analysis of the measurements. The CM4D tool does not generate any further output flow, marking the end of the data flow within the scope of the study. 27 4. Results The demand that was not within the specification and thus turned red is managed in PIRA. PIRA is a portal used within Polestar to report on product or process- related issues identified in product projects. In order to meet the final demands, an applicable action needs to be carried out for the unmet demand. One of the options is to modify the demand, which implies that feedback on the results is communicated back to R&D Geo team where the demand can be changed. To decide on an appropriate action, several different methods can be utilized to support the decision. A common practice is to identify the root cause of the demand not being within specifications, which can facilitate addressing the issue. This is termed Root Cause Analysis, or RCA, and is performed by the R&D Geo team and ME Geo team. 4.1.3 Significant Gates in the Current Process In the current process, several key milestones are navigated within the scope of the study. These include the project start, PS, representing the commencement of the vehicle program and the formulation of project requisites necessary for advancement to the current process; the final data judgment, FDJ, representing the completion of the vehicle design for verification purposes, signifying no further alterations re- quired; and the prototype builds phase, VP, wherein prototype production begins to verify previous calculations on a complete vehicle before commencing mass pro- duction. To summarize, the PS marks the initiation of the car program, allowing for the establishment of necessary project requirements. The FDJ stage indicates the com- pletion of design verification with no further adjustments necessary. Finally, the VP entails the commencement of prototype production to validate previous calculations before proceeding to full-scale production. 4.1.4 Graphical Representation of the Current Process This section provides a visual representation of the current process through a de- tailed presentation of a process and data flow diagram. Subsequently, a compre- hensive narrative elaborates upon the activities of the depicted process, enhancing understanding of the work flow. Additionally, a structured SIPOC diagram is in- cluded to further illuminate the process and to provide a simplified overview of the complex process flow. 4.1.4.1 Process and Data Flow Diagram The outcomes derived from the process and data flow are documented in Appendix F. The diagram successfully illustrates the activities within the three phases, incor- porating interconnected data flows between them, alongside details indicating the responsible department for each phase or activity. The different types of symbols and line styles indicate various items, which also is explained in Figure 4.3 below. 28 4. Results Manual operation Role/responsible Iteration Flow direction Feedback direction Software  Activity Decision Stage Data Database Stored Data Figure 4.3: Description of symbols and line styles used in the flow diagram. The initial segment of the comprehensive diagram illustrates the activities between the PS and the FDJ including the ones within the Concept Phase, as depicted in Figure 4.4. 29 4. Results Section points Measures Defined demands Simulation results Create and fill in the Gap Plan for the current program Create measures of PQ demands Define section points on two surfaces Assign tolerances Run variation and contribution analyses Specify geometry demands Engineering models Concept Phase RD&T: Simulation Create MP sections Verification Phase RD&T: Inspection Preparation Import PKI drawings Final Demand Application: Part Verification Measure the demand on the vehicle / nominal buck with a gauge Geo Engineer (R&D) Geo Engineer  (ME Plant) Build model Model Modify model Change tolerances Run variation and contribution analyses Looped analysis results Stored section points and measures Loop #n Demand data the Gap Plan OK? YesNo Optimized calculation? NoYes Balancing of demand Operator (Plant) R & D G e o T e a m M E G e o T e a m F D J Project content and prerequisites PS Feedback information Feedback information Data VP - X weeks Request for design change 1.1 0.1 1.2 0.2 1.3 1.4 1.5 2.11.6 a. b.b.1 1.a 2.6 1.b 2.7 1.c Figure 4.4: Initial segment of the flow diagram. As depicted in the first segment of the flow diagram, as seen above, the initial step occurs at PS, where the project content and prerequisites are established. Since this step is crucial before commencing the actual process, it was considered necessary to include it. The flow then progresses to the design and engineering team within the R&D department, where the geometry demands are specified before transmitting the demand data to the R&D Geo team. The Concept Phase numbered as 1 in the figure above, commences with the creation of the Gap Plan in Excel, which is completed by the Geo Engineer from R&D. Sub- sequently, the Geo Engineer progresses to the simulation, utilizing RD&T, where the input data consists of demand names and types extracted from the Gap Plan and relevant CAD files. The software process initiates by constructing the simulation model, which also serves as the output data, followed by defining section points be- tween the surfaces to facilitate subsequent measurements. This section point data is then utilized to generate measures of PQ demands, such as gap and flush. Following this, tolerances are assigned, resulting in finalized simulation models. This enables 30 4. Results the execution of various analyses within the software. The output comprises the analysis results, which are then manually entered into the Gap Plan to determine if the demand aligns with the target value for proceeding to the next phase. This evaluation is visually depicted in the Gap Plan, where green signifies compliance with the demand and red denotes that it failed to meet the demand, providing the necessary information to address decision point a in the flow diagram. In the case of the demand being indicated as red, meaning it does not comply with specified demand data in the Gap Plan, the information is looped back to a deci- sion point. Here, the calculation undergoes evaluation to determine if it has been adequately optimized. If the answer is affirmative, the demand is balanced. This involves adjusting the demand or altering the design using various methods, de- pendent upon the desired outcome. Subsequently, the output data flows back into the Gap Plan. In cases where the design is modified, resulting in a less sensitive demand, the calculation yields an affirmative result. Another approach to design modification involves identifying the demand as unnecessary and removing it. For instance, this could entail eliminating a demand. However, if the answer is negative, indicating that optimization is not sufficient, the simulation undergoes looping, as illustrated in the zoomed-in section of the figure above. The looping process may oc- cur multiple times until the result is deemed satisfactory or accepted without further modification. The looped simulation begins with the modification of the model to optimize the results. The section points and measures generated in the initial loop are stored within the software and reused in subsequent iterations. Subsequently, the process moves to adjusting tolerances, followed by another round of analyses. The results of these looped analyses are then integrated into the Gap Plan to re- assess compliance with the requirements.If the changed results are satisfactory, the flow proceeds to the Verification Phase. This phase often occurs immediately after the FDJ is established, coinciding with the ordering of the nominal buck. The Verification Phase is commenced by conducting the inspection preparation in RD&T, using the same session and performed by the same Geo Engineer from R&D Geo team, as depicted in Figure 4.5. 31 4. Results Depicted MP section layouts PKI drawings and data Create needles for each demand Create MP sections Define section for measuring Create PKI:s with gauges Verification Phase RD&T: Inspection Preparation Export PKI as text and JPEG files   Distance measure Document the measurement result Import PKI drawings Final Demand Application: Part Verification Measure the demand on the vehicle / nominal buck with a gauge Verfication output Deviation within tolerance?  PKI text and JPEG files Send output to CM4D Perform RCA and Create PIRA Geo Engineer (ME Plant)Geo Engineer  (ME Plant) Output Geo Engineer (ME Plant) Action O th er Demand change Create MP drawings Measurement Points New PKI necessary? No Yes Operator (Plant) Go to activity 2.3 or remove gauge from PKI Feedback information VP - X weeks No Yes 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 c. c.1 d. 2.9 e.d.1 Figure 4.5: The Verification Phase in the flow diagram. The ongoing process in RD&T is followed by the creation of MP sections, resulting in the output of MP section layouts. This data is subsequently utilized to gener- ate needles to create measurement points for each demand found in the Gap Plan, which the Geo Engineer is currently working on. In parallel to the main work flow, the output of measurement points may also be employed to generate MP drawings if necessary. The output data, comprising measurement points, is then employed to create PKI:s using Polestar’s gauges. The digital gauges represent the physical gauges utilized in the plant for the part verification. This results in the generation of PKI drawings and associated data. All data from preceding activities is then uti- lized to export the PKI:s as text and JPEG files. That output can then be shared via SharePoint, or sent via email if needed to the next Geo Engineer, who is a part of the ME Geo team working at the plant. Between the last activity in RD&T and the part verification in FDA, there is a stage labeled as "VP - X weeks." This signifies that it occurs a few weeks before the construction of the VP begins. Until this point, any necessary changes to the PKIs can be made. Otherwise, the engineers will wait for the construction to com- 32 4. Results mence. Subsequently, the Geo Engineer from the plant initiates the necessary work in FDA. The PKI data provided by the Geo Engineer from the R&D Geo team is manually imported into FDA. To progress within the application, an operator at the plant measures the demand on either the vehicle or nominal buck, using a gauge. At this point, it becomes apparent whether the measurement method and instruction can be used. This evaluation determines if there is a need for a new PKI, which is assessed by the Geo Engineer. If there is a requirement for a new PKI, the information is utilized to revise the existing PKI in RD&T. Conversely, if there is no need for a new PKI, the operator’s measurement results from a vehicle are manually inputted into FDA by the Geo Engineer for documentation. This process yields the verification output, which is then assessed in the software to determine if it falls within acceptable deviation limits. Similar to previous steps, this assessment is also represented by green and red indicators, with green indicating an accepted result and red a rejected result. All verification outputs are directed to the activity of sending the output to CM4D within the application. However, if the result is marked as red in the FDA, a parallel flow is initiated. To assess the cause of the problem, an RCA is performed, followed by the creation of a PIRA. The output of this process then leads to taking some form of action. If a demand change is deemed necessary, the information is fed back to the R&D Geo team and their Gap Plan. Conversely, if any other action is required, the output is directly sent to CM4D, which is automatically done by FDA. As the output is sent to CM4D, the Production Phase is automatically initiated, comprising the third phase in the process flow diagram, as illustrated in Figure 4.6 below. The scope of the study does not proceed further than this point, marking the termination of the analyzed data flow. 33 4. Results Analyze  Production Phase CM4D: Storing and Analysis of Measurement Data Geo Engineer (ME Plant) 3.1 Figure 4.6: The Production Phase in the flow diagram. 4.1.4.2 SIPOC Diagram An overview of the process flow can be seen as a SIPOC diagram in Figure 4.7. The respective departments involved in the process are listed as suppliers of the inputs and customers, meaning receivers, of the outputs from the process. A brief overview of the workflow with the key steps is seen in the process column in the di- agram, aligned with the corresponding inputs and outputs linked to their respective suppliers and customers. 34 4. Results R&D Eng S I P O C SUPPLIERS INPUTS PROCESS OUTPUTS CUSTOMERS R&D Geo ME Geo Engineering models Demand data R&D Geo Demand data PKI Data PKI Data ME Geo Stored and analyzed measurement data Simulation Inspection preparation Part verification Analysis and storing of measurement data  Start End Verification output Verification output Simulation results Simulation results Car program Figure 4.7: SIPOC diagram. R&D Engineering, R&D Eng, is included as a supplier in the diagram, as they provide the R&D Geo team with the input Engineering models. As their role is not included in the scope of the process that is studied, their work is separated with a dashed border. Further, Car program, which is the entire development project of the final product, is placed as the customer for the final output of the process. 4.1.5 Pilot Process This subsection presents the results of conducting the Pilot Study of the current process in three phases: Concept, Verification and Production. 4.1.5.1 Concept Phase The two selected demands, along with their corresponding components, were EX076 and EX077, which included the components “Rear Fender” and “Rear Body Lamp”, placed in the backside of the vehicle. Both demands include gap and flush require- ments with imaginary values, as there are split-lines between them that require measurement and tolerances. The demands can be seen in Figure 4.8, adapted from the Gap Plan. 35 4. Results Figure 4.8: Visual representation of the placement of demands EX076 and EX077 with imaginary values. . As seen in the figure above, the demands exhibit both a nominal value and a toler- ance for both the gap and flush requirements, as detailed in Table 4.1 below. Table 4.1: Demands EX076 and EX077. The chosen components and demands put into RD&T yielded the following results of the measurements shown in Figure 4.9. 36 4. Results Figure 4.9: The gap and flush demands as measures in RD&T (EX076 to the left and EX077 to the right). The figure illustrates that the measurement of the gap demand is taken at the outermost corner of the curved edges of the components, while the flush demand is measured from the visible Class A surfaces. Further, the measurements, along with the assigned tolerances from the tolerance chain in the components, produced the results of the variation analyses for the gap demands, shown in Figures 4.10 and 4.11. Figure 4.10: Variation analysis results from RD&T for EX076 (Gap demand). 37 4. Results Figure 4.11: Variation analysis results from RD&T for EX077 (Gap demand). Similarly, the variation analysis was conducted for the flush demand and that yielded the results presented in Figures 4.12 and 4.13 below. Figure 4.12: Variation analysis results from RD&T for EX076 (Flush demand). 38 4. Results Figure 4.13: Variation analysis results from RD&T for EX077 (Flush demand). . The final results were compiled in an RD&T-based document, which can be seen in Figures 4.14 and 4.15. These results were controlled by comparing them with the values in the Gap Plan. Figure 4.14: Results for demand EX076. 39 4. Results Figure 4.15: Results for demand EX077. As the results, seen above, were not affirmative the values were compared with the target values in the Gap Plan to see if it was within acceptable tolerance or not. With help from experts, the demands were accepted to proceed to the next step. If it were not accepted, it would have needed a balancing of the demand or design. 4.1.5.2 Verification and Production Phase The needles creating measurement points can be observed in the MP sections. In Figure 4.16, the MP section for demand EX076 is depicted, which looks the same as the MP section for demand EX077. It visualizes the cross section that shows the relation between the components. 40 4. Results Figure 4.16: MP section for demand EX076. In Figures 4.17 and 4.18, the MP drawings can be found where the placement of the demands is visible on the components. Figure 4.17: MP drawing for the rear fender. 41 4. Results Figure 4.18: MP drawing for the rear body lamp. PKI drawings were created in the MP sections for the EX076 gap and flush demands, which can be found in Figures 4.19 and 4.20. The gauges measuring the relations are visualized in the drawings, with the measured values presented in the table below the pictures. Figure 4.19: PKI drawing with gauge for EX076 gap demand. 42 4. Results Figure 4.20: PKI drawing with gauge for EX076 flush demand. The PKI drawing imported into FDA prompts the application and several inputs are chosen to proceed. Site and project were assigned. Once the inputs were assigned the button “Measure” was clicked to prompt the “Measurement” page, which can be seen in Figure 4.21. Figure 4.21: Measurement page in FDA. The imported PKI drawings are shown in the figure above as features, where the drawing changes as the measurements are filled in. Parallelly, the results from the operator measuring the nominal buck with the assigned gauges were conducted and the documented result can be seen in Figure 4.22. 43 4. Results Figure 4.22: Results of measurements on the nominal buck. In the figure above, it can be seen that three out of four demands are marked as green, while the fourth one is marked red. Next, the results are filtered according to the input boxes, such as Body ID, Measurement Type and Build Series, which can be seen in Figure 4.23. 44 4. Results Figure 4.23: Filtering of the measurement results. The results sent into CM4D are graphically visualized to show the correlation be- tween the final demands and components, which is depicted in Figure 4.24. 45 4. Results Figure 4.24: The results in CM4D where the confidential information is covered. In the figure above, the graphical representation of the final demands can be ob- served, as well as the iterated measurement results in the lower right corner. Ad- ditionally, the user can ascertain whether it is fulfilled and review the imported numerical results measured by the FDA. 4.2 Facets of the Current Process The facets of the current process are addressed in this section, where the findings derived from in-house interviews and the Pilot Study will be presented to provide a comprehensive understanding of the various aspects of the current process. 4.2.1 Interviews within Polestar In this subsection, the facets of the current process derived from interviews con- ducted with engineers within the organization are presented. This includes the most apparent challenges, potential consequences of those challenges, strengths, improve- ments, and lastly, a summary of those findings. The transcription of the interviews can be found in Appendix D Challenges Firstly, the content analysis resulted in several findings of similarities in the answers given regarding challenges with the current process. The time perspective was found to be the most significant challenge, where several respondents claim that the pro- 46 4. Results cess lacks a proper time plan including deadlines for deliverables. Further, some respondents experience a lack of a standardized work process with clear instructions and a common understanding of the work. Thus, some of the respondents claim that they have established an individual-based working procedure, due to lack of instruc- tions. It was also mentioned that there is not always a clear, common agreement on the work among the teams, as official hand-shakes between the teams are missing in some cases. Additionally, the storing and hand-overs of data were described as un- structured, specifically in the Concept Phase. This is believed to be a consequence of manual data management in some activities, which was also indicated to be a common challenge among the respondents. Another challenge emphasized particularly in the Concept Phase, is the absence of a department that specializes in PQ at Polestar. The respondents engaged in the Concept Phase explained that PQ is informally integrated into their process, as there is a collective responsibility within the team. The R&D Geo team involved in the Concept Phase is expected to consider PQ when conducting their work, relying on their personal experience and expertise although none of them are PQ experts. Furthermore, one of the respondents from the Concept Phase emphasized that their team primarily focuses on dimensional engineering and lacks the perspective of PQ engineers. PQ engineers not only consider dimensional appearance but also conduct thorough historical research and market benchmarking. Consequently, the absence of in-depth work from a dedicated PQ department, as highlighted by both respon- dents from the Concept Phase, may interfere with the holistic view of the vehicle’s appearance. It was also emphasized that it is challenging to ensure that other teams understand the data, particularly from the R&D Geo team in the Concept Phase, which ends up with irrelevant analyses made by other teams. Potential Consequences of the Challenges The respondents claim that as a result of the above mentioned challenges, the pro- cess can become time-consuming which can cause delays. As the process appears unstructured, it leads to difficulties with navigating through the work process and tracking the data flow. Other consequences include sensitivity to variation and de- viation. Most respondents claimed that as a result of the absence of a defined, standardized work process, a lot of the work conducted by the engineers has become individual-based. Thus, deviation factors such as a personnel change would imply a significant challenge in the work process. Additionally, one respondent explained that this highlights the necessity for proper gateways, as agreements were often based on trust. Without such formal checkpoints, potential challenges, as implied by the respondent, may arise in reality. This could lead engineers to proceed with activities even though prior tasks have not been completed or approved. Finally, unspecified collective responsibility, like in the case of PQ, can cause issues with a lack of responsibility. Strengths 47 4. Results The most commonly mentioned strength of the process is that the communication works well, although it is not entirely formal since it is carried out through mail conversations without following a set standard. Another strength mentioned by the respondents is the fact that eventually, the work gets done, although it is not done through a well-defined process. The quality of the work relies highly on the exper- tise of the engineers, meaning that they manage to obtain good results even without clear instructions. This demonstrates a strength within Polestar as an organization, as it enables agile and rapid decision-making processes. Improvements Among the suggestions on how the process can be improved and managed more efficiently with regards to data management is firstly a more evident time plan, including clear deadlines and deliverables. Further, more efficient hand-overs and official hand-shakes between the teams are necessary, along with options to track and store data. Another improvement is establishing a collective agreement on the work process, which can be obtained by clear instructions and objectives. This in- cludes an evident process for iteration and feedback between teams. Finally, smart solutions such as automated management of data transfer and formatting between steps in the process would enable it to be more structured. As one respondent said: “The culture at Polestar is not used to standards but it needs a change...a standard- ized process is needed with room for own initiative” The quotation suggests that the organization requires a standardized process that is sufficiently detailed to allow for individual initiative without hindering creativity. Summary The key takeaways from the analysis highlighting significant reoccurring findings from the interviews are summarized in Table 4.2. 48 4. Results Table 4.2: Key takeaways from analysis of interview answers. 4.2.2 Findings from the Pilot Study The identified bottlenecks and ineffective methods encountered during the Pilot Study can be found in Table 4.3. The findings are categorized based on the corre- sponding activity or document that was found to have bottlenecks. 49 4. Results Table 4.3: Identified bottlenecks from the Pilot Study. In the table, findings coded with a star sign (*) indicate that they fall outside the scope of this report. For instance, the issue with calibrating the gauges numbered 4.3 requires precision from the engineer and cannot be solved in any other way, unless changes are made to the construction of the gauge. This is not within the focus area of the thesis as it concerns the manufacturer of the gauges and does not directly relate to the process and data flow at Polestar. Similarly, the manual input to the Gap Plan from the demand data numbered 1.1, relates to a department that provides the demand data before the process commences, which is why that step is recognized but not within the scope. Furthermore, findings numbered in the table without a star sign (*) are those that will be processed and evaluated for improvement. 4.3 Facets of Processes at Competing Companies The interviewed engineers, outside the company, were the following: • Principal Engineer at the department of PQ at Volvo Car Corporation, VCC, in Gothenburg, Sweden. 50 4. Results • Head of Project Quality and Vehicle Integration at ZEEKR (former CEVT) in Gothenburg, Sweden. The conducted interviews lasted approximately one hour each and the transcribed answers for each participant can be found in Appendix G. The content analysis of the answers revealed more similarities than differences, de- spite the interviewees being from different companies. The key findings gathered from the interviews are presented below. Challenges The challenges identified in their GA processes, consistent for both companies, in- clude the need to balance technical attributes, such as safety requirements, with PQ attributes. This necessity leads to compromises between the two during the vehicle development process. The interviewee from VCC mentioned that they constantly hold meetings to discuss the overall perception the car needs to give to its customers, making trade-offs accordingly. Similarly, the interviewee from ZEEKR mentioned iterative meetings with other departments to assess trade-offs, all stemming from the core feeling and perception the car and brand aim to convey. This was conveyed as a challenge, as the PQ departments are often not the ones making the final deci- sion. Consequently, the PQ suffers, impacting the overall GA. Furthermore, the competing companies, being larger and more established, en- counter difficulties in making changes later in the process. This challenge arises due to clearer delivery times and defined points when the design is closed for changes. Additionally, because of their size, these companies have production facilities in dif- ferent countries, leading to various suppliers and, consequently, variations in meeting requirements. This diversity imposes a challenge in identifying patterns of the prob- lem, as finding the root cause becomes more challenging. Strengths The competing companies, each with a dedicated