Flexible layout design for the battery cell manufacturing industry to optimize for the unknown future A comparative analysis of how different project models increases flexibility in layout design Master’s thesis in Production Engineering Pontus Axelsson Maja Jansson Department of Industrial and Material science CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2024 Master’s thesis 2024 Flexible layout design for the battery cell manufacturing industry to optimize for the unknown future A comparative analysis of how different project models increases flexibility in layout design Axelsson, Pontus & Jansson, Maja Department of Industrial and Materials Science Chalmers University of Technology Gothenburg, Sweden 2024 Flexible layout design for the battery cell manufacturing industry to optimize for the unknown future A comparative analysis of how different project models in- creases flexibility in layout design Pontus Axelsson Maja Jansson © Axelsson, P & Jansson, M. 2024. Supervisor: Bokrantz, Jon, Department of Industrial and Material science Advisor: Sigvardsson, Johanna, NOVO Energy AB Examiner: Skoogh, Anders, Department of Industrial and Material science Master’s Thesis 2024 Department of Industrial and Material science Chalmers University of Technology and University of Gothenburg SE-412 96 Gothenburg Telephone +46 31 772 1000 Gothenburg, Sweden 2024 iv Abstract The worlds transition towards a more sustainable society is a driver for the demand in electrical batteries for vehicles. To be able to meet the demand new battery cell factories needs to be build, and start their production as fast as possible. This master thesis has therefore investigated how the factory layout design can increase its flexibility by using different project models as a basis. The results shows that the flexibility can be increased, but the investment cost is in most of the cases higher than if flexibility in the layout is disregarded. The highest level of flexibility would be reached with help of the agile methodology where the generated layout is based on smaller modules. The set-based method could lead to an increase in flexibility depending on the scenario, but has several other advantages. The waterfall methodology would on the other hand not increase the flexibility of the layouts if not re-configurable manufacturing systems (RMS) are included. Therefore the battery cell manufacturers needs to make a choice between saving investment cost and instead face the risk of re-layout costs or if they want to create a flexible factory that can change its production together with the changing customer demands. Keywords: Battery Cell, Factory Layout, Project Models, Flexibility, Production Processes, Li-Ion battery. v Acknowledgements This Master Thesis were carried out during the spring of 2024 at NOVO Energy AB and the deparment of Industrial and Material science at Chalmers University of Technology. We would like to express our gratitude to everyone involved in this project. We are very grateful for the opportunity to write our Master Thesis at NOVO Energy AB under the supervision of Johanna Sigvardsson. Johanna has been a huge resource and helped us get to where we are today. We would also like to express our gratitude to everyone at NOVO Energy AB for helping us, answering all of our questions and taking the time to help produce this thesis. We would also like to express our gratitude to our supervisor at Chalmers University of Technology, Jon Bokrantz. Thank you for your inputs and ideas throughout this project and for helping us finding the right directions in situations were we needed it. This is the final project of our Civil Engineer education after five long years at Chalmers University of Technology and therefore we would also like to thank every- one that has been a part of our journey to this day. Pontus Axelsson & Maja Jansson, Gothenburg, 2024-06-05 Declaration of AI technologies in the thesis work and writing process This thesis have been using different AI technologies in order to improve the writing process. However, all the information and provided text have been written and formulated in advance by the thesis workers. Chat GPT - To improve and formulate different sentences and sections to provide an clear and understandable text for the reader. Grammarly - To correct spelling errors and bad structure in the text. vii Contents List of Figures xiii List of Tables xv 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.1 Problem formulation . . . . . . . . . . . . . . . . . . . . . . . 2 1.2 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Research question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.4 Delimitation’s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2 Theoretical Framework 5 2.1 Production Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.1.1 Decision Areas . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.1.2 Production Development . . . . . . . . . . . . . . . . . . . . . 7 2.1.3 Reconfigurable Manufacturing Systems . . . . . . . . . . . . . 8 2.2 Li-Ion Batteries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2.1 Different Recipies . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2.2 Different Products and Future Batteries . . . . . . . . . . . . 10 2.3 Production of Prismatic Li-Ion Batteries . . . . . . . . . . . . . . . . 11 2.3.1 Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.3.2 Electrode Manufacturing . . . . . . . . . . . . . . . . . . . . . 12 2.3.2.1 Slurry Mixing . . . . . . . . . . . . . . . . . . . . . . 13 2.3.2.2 Coating . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.3.2.3 Calendering . . . . . . . . . . . . . . . . . . . . . . . 13 2.3.2.4 Slitting . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.3.3 Cell Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.3.3.1 Stacking and Winding . . . . . . . . . . . . . . . . . 14 2.3.3.2 Welding . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.3.3.3 Packaging . . . . . . . . . . . . . . . . . . . . . . . . 15 2.3.4 Cell Finishing . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.3.4.1 Electrolyte Filling . . . . . . . . . . . . . . . . . . . 16 2.3.4.2 Soaking . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.3.4.3 Pre-Charge . . . . . . . . . . . . . . . . . . . . . . . 17 2.3.4.4 Tray Exchange . . . . . . . . . . . . . . . . . . . . . 17 ix Contents 2.3.4.5 Electrolyte Filling . . . . . . . . . . . . . . . . . . . 17 2.3.4.6 Soaking . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.3.4.7 High and Room Temperature Ageing . . . . . . . . . 18 2.3.4.8 Formation . . . . . . . . . . . . . . . . . . . . . . . . 18 2.3.4.9 Degassing . . . . . . . . . . . . . . . . . . . . . . . . 18 2.3.4.10 Sealing . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.3.4.11 Room Temperature Ageing . . . . . . . . . . . . . . 18 2.3.4.12 End Of Line . . . . . . . . . . . . . . . . . . . . . . . 18 2.4 Project Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.4.1 Waterfall Method . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.4.1.1 Requirements phase . . . . . . . . . . . . . . . . . . 19 2.4.1.2 Design phase . . . . . . . . . . . . . . . . . . . . . . 20 2.4.1.3 Implementation phase . . . . . . . . . . . . . . . . . 20 2.4.1.4 Verification phase . . . . . . . . . . . . . . . . . . . . 20 2.4.1.5 Maintenance phase . . . . . . . . . . . . . . . . . . . 20 2.4.2 Concurrent Engineering . . . . . . . . . . . . . . . . . . . . . 20 2.4.2.1 Set Based Concurrent Engineering . . . . . . . . . . 21 2.4.2.2 Principles for set-based Concurrent Engineering . . . 21 2.4.3 Agile Method . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.4.4 Other Project Models . . . . . . . . . . . . . . . . . . . . . . . 27 2.5 Factory Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.5.1 Traditional Layouts . . . . . . . . . . . . . . . . . . . . . . . . 27 2.5.1.1 Fixed Position Layout . . . . . . . . . . . . . . . . . 27 2.5.1.2 Functional Layout . . . . . . . . . . . . . . . . . . . 28 2.5.1.3 Cell Layout . . . . . . . . . . . . . . . . . . . . . . . 28 2.5.1.4 Line Flow Layout . . . . . . . . . . . . . . . . . . . . 29 2.5.2 Dynamic Layout Problem . . . . . . . . . . . . . . . . . . . . 30 2.5.2.1 Re-configurable approach . . . . . . . . . . . . . . . 30 2.5.2.2 Robust Approach . . . . . . . . . . . . . . . . . . . . 32 3 Methodology 35 3.1 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 3.2 Research tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.2.1 Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.2.2 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.2.3 Qualitative analysis . . . . . . . . . . . . . . . . . . . . . . . . 38 3.3 Phase 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3.4 Phase 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3.5 Phase 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.6 Phase 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.6.1 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.6.2 Requirements and wishes . . . . . . . . . . . . . . . . . . . . . 40 3.6.3 Waterfall Method . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.6.4 Set-based Concurrent Method . . . . . . . . . . . . . . . . . . 42 3.6.5 Agile Method . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.6.6 Further development of the layouts . . . . . . . . . . . . . . . 44 x Contents 3.6.7 Layout Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.7 Phase 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4 Results 45 4.1 Phase 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.2 Phase 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.3 Phase 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.4 Phase 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 4.4.1 Requirements and wishes . . . . . . . . . . . . . . . . . . . . . 46 4.4.2 Waterfall Layout . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.4.3 Set-based Concurrent Layout . . . . . . . . . . . . . . . . . . 49 4.4.4 Agile Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 4.4.5 Comparison between the methods . . . . . . . . . . . . . . . . 60 5 Discussion 63 5.1 Waterfall methodology . . . . . . . . . . . . . . . . . . . . . . . . . . 63 5.2 Waterfall Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 5.3 Set-based Concurrent method . . . . . . . . . . . . . . . . . . . . . . 64 5.4 Set-based Concurrent Layout . . . . . . . . . . . . . . . . . . . . . . 65 5.5 Agile Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 5.6 Agile Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 5.7 Combination of methods . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.8 Answer to research questions . . . . . . . . . . . . . . . . . . . . . . . 68 5.9 Research Methodology Discussion . . . . . . . . . . . . . . . . . . . . 70 5.10 Ethical, societal and ecological perspectives . . . . . . . . . . . . . . . 71 5.11 Future battery cell factories and further research . . . . . . . . . . . . 72 6 Conclusion 75 A Appendix A I B Appendix B III C Appendix C V D Appendix D VII xi Contents xii List of Figures 1 Product-process matrix between volume and production variants/process based on Bellgran & Säfsten (2010). . . . . . . . . . . . . . . . . . . . 7 2 The product realization process based on Bellgran & Säfsten (2010). . 8 3 The production realization process, part of the innovation process and the product life-cycle based on Bellgran & Säfsten (2010). . . . . . . 8 4 The production processes at NOVO Energy . . . . . . . . . . . . . . 12 5 The production processes for electrode manufacturing . . . . . . . . . 12 6 The production processes for Cell Assembly . . . . . . . . . . . . . . 14 7 Cell Finishing Process Sequence . . . . . . . . . . . . . . . . . . . . . 16 8 The Waterfall methods phases and order based on Model (2015). . . . 19 9 Principles of SBCE based on (Raudberget, 2010). . . . . . . . . . . . 22 10 Evaluation matrix based on Sobek et al., (1999) . . . . . . . . . . . . 23 11 Scrum method based on Rösiö et al., (2020). . . . . . . . . . . . . . . 26 12 Example of Fixed Position Layout . . . . . . . . . . . . . . . . . . . . 28 13 Example of Functional Layout . . . . . . . . . . . . . . . . . . . . . . 28 14 Example of Cell Layout . . . . . . . . . . . . . . . . . . . . . . . . . . 29 15 Example of Line Flow Layout . . . . . . . . . . . . . . . . . . . . . . 29 16 Example of a Modular Layout Design . . . . . . . . . . . . . . . . . . 31 17 Distributed layouts based on Lahmar & Benjaafar (2005). . . . . . . . 34 18 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 19 The used project models and their structure . . . . . . . . . . . . . . 40 20 Relationship between the different processes . . . . . . . . . . . . . . 47 21 Waterfall Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 22 Waterfall Layout with further development . . . . . . . . . . . . . . . 49 23 Process Function Layout . . . . . . . . . . . . . . . . . . . . . . . . . 52 24 Facility Function Layout . . . . . . . . . . . . . . . . . . . . . . . . . 53 25 Product Function Layout . . . . . . . . . . . . . . . . . . . . . . . . . 53 26 Material Handling Function Layout . . . . . . . . . . . . . . . . . . . 54 27 Evaluation matrix for the function layout design . . . . . . . . . . . . 55 28 Set-based layout after intersection . . . . . . . . . . . . . . . . . . . . 56 29 Set-based layout after workshop . . . . . . . . . . . . . . . . . . . . . 57 30 Set-Based Layout with further development . . . . . . . . . . . . . . 58 31 Layout based on the requirements . . . . . . . . . . . . . . . . . . . . 59 32 Agile Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 xiii List of Figures 33 Agile Layout with further development . . . . . . . . . . . . . . . . . 60 34 First Iteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII 35 Second Iteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII 36 Third Iteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII 37 Fourth Iteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX 38 Fifth Iteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX 39 Sixth Iteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X 40 Seventh Iteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X xiv List of Tables 1 Decision areas and related examples . . . . . . . . . . . . . . . . . . . 6 2 Principles for Set-based concurrent engineering (Sobek, Ward, & Liker, 1999). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3 interviewee in the Pre-Study . . . . . . . . . . . . . . . . . . . . . . . 39 4 interviewee in the third phase . . . . . . . . . . . . . . . . . . . . . . 39 5 Employees participating in the workshop . . . . . . . . . . . . . . . . 41 6 Requirements on layout . . . . . . . . . . . . . . . . . . . . . . . . . . 46 7 Wishes on layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 8 Advantages and Disadvantages of the Waterfall layout . . . . . . . . . 49 9 Process Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 10 Product requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 11 Material Handling Requirements . . . . . . . . . . . . . . . . . . . . . 51 12 Facility Layout Design Requirements . . . . . . . . . . . . . . . . . . 51 13 Advantages and Disadvantages of the Set-based layout . . . . . . . . 58 14 Advantages and Disadvantages of the Agile layout . . . . . . . . . . . 60 15 Method Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 xv List of Tables xvi 1 Introduction This chapter aims to provide a brief background for NOVO Energys problem state- ment and why the master thesis is relevant for both the company and the academic world. The chapter also includes the aim, limitations as well as the research ques- tions for the project which provide insight into the intended direction and boundaries of the thesis. 1.1 Background The demand for electrical batteries is increasing due to the increase in electrical vehicles (EV’s) and more battery cell factories are needed to meet the demand (Duffner, Mauler, Wentker, Leker, & Winter, 2021). In 2021 the production of battery cells was around 1100 GWh/year globally (Bhutada, 2021). However, the battery cell production within Europe was only 30 GWh/year (Breiter, Horetsky, Linder, & Rettig, 2022; ees Europe, 2021). Therefore the building of new factories in Europe needs to speed up and increase from the current capacity. Currently, the largest production of battery cells is occurring in Asia, but the European Union launched the European Battery Alliance in 2017 to build up the battery technology and the production capacity within the EU to reach the climate goals (European Battery Alliance, n.d; Bhutada, 2021). This master thesis is written in cooperation with NOVO Energy AB - a joint venture between Northvolt and Volvo Cars which is a Swedish company that is currently building a gigafactory in Gothenburg with a potential of producing up to 50 GWh per year when the factory is completed (NOVO Energy AB, n.d). To be able to reach the demand, new battery cell factories have to be built and according to NOVO Energy when creating a battery cell factory, many parameters need to be considered and the four most important are: 1. The product 2. The capacity 3. The process/technology 4. The enablers Given the urgency to meet the escalating demand for battery cells in Europe, tra- ditional project models present challenges. Due to lengthy lead times inherent in 1 1. Introduction these models, there is a risk of rendering initial customer needs and product designs obsolete by the time production starts. Therefore, exploring alternative project mod- els becomes imperative to ensure adaptability to evolving customer needs, product designs, and process requirements throughout the project lifecycle. In traditional project models, the customer needs are first clearly defined before the product design phase can begin. Then when the product is defined the factory layout can be designed to be able to produce the product that meets the customer’s demand (Thesing, Feldmann, & Burchardt, 2021). The problem with this traditional project model occurs when companies have to start building the factory X years before the product design and production phase due to long lead times. This results in the customer needs and product design be- coming obsolete concerning the factory layout once the production phase is initiated. Therefore, NOVO Energy requires new ways to create a more iterative process that is more prone to handle changes and disturbances in customer needs and product design as well as process design to have a factory layout design that will meet the requirements that is expected by the previous steps. 1.1.1 Problem formulation The core-problem for this thesis is therefore to investigate if the use of project-models could enhance the layout generation when the uncertainty in capacity, process, and product design is high. And therefore provide a structural way in order to increase the flexibility and robustness of the factory layout design. It should also guide the company in the decision making process when determining the factory layout. 1.2 Aim This Master Thesis aims to find a solution to optimize the layout of battery cell factories for the so-called unknown future where the customer need and product design isn’t clearly defined in advance before the building of the production to help improve future factories. The project aims to help pace up the industry and the time it takes to build each factory to enhance the transition to a more sustainable society. 1.3 Research question To achieve the project’s aims, the research questions that are being investigated is: 1. How can the use of project-models increase the flexibility of battery cell factory layout design in order to handle fluctuations in customer needs, product, and process design? 2. What is the value for the company and industry to increase the flexibility of the factory layout design? 2 1. Introduction By answering these research questions the project will help NOVO Energy to opti- mize future gigafactories and help pace up the battery cell industry. 1.4 Delimitation’s The project is scoped to focus on prismatic lithium-ion (Li-Ion) batteries and inves- tigate how the production can become more flexible in regards to changes of the customer need, product and process designs. The focus was therefore on the process steps regarding Li-Ion batteries and focused on how the factory layout could be de- signed to increase flexibility to cope with changes in the product and process design. In regards to the timeframe the production capacity was not be the primary focus and therefore was disregarded. Since the master thesis was limited by time the project’s scope was limited to the main process steps of battery cell production and scoped down to one of the major steps. The project step that was investigated was Cell Finishing due to its nonlin- earity and was therefore seen as the most suitable for flexible layouts. Cell Assembly and Electrode Manufacturing, however, are more linear and were therefore excluded due to the limited time. Another limitation of the thesis is that it focuses on the manufacturing part of the processes, not the research and development (R&D) which involves the chemical design of the product. The main focus was also on European and North American battery cell production due to easier access to research papers and other information. As battery development & manufacturing is an emerging and competitive market, the relevant research literature on the subject is in short supply with many domestic resources, especially from Asia. In this thesis, the term "optimize" is used with a focus on refining processes for generating flexible layouts suitable for various battery cell manufacturing, rather than solely pursuing an ultimate, singularly mathematical optimal layout. Therefore, the reader should acknowledge this throughout the thesis. 3 1. Introduction 4 2 Theoretical Framework The world is currently in a transition to become more sustainable and one of the movements is the shift from combustion engines to electrical motors. As the demand for batteries for EV are increasing, the production of battery cells needs to increase. Different manufacturers around the world focus on speeding up the production pace and invest in new factories. In traditional industries the project follows a logical order where the project is defined first, then the product, then the production system and lastly the layout of the production system. The battery cell industry on the other hand is expanding quickly and the way of working can instead look like that the project is firstly defined, then the production system and then the layout before the product is defined. In this chapter the theoretical framework is presented and regards production sys- tems, Li-ion batteries, project models and factory layouts. 2.1 Production Systems The traditional definition of production is "the transformation of raw materials into products by a series of energy applications, each of which affects well defined changes in the physical or chemical characteristics of the materials" (Danø, 1966). Today manufacturing and production has a broader definition compared to the original def- inition. Today, "manufacturing" encompasses a series of interrelated activitiesfrom design and materials selection to production, quality assurance, and marketinghigh- lighting the integration of design within the manufacturing process (Hitomi, 2017). A production system comprises a number of elements between which there are re- ciprocal relations (Bellgran & Säfsten, 2010). This could for instance be buildings, humans, equipment or software. However, a production system could also involve di- mensions of the decision-making process and therefore include, capital management, business management and production management to the system (Bellgran & Säf- sten, 2010) Each component is an important system resource, but are also potential sources for variation and disturbances which can be difficult to predict (Bellgran & Säfsten, 2010). Traditionally production systems have been created to suit a specific product or product family under a long time period (Rösiö et al., 2020). As the competitive environment and market is changing this production system isn’t longer suitable as 5 2. Theoretical Framework new products need to be introduced into the system which requires changes within it (Rösiö et al., 2020). Therefore the new production systems need to become more flexible and reconfigurable (Rösiö et al., 2020). 2.1.1 Decision Areas In order to create a functional production system there are areas where companies need to make decisions, which are called decision areas and can be seen in Table 1 (Bellgran & Säfsten, 2010). These areas comprise a number of issues and questions which a company has to deal with and make decisions on (Bellgran & Säfsten, 2010).This thesis will focus on the first decision area (production processes) but as stated before there are interdependencies between several areas which will affect the production processes like capacity, facility and product design. Table 1: Decision areas and related examples Decision areas Examples of decision questions Production process Process type, layout, technical level Capacity Amount, acquisition point Facility Localisation, focus Vertical integration Direction degree, relation Quality Definition, role, responsibility, control The production process involves converting resources into products, and decisions related to the production process includes considerations of process type, layout, and technological level. Process type relate to the organization of various processes and activities, with direct connections to production volume and the number of variants (Bellgran & Säfsten, 2010). A fundamental principle for categorizing production is frequency, which is how a certain product family is run in production. These are categorized in terms of single unit process, intermittent process and continuous process where intermittent process implies a production is run with a certain interval in production (Bellgran & Säfsten, 2010). Another dimension of intermittent process can also be based on decoupled or coupled flow of products. The classical product- process matrix, depicted in Figure 1, identifies the optimal process type for managing product volumes. The diagonal line in Figure 1 represents the standard position, and any deviation from this line heightens the risk of elevated costs to offset flexibility or suboptimal utilization of process flexibility (Bellgran & Säfsten, 2010). 6 2. Theoretical Framework Figure 1: Product-process matrix between volume and production variants/process based on Bellgran & Säfsten (2010). The next decision area concerning the production process is about the layout de- sign, which handles the physical arrangement of different equipment in the factory. Aspects influencing the different layout options are production volume, number of variants and relevant competitive factors (flexibility, cost, etc.). A division can be based on several different layout options and are discussed more in section 2.5. 2.1.2 Production Development When developing and introducing new products within a manufacturing context most often new assembly systems need to be developed (Bellgran & Säfsten, 2010). The reason for introducing new products differs between companies and industries and the old production system might not be suitable for the new product (Bellgran & Säfsten, 2010). Production development can be considered as a natural part of the production real- ization process where product realization refers to the process from product planning to the completed product (Bellgran & Säfsten, 2010). Product realization concerns development and production of products attractive to customers and therefore com- prises all activities necessary to develop solutions satisfying and identified customer needs, and all activities required to realize these solutions in terms of physical prod- ucts and services (Bellgran & Säfsten, 2010). In this thesis product realization is therefore considered to be a concept where product and production development are integrated processes with dependencies over each other for efficient development. The product realization process also consists of required supporting functions like production engineering, quality, IT, engineering material and process development for instance. The production realization process is illustrated in Figure 2 with the different parameters (Bellgran & Säfsten, 2010). 7 2. Theoretical Framework Figure 2: The product realization process based on Bellgran & Säfsten (2010). The production realization process is part of the innovation process which in turn comprises all the necessary activities to make a new product available for use in the market, from research and design, process and production planning to the use and service phase (Bellgran & Säfsten, 2010). The innovation process is part of the product life-cycle which involves end-of-life treatment of worn and scrapped products. These activities are illustrated in Figure 3, and are necessary to consider before, during and after the product realization process (Bellgran & Säfsten, 2010). Figure 3: The production realization process, part of the innovation process and the product life-cycle based on Bellgran & Säfsten (2010). 2.1.3 Reconfigurable Manufacturing Systems Reconfigurable manufacturing systems (RMS) is a class of manufacturing system which are capable of efficiently and quickly adapt to changes in the market, such as changes in demand, product mix or variants (Brunoe, Soerensen, & Nielsen, 2021). The need for RMS arises from unpredictable market changes, such as increasing frequency of introducing new products, changes in existing products, and shifts in process technology (Koren et al., 1999), driven by escalating economic competition 8 2. Theoretical Framework and rapid technological advancements (Koren et al., 1999). RMS encompasses both hardware and software adaptable to changes, optimizing resource utilization as pro- duction requirements evolve (Rösiö et al., 2020), designed for upgradability and flexibility to ensure optimal functionality and capacity alignment over time (Rösiö et al., 2020). The overarching objective of RMS is to establish responsive production systems capable of navigating rapid shifts in product and process technologies result- ing from fluctuations in customer demand (Koren & Shpitalni, 2010), distinguishing it from flexible production systems (FMS) which address product changes but lack adaptability for structural shifts, thus proving unsuitable for abrupt market fluc- tuations. FMS also grapples with high throughput and relies on costly equipment compared to RMS and dedicated manufacturing lines (DML) (Koren & Shpitalni, 2010). RMS are marked by six core reconfigurable characteristics for the system which are mentioned in (Koren & Shpitalni, 2010). These characteristics are customization, where the system is designed by allowing flexibility for a specific product family. It should also boast convertibility enabling easy adaptation of existing systems and machines to meet new production needs. Another characteristic is scalability per- mitting adjustments to production capacity by adding or removing manufacturing resources or system components. Modularity which divides operational functions into units, which can be manipulated to optimize the production arrangements. In- tegrability, which ensures rapid and precise module integration through mechanical, informational and control interfaces. And, lastly, diagnosability which enables au- tomatic assessment of system states to detect and address root causes of product defects efficiently. The first three characteristics are critical for RMS, while the last three allow rapid configuration and do not guarantee modifications in capacity or functionality (Koren & Shpitalni, 2010). Moreover, RMS are designed according to reconfiguration principles. These princi- ples are intended to improve reconfiguration speed and speed of responsiveness to (i) unpredictable external resources (market changes), (ii) planned product model changes, and, (iii) unexpected intrinsic events system events. As described in (Koren & Shpitalni, 2010) these principles results in that: 1. A RMS system provides adjustable production resources to respond to unpre- dictable market changes and intrinsic system events. - RMS capacity can be rapidly scalable in small increments. - RMS functionality can be rapidly adapted to new products. - RMS built in adjustment capabilities facilitate rapid response to unexpected equipment failures. 2. An RMS system is designed around a product family, with just enough cus- tomized flexibility to produce all members of that family. 3. The RMS core characteristics should be embedded in the system as a whole, as well as in its components. 9 2. Theoretical Framework 2.2 Li-Ion Batteries The development of Li-Ion batteries started in the 1980s and in 1991 they reached the commercial market (Pistoia, 2013). The major parts in Li-Ion batteries are cathode which is the positive electrode, anode which is the negative electrode, a separator and an electrolyte (Duffner et al., 2021; Pistoia, 2013). For Li-Ion batteries it is important that the battery specifications are properly designed to meet the requirements that are set upon the system as the system utilizes chemical reactions to be able to provide energy to the system the battery is connected to (Pistoia, 2013). Recently the major focus has been on increasing the energy density of the Li-Ion batteries by improving anode and cathode materials for higher voltage and energy density. 2.2.1 Different Recipies The production of prismatic Li-Ion cells can differ within the chosen recipe which in turn can lead to differences within the production processes. During the development of a new battery cell there are different samples being produced. The A samples, which are the first samples of a battery design, are produced in small sample volumes on a pilot line and undergo simple function and performance tests (Örüm Aydin et al., 2023). The B samples are manufactured on larger test lines and undergo customer tests and verification of the required product properties with the aim of freezing the product design (Örüm Aydin et al., 2023). The C samples are produced on the series production line, and checked in detail to the customers specification (Örüm Aydin et al., 2023). In the C sample phase only small adjustments are made to the product as the process design is freezed at the end of the phase and the process release occurs (Örüm Aydin et al., 2023). The problem that can occur is that if the factory has to be built before the C-sample is created, changes could be made to the product design and therefore changes in the production processes may occur. This is also possible if new batteries are developed after the factory is built. 2.2.2 Different Products and Future Batteries There are different kinds of batteries that can be used for EVs. Other than Li- Ion there are Lead-Acid batteries which are the cheapest in regards to the raw material but they also have low energy density and therefore are usually used when the operating distance and weight are low (Manzetti & Mariasiu, 2015). Another battery type is the Nickel-Metal Hydride battery which for example Toyota, Honda and Lexus uses in their hybrid EVs but no one uses them for fully electrical vehicles due to its high material cost (Manzetti & Mariasiu, 2015). Looking at future batteries for EVs one possibility is the solid state batteries (SSB). In comparison to Li-Ion batteries with liquid electrolyte and a separator SSBs would instead contain a solid electrolyte which is impenetrable to Li-metal dendrites which would therefore allow Li-metal as the anode which would result in increased energy 10 2. Theoretical Framework density of the batteries (Ulvestad, 2018). There are three types of battery cells formats which are commonly used in electric vehicles, these are cylindrical cells, pouch cells and prismatic cells. These different cell types have different types of advantages and drawbacks which need to be con- sidered as well as different manufacturing processes. The battery cell is the smallest unit of the battery packs or the single unit of battery packs used in electric vehicles (Halimah, Rahardian, & Budiman, 2019). Cylindrical cells’ most distinguished advantages is that it needs no additional mech- anism to control the pressure change during charging and discharging due to its cylindrical metal casing that helps maintain the cell (Halimah et al., 2019). An- other advantage is that it can be conveniently densely packed due to its slender cylindrical shape and fitted in the vehicle space. Some disadvantages include high weight per energy storage (Halimah et al., 2019). The pouch cell however offers the advantage of low weight per energy storage com- pared to the cylindrical cell. It is mostly used for lighter products such as electronic devices or drones (Halimah et al., 2019). A big disadvantage with this cell is that the casing is not designed to protect the battery from heavy loading (Halimah et al., 2019). The prismatic cell offers very high energy density compared to other battery types and provides a compact packing. It is very suitable for heavy-weight electric vehi- cles, a big disadvantage is that the battery cell is more costly to produce than the cylindrical and pouch cell (Halimah et al., 2019). 2.3 Production of Prismatic Li-Ion Batteries The production of Li-Ion batteries can be divided into three main steps which con- tains more processes within each step. The first main step is the Electrode Manufac- turing which contains slurry mixing, coating, calendering and slitting. The second step is Cell Assembly which contains stacking, cell assembly and packaging and the last step is Cell Finishing which contains electrolyte filling, soaking, pre-charging, degassing, sealing and ageing. the formation steps as well as the aging process. Different manufacturers divided the processes in different ways depending on their situation. In this thesis the chosen division of processes can be seen in Figure 4. There are other processes that can be included in the battery manufacturing but they have been disregarded due to lack of published information. 11 2. Theoretical Framework Figure 4: The production processes at NOVO Energy The complex part of battery manufacturing is that the electrode production belongs to process engineering, cell assembly to assembly technology and Cell finishing to electrical engineering and therefore the three different disciplines need to work to- gether to make the production work (Örüm Aydin et al., 2023). 2.3.1 Environment Another aspect that makes the battery production complex is the environment that is needed. Some processes must take place in clean and dry rooms where the particles, temperature, and humidity are strictly controlled due to the high moisture sensitivity of different materials (Lechner, Mothwurf, Nohe, & Daub, 2023; Plocher et al., 2023). The need of clean and dry rooms are a contributor to the cost and energy consumption regarding the Li-Ion battery production and as the cost for the rooms depends on their volumes it is important to regard this aspect in the designing phase of the factories (Lechner et al., 2023). 2.3.2 Electrode Manufacturing The first main process is electrode manufacturing where the electrode for anode and cathode is produced separately, due to the risk of cross-contamination. The processes within electrode manufacturing can be seen in Figure 5. Figure 5: The production processes for electrode manufacturing 12 2. Theoretical Framework 2.3.2.1 Slurry Mixing The first step in manufacturing Li-Ion batteries is the slurry mixing where the active material (most commonly graphite for anode and lithium nickel manganese cobalt oxide for cathode), conductive additives and solvent binders are mixed to create a slurry together with a solvent (Günther et al., 2016; Y. Liu, Zhang, Wang, & Wang, 2021). The active materials as well as the conductive additives are in a solid powder form and bound together with the binder which is either an organic solvent or an aqueous, in a mixer (Duffner et al., 2021; D. Liu et al., 2014). The aim of the process is to achieve the right homogeneity and viscosity of the slurry (Duffner et al., 2021). The slurry mixing process is crucial for the performance and quality of the batteries and therefore the selection of mixing devices as well as procedures are an important step (D. Liu et al., 2014). 2.3.2.2 Coating In the coating process the slurry is coated on both sides of the current collector which for anode is copper foil and for cathode aluminium foil, and then dried with hot air or thermal radiation throughout the coating machine where the solvent is evaporated and the slurry solidified to the right porosity (Duffner et al., 2021; Günther et al., 2016; Y. Liu et al., 2021). As mentioned in section 2.3.2.1 it is therefore important that the slurry has the right properties so that adhesion occurs with the foil (Duffner et al., 2021). Depending on the manufacturer and the used coating machine an additional step of drying may be required. In that case the drying is occurring in drying chambers with either air heating, infra-red or laser heating and consists of different temperature zones ranging from 50-180 degrees Celsius (Jinasena, Burheim, & Strømman, 2021). For both drying within the coating machine and the additional step the drying rate is important as too high rates can lead to binder migration and accumulation to the surface of the electrode which in turn could lead to poor adhesion properties and in worst case crack formations in the electrode (Jinasena et al., 2021). Depending on if the manufacturer wants to use stacking or winding the coating process differs as the two processes require different types of electrode sheets. 2.3.2.3 Calendering When the foil has been coated and dried the next step is calendering which adjusts the physical properties of the electrodes, for instance bonding, conductivity and the thickness (Y. Liu et al., 2021). The electrode thickness affects both the cell properties and the cost, if the thickness is increased it results in higher energy density but it decreases the power density and increases the cost (Duffner et al., 2021). 13 2. Theoretical Framework 2.3.2.4 Slitting The last step of the electrode manufacturing is slitting which cuts the electrode to reach the required dimensions to fit the cell’s design (Y. Liu et al., 2021). The working width of the coater is usually up to 1500mm which is wider than commonly used electrodes, so the foil has to be cut to the right size in the slitting process (Duffner et al., 2021). The sheets are then vacuum dried to remove any solvent that may remind on the roll (Jinasena et al., 2021). The electrode rolls are then, if the cell design includes stacking, cut into separate electrodes sheets by shear cutting or laser cutting (Jinasena et al., 2021) 2.3.3 Cell Assembly The assembly process accounts for a large contribution of the production costs of batteries, including factors like equipment depreciation, labor costs, and plant floor space expenses (Nelson, Ahmed, Gallagher, & Dees, 2019). According to (Nelson et al., 2019) that different stages of the assembly process contribute to the overall costs in varying degrees. Typically, stacking, welding, and enclosing operations are among the primary cost drivers. However, the specific contribution of each stage can vary depending on factors like the type of cell design being produced and the throughput of the plant. To maintain high-quality outcomes, these processes are often conducted in controlled environments, like dry rooms. In cases where such conditions cannot be maintained, additional steps, such as vacuum drying for 12-24 hours, may be necessary to remove moisture from the cells (Tagawa & Brodd, 2009). The production processes for Cell Assembly can be seen in Figure 6 Figure 6: The production processes for Cell Assembly 2.3.3.1 Stacking and Winding The first step within cell assembly is to either stack or wind the electrodes together with a separator. Different manufacturers uses different processes and they have different advantages and disadvantages. In the winding process the anode and cathode electrode are together with a separator wrapped to create an endless band (Duffner et al., 2021). The winding process is 14 2. Theoretical Framework very productive and precise but cause stress on the electrodes which in turn limits the energy density that can be produced within a cell (Duffner et al., 2021). Stacking on the other hand involves cutting the electrode rolls created in the cal- endering process into appropriately sized sheets, both anode and cathode by either die-based punching or with laser cutters (Y. Liu et al., 2021). The edge quality of the cutting process affects the batteries quality and safety (Y. Liu et al., 2021). These sheets are subsequently arranged in a precise alternating sequence of anode and cathode layers, separated by insulating material known as a separator, to achieve required tightness (Jinasena et al., 2021; Wu, 2015). The number of layers is contin- gent upon the specific requirements of the battery cell type and size. This assembly results in what is known as a ’jelly roll’ structure. The process of stacking ensures that the requisite number of layers is achieved (Y. Liu et al., 2021). Stacking has the advantage over winding by applying uniform mechanical load to the sheets which results in higher energy density (Kwade et al., 2018). 2.3.3.2 Welding The process initiates with the pre-welding of aluminium and copper tabs onto the anode and cathode current collectors, respectively. This step is crucial for ensuring a continuous electrical connection throughout the jelly roll. Typically, ultrasonic welding is employed for this purpose, although resistance welding may be utilized in certain cell designs (Y. Liu et al., 2021). Subsequently, the tabs on each side are collectively connected to a current collector, along with the cell lid. This sequence of actions ensures the establishment of robust electrical pathways within the cell (Michaelis et al., 2018). 2.3.3.3 Packaging In the packaging process the jelly-roll is inserted in a robust metal housing, while being insulated in insulation foil which protects the jelly roll during the insertion into the metal can (Michaelis et al., 2018). For the prismatic cell the edges of the jelly roll are typically compressed, fixed and then ultrasonically welded to contact the terminals attached to the lid of the battery (Michaelis et al., 2018).The housing is then typically sealed using an additional laser welding process (Michaelis et al., 2018). 2.3.4 Cell Finishing The cell finishing process is the final stage in the production of the battery cell (Kampker et al., 2023). Depending on the manufacturer’s protocol the cells pass through the process steps and measurements in different order (Plumeyer, Kokozin- ski, & Kampker, 2023). Therefore the process sequence that can be seen in Figure 7 is literature based and can differ between manufacturers. 15 2. Theoretical Framework Figure 7: Cell Finishing Process Sequence The formation part is the most expensive process and accounts for about 6% of the total battery cost. The process is also time-consuming and can range between 1,5-4 weeks where the batteries are undergoing different charging voltages, rest steps and degassing stages depending on cell chemistry and format (Pathan, Rashid, Walker, Widanage, & Kendrick, 2019). As cell finishing requires a large amount of equipment it can can account for 25% of the factories floor space as each individual cell must pass through the processes (Örüm Aydin et al., 2023). Due to the high cost different manufacturers are investigating new formation processes that decrease the total time within cell finishing by having rapid charge-discharge cycles and optimized temperatures etc. (Weng et al., 2021). 2.3.4.1 Electrolyte Filling Depending on manufacturer the first electrolyte filling can be placed within the cell assembly or within cell finishing. In the first electrolyte filling process the cell is filled with electrolyte in dry conditions with controlled temperatures under weak vacuum (Kwade et al., 2018). Several quality controls are usually in procedure like weighing the cell before and after filling in order to ensure that the correct amount is filled in each cell before the cells are sealed (Kwade et al., 2018). 2.3.4.2 Soaking After the cells are filled with electrolyte they are typically temporarily sealed and packed in a compression tray in order to be transported into a high-temperature room (around 40-60 degrees) for soaking (Wood, Li, & An, 2019). The soaking step is typically done in multiple steps in order for complete solid electrode interface and cathode electrolyte interface to occur. This is done by allowing the electrolyte to evenly distribute within the layers of the cell, and after they are sent for cooling (Wood et al., 2019). The duration of the soaking process at a given temperature depends on the size and the format of the cell as well as the cell chemistry and the electrolyte-filling process (Plumeyer et al., 2023). The soaking process usually occurs more than one time during the formation cycle to ensure that the process is successful (Wood et al., 2019). 16 2. Theoretical Framework 2.3.4.3 Pre-Charge Next the cells are typically pre-charged after being inspected for any electrolyte leakage. This is done by a precisely controlled charging cycle in order to activate the cell. The cells are monitored in order to address any non-conformities. During the first charging when electrolyte is accessible to electrons at the electrode while at the same time the electrolyte is experiencing an unstable voltage range the anode solid electrolyte interphase (SEI) and the cathode electrolyte interphase (CEI) are formed (An, Li, Du, Daniel, & Wood III, 2017; Brodd & Tagawa, 2002). The most electrolyte interphase forms during the first charge and discharge cycle due to that the anode and cathode have not formed any passivation layers that electronically insulates the electrode from electrolyte before (An et al., 2017). The SEI layer that forms on the anode is there to protect it from reacting with the electrolyte spontaneously during the cell’s normal operation (Brodd & Tagawa, 2002). The SEI layer is essential to the performance of the battery as it has an impact on its initial capacity loss, the self-discharge characteristics, the batteries cycle life, and safety (An et al., 2016). During the pre-charge process the cells are placed in a compression tray which is placed in the pre-charge chamber that can only fit one tray at a time (Ulfsparre, 2020). The compression tray compresses the cells into a certain dimension which is necessary due to the fact that during the charging process the cells have a tendency to swell (Rai, personal communication 13th of February 2024). 2.3.4.4 Tray Exchange For the precharge process as well as formation there is a need to have compression trays that keeps the cells dimension as mentioned in section 2.3.4.3. The compression trays are very expensive and therefore in many cases there isn’t an option to only use them for all the processes and therefore aging trays are also used during the other processes. There are also other trays such as electrolyte filling trays and degassing trays but this thesis is focused on just aging and compression trays due to simplification and limited published information on the subject. 2.3.4.5 Electrolyte Filling After this stage, the battery is typically filled with electrolyte again to compensate for any fluid that has been evaporated into gas in order to adjust to the correct amount. The same process as described in section 2.3.4.1 is occurring but a smaller amount of electrolyte is filled. 2.3.4.6 Soaking After the second electrolyte filling the cells are then soaked again as mentioned in the previous soaking step which can be seen in section 2.3.4.2. 17 2. Theoretical Framework 2.3.4.7 High and Room Temperature Ageing In the aging process the cells enter the first cycle of aging where they are stored in a high temperature room for a period of time. After this they are typically kept in room-temperature to cool down followed by the first formation where the cell is exposed to a charge-discharge cycle (Wood et al., 2019). 2.3.4.8 Formation The formation process consists of charging and discharging steps that are completed at slow rates to ensure that the properties of the SEI and CEI are optimized for minimum capacity fade over cycle life of the battery (Wood et al., 2019). The cells are placed in compression trays and placed in a formation chamber where the charging and discharging is occuring. 2.3.4.9 Degassing During the soaking and pre-charge some of the components of the electrolyte are reduced which results in gasses forming inside the battery cell which increases the pressure (Plumeyer et al., 2023). For both quality and safety reasons the gas is extracted from the cell during the degassing process which in the case of prismatic cells is done through a port (Plumeyer et al., 2023). 2.3.4.10 Sealing The next step is to seal the battery. The sealing process ensures that the battery cell is air and liquid tight as it is closed (Kampker et al., 2023). 2.3.4.11 Room Temperature Ageing The previous processes are usually followed by an aging step which usually takes an additional of 1-2 weeks to complete (Wood et al., 2019). This stage is necessary to check leak currents, which range from 20-50 uA/cm2 after formation to <1 uA/cm2 after the aging stages (Wood et al., 2019). The aging process also requires many electrochemical cyclers, environmental chambers and requires a large amount of factory floor space up to around 25% of the plant capacity (Wood et al., 2019). 2.3.4.12 End Of Line The last process in the manufacturing of Li-Ion cells is the end of line where the cell’s electrical characteristics are tested before the product is finished and can be placed in a battery pack (Wolter, Fauser, Bretthauer, & Roscher, 2012). 2.4 Project Models A project model is a structure of how a project should be executed and different projects requires different models to succeed (Thesing et al., 2021). The way a project is managed is influenced by many different factors, for example the rising 18 2. Theoretical Framework of new technologies and shorter time-to-market cycles (Thesing et al., 2021). There are many different kinds of project models that are being used in different industries and projects but it is important to remember that there is no model that works for all projects as all of the different models are suited for certain projects with specific defined criteria (Thesing et al., 2021). 2.4.1 Waterfall Method In traditional project models, the customer needs are first clearly defined before the product design phase can begin. Then when the product is defined the factory layout can be created to be able to produce the product that meets the customer’s demand (Thesing et al., 2021). Within the traditional project models the processes are usually carried out in a linear manner where the next step cannot begin before the previous one is finished (Putnik & Putnik, 2019). One example of a traditional project model is the waterfall method which con- tains five different phases that are conducted in a linear manner and where each of the phases requires a deliverable from the previous phase (Model, 2015). The five phases are Requirements, Design, Implementation, Verification and Maintenance which gives the user a good understanding of how to work with the project. Within each phase a list of tasks containing details of each step is prepared as well as the requirements and the success criteria (Leong, May Yee, Baitsegi, Palanisamy, & Ramasamy, 2023). The methodology can be seen in Figure 8. Figure 8: The Waterfall methods phases and order based on Model (2015). 2.4.1.1 Requirements phase The requirements that are created within the first phase regards the purpose, scope, perspective and functionalities etc. of the project (Senarath, 2021). The purpose of the phase is to outline the big picture of the project’s requirements that could be implemented in different ways (Hoory & Bottorff, 2022). 19 2. Theoretical Framework 2.4.1.2 Design phase When the requirements for the project are outlined and delivered the next phase is the design phase. Within this phase the goal is to design different solutions that can reach the requirements of the previous phase (Hoory & Bottorff, 2022). 2.4.1.3 Implementation phase During the third phase a single solution is chosen to be implemented with the help of technology and then inspect whether the designed solution is able to support the stated requirements from phase one (Hoory & Bottorff, 2022). 2.4.1.4 Verification phase In the verification phase the implementation is tested to see if it validates the re- quirements and if the solution is suitable for the project (Hoory & Bottorff, 2022). 2.4.1.5 Maintenance phase The last phase of the waterfall method is the maintenance phase in which the system created needs to be maintained (Hoory & Bottorff, 2022). The phase involves testing for errors and fixing them if they occur as well as maintaining the solution as a whole (Hoory & Bottorff, 2022). One of the negative aspects of the traditional model is that it doesn’t cope well with late changes and can produce high costs and efforts during the project (Petersen, Wohlin, & Baca, 2009). A consequence of this is that the customers current needs in the beginning of the project aren’t met at the end of it (Petersen et al., 2009). As the phases are conducted in a linear manner and each step needs to be finished before moving to the next it can also be a costly method to use if changes are needed. 2.4.2 Concurrent Engineering Concurrent engineering (CE) is a comprehensive, systematic approach to the inte- grated, concurrent design and development of complex products and the related process e.g. manufacturing and logistics.The goal of CE is to achieve higher produc- tivity, lower costs and a shorter time-to-market by taking to account downstream requirements and constraint in the design phase (Stjepandi, Wognum, & Verhagen, 2015).The adoption to CE was a reaction from manufacturing companies from tra- ditional sequential engineering in order to handle reduced product life cycles and to meet new market and customer demands (Stjepandi et al., 2015). CE consists of three basic elements according to Stjepandi et al., 2015 which incor- porates different functions e.g. product engineering, process engineering, manufac- turing planning and sourcing: • Early involvement of participants • The team approach 20 2. Theoretical Framework • Simultaneous work on different phases of product development 2.4.2.1 Set Based Concurrent Engineering Toyotas Set-Based Concurrent Engineering (SBCE) is a product development method where the developers consider a set of solutions in parallel instead of just a single solution (Sobek, Ward, & Liker, 1999). Toyotas set-based Concurrent Engineering frequently bypasses numerous practices typically regarded as crucial for the effec- tiveness of concurrent engineering (Sobek et al., 1999). Historically, design practice focused on quickly arriving at a single solution while then iterating the solution until it meets the design objectives. Subsequent itera- tions can however be very time consuming if the single solution requires substantial change and can also lead to suboptimal design both in terms of product design and manufacturing system design (Sobek et al., 1999). During the learning cycles the solutions are narrowed down with help of additional information such as customers and tests until the best solution from all aspects are found (Silvestre De Oliveira, Fidelis Peixer, Forcellini, Riso Barbosa Jr, & Lozano Cadena, 2023; Sobek et al., 1999). One of the main parts of the SBCE is the continuous communication be- tween different parts of the development team, such as the design engineers and manufacturing engineers which leads to the final idea being suitable for all teams involved (Sobek et al., 1999). Most of the projects where SBCE is used have a solid knowledge background and well established technologies and the model is supported by continuous development (Silvestre De Oliveira et al., 2023). The SCBE is performed with various principles identified by (Sobek et al., 1999), with different approaches to attain convergence in the product design and manufac- turing design process. This creates a framework in order to work in parallel in order to create flexible system design. This is illustrated in Figure 9. 2.4.2.2 Principles for set-based Concurrent Engineering This step is done in order to develop and characterize different alternatives used within the convergence process. Furthermore, this is done on two levels. First on individual projects to explore and communicate many alternatives in order to map out possibilities within the cross-functional teams. Secondly, the engineers try to capture insights and experience from previous projects on an ongoing basis in order to improve current projects. These principles are shown in Table 1, and each principle is discussed more thoroughly in the next section. The principles for set-based concurrent engineering are based on (Sobek et al., 1999). Principle 1 - Map the design space The first step in the process is to Define Feasible Regions which consist of each functional development team for instance for product engineering or production en- gineering defining feasible regions for the project from their perspective. This is done by independently determining the primary design constraints on each subsys- tem - what can and what should not be done. Typically this is based on experience, analysis, experimentation, and testing. The functional teams then create checklists 21 2. Theoretical Framework Figure 9: Principles of SBCE based on (Raudberget, 2010). that have detailed guidelines for the design in relevant areas such as functionality, manufacturability, reliability, and so on. It can also outline what is economically fea- sible as well as new emerging technologies and production methods that can suggest enhancement in quality, cost etc. These checklists are then passed along functions in order to update other departments on what is feasible and what problems have been solved previously. The next step in principle one is to explore trade-offs by designing multiple alterna- tives since merely identifying alternatives is insufficient. This is done by intelligently deciding alternatives by exploring trade-offs by designing, prototyping, and simulat- ing alternative systems and subsystems. The third and last step in principle one is to communicate sets of possibilities. This is done through the sets of possibilities, trade-offs, and implications of choosing an alternative over another then trying to communicate and understand feasible regions of other functions. This is done in order to avoid sub-optimal system solutions where one alternative is excellent for one functional team but poor for another. In order to evaluate this one option is to create a sample matrix with design alternatives along with different criteria for different options which can be seen illustrated in Figure 4. The evaluation criteria can be determined by absolutes (acceptable/unacceptable) or in terms of bounded intervals from (optimal to unacceptable) in order to find convergence. An example of the evaluation matrix can be seen in Figure 10. 22 2. Theoretical Framework Table 2: Principles for Set-based concurrent engineering (Sobek, Ward, & Liker, 1999). Principle 1: Map the design space • Define Feasible Regions • Explore trade-offs by designing multiple alternatives • Communicate sets of possibilities. Principle 2: Integrate by intersection • Look for intersections of feasible sets • Impose Minimum Constraint • Seek Conceptual Robustness Principle 3: Establish Feasibility before commitment • Narrow Sets Gradually While Increasing Detail • Stay within Sets Once Committed • Control by Managing Uncertainty at Process Gates Figure 10: Evaluation matrix based on Sobek et al., (1999) 23 2. Theoretical Framework Principle 2 - Integrate by Intersection After the steps for principle one is done it begins with principle two as the cross- functional teams now understand the considerations from the different perspectives and try to integrate sub-systems that identify solutions suited for every area. This step begins with looking for intersections of feasible sets. After communicating the different possibilities teams should look for intersections for different functions i.e. where the feasible regions overlap. This is done in order to avoid integrating independently optimized components in order to optimize the total system performance. Communication in this step is crucial and its optimal to have several cross-functional meetings in order to argue for different solutions and find solutions for the intersection of the feasible regions to find the best solution for the overall system. The next step of principle two is to Impose Minimum Constraint, in traditional engineering key decisions are made early on in order to simplify interactions among subsystems. Traditional engineering therefore aims to maximally constrain design in order to achieve the desired effect (early freeze on hard points like vehicle dimen- sions) in order to avoid confusion for the different functions. Set-based concurrent engineering however tries to in contrast impose minimum constraints needed at the time ensuring flexibility for further exploration or adjustments that can help improve the integration. The third step of principle two is to seek conceptual robustness that applies both to product design, but also robustness in market variation. This applies to creating strategies in shorter development cycles, manufacturing flexibility, and standard- ization in order to decrease design susceptibility to changes in market demand or competition. For instance, both suppliers and the manufacturer need to make pro- jections on product-design improvements for next-generation products in order to create the manufacturing system. Principle 3 - Establish Feasibility before commitment The last principle emphasizes a flexible approach to product and production devel- opment in order to enable overall system optimization and to seek an understanding of all possibilities and interactions before committing to a particular design. This is done by exploring multiple designs in parallel, and gradually converging instead of making late changes to the design. The first step of principle three is Narrow Sets Gradually While Increasing Detail which emphasizes gradually eliminating possibilities until a final solution remains instead of picking the best solution from a set. As the set grows smaller the detail in the design is increased and enables functional prototyping and simulations. This en- sures full understanding the relevant considerations before committing to a specific design. This ensures that each function narrows the respective sets with communi- cation with the other functions in order to converge to a solution that integrates with the whole system. Narrowing the options down also enables the possibility to consider the most important alternatives to a larger extent while allowing flexibility in the development. 24 2. Theoretical Framework The second step of principle three is to Stay within Sets Once Committed which means that the functional teams have to stay within the narrowing funnel when continuing to improve the current alternatives. This means that changes that cause rework to the whole system cannot be made. The third step of principle three is to Control by Managing Uncertainty at Process Gates as the need to make decisions is vital in order to progress with the design set-based concurrent engineering aims to remove certain uncertainty along with each gate. The gates represent an integrating event like a prototype of the design. This differs from traditional engineering where typically system functions hand off partial solutions to each other knowing changes result later on. Instead, set-based concurrent engineering obliges the functions to report in effect of each gate knowing that a good solution lies between the set of possibilities defined at each gate. 2.4.3 Agile Method One of the newer methods is the agile method . The agile methodology arised from the software development industry and was developed by software developers in 2001 (Rigby, Sutherland, & Takeuchi, 2016). The method is based on iterative and incremental development and has been proven to be able to solve complex issues and adapt fast to new changes (Kaur & Jajoo, 2015; Marnada, Raharjo, Hardian, & Prasetyo, 2022). The method is most suitable when the project is surrounded by changes as it is more flexible than traditional methods such as the waterfall method (Chovanova, Husovic, Babcanova, & Makysova, 2020). Within the software industry,the need of operating in dynamic and competitive en- vironments where speed, quality and cost are important the method has been seen as successful within all three areas due to the methods focus on customer demand, the responsiveness to change and the continuous iterative processes (Stavru, 2014). Since the rise of agile methodology within the software industry it has been spread- ing to different industries such as automotive, machine development and marketing (Rigby et al., 2016). But as most of the agile frameworks have been developed from software development they might not always be easily replicated in other ar- eas (Heimicke, Dühr, Krüger, Ng, & Albers, 2021). One big difference from other methods is that within the agile method documentation and planning is kept to a minimum so that flexibility and fast response to the changing environment is possi- ble (Tena., Gosar, Kuar, & Berlec, 2020). This goes in hand with the four values that the agile methodology is built upon are (1) individuals and interactions over processes and tools; (2) working software over comprehensive documentation; (3) customer collaboration over contract negotiation; and (4) responding to change over following a plan (Beck et al., 2001). As the agile methodology is about innovation it is less suitable for routine processes but as many industries are operating within dynamic environments there is a need for continuous innovation in functional processes (Rigby et al., 2016). The agile methodology is most effective if it is used where the problem that needs to be solved is complex, the possible solutions are unknown, the product requirements are unknown and where close collaboration with both customers and within the team 25 2. Theoretical Framework is possible (Rigby et al., 2016). There are different frameworks that are included within the agile methodology, the most popular is Scrum which is an iterative framework that focuses on flexibility and a holistic product development strategy (Kaur & Jajoo, 2015). The Scrum framework consists of iterative phases of planning, requirements analysis, design, execution, tests, and delivery to customers and stakeholders until the project is released (Salameh, 2014). A visualization of the method can be seen in Figure 11. Figure 11: Scrum method based on Rösiö et al., (2020). Scrum is structured in iterative cycles that are called sprints which are never longer than a certain time period and all of the sprints take place without any break in between until the final date (Deemer, Benefield, Larman, & Vodde, 2010). The people involved in the Sprint should be a cross-functional team that works together during the scrum process. The first step of a sprint is to select a number of customer requirements from a prioritized list which will be completed until the end of the sprint and they should not be changed during the period (Deemer et al., 2010). After the first sprint the people involved review the work with stakeholders and get feedback for the next sprint which then is executed in the same way as the last one (Deemer et al., 2010). The method emphasizes on using short cycles of development and continues to develop the solution (Deemer et al., 2010). Changes and modifications of the project are allowed within the agile method as the requirements are evaluated in each interaction (Salameh, 2014). The method also focuses on defining the project’s scope and requirements based on the value it brings to the customers and market share and it is therefore important to involve the customers in the iterative processes to make sure that the project is heading in the right direction (Salameh, 2014). On the other hand a disadvantage of the agile methodology is that in many cases there are low levels of documentation as short and face to face communication is preferred (Dzanic, Toroman, & Dzanic, 2022). This can lead to that information being lost and that it can create a problem of how to maintain the project later (Dzanic et al., 2022). 26 2. Theoretical Framework 2.4.4 Other Project Models As mentioned, there are many different project models that can be used for different projects. As this master thesis is time limited the three models that have been pre- sented are the ones that will be used during this project but there are other models such as Platform Based Product and Production Development and Product Evolu- tion Process and which are both two project models often used in the automotive industry that could have been applied within this thesis (Göpfert & Schulz, 2013). 2.5 Factory Layout The project of planning a factory’s layout is of importance for the future perfor- mance of the manufacturing system but it is a complex process due to the large variety of possible solutions (Klar, Glatt, Ravani, & Aurich, 2023). The most impor- tant objectives within factory layout planning is the flexibility and changeability in companies where changes are occurring frequently (Burggräf, Dannapfel, Hahn, & Preutenborbeck, 2021; Pérez-Gosende, Mula, & Díaz-Madroñero, 2021). Therefore having a sufficiently flexible layout is of high importance as by having an effective fa- cility layout the manufacturing expenses can be reduced by 10-30% (Pérez-Gosende et al., 2021; Yang & Peters, 1998). Layouts are usually designed for the initial conditions of the factory but when inter- nal or external changes such as changes in production volumes, changes in processes and technology and changes in the product re-layouts are usually necessary (Monga & Khurana, 2015). 2.5.1 Traditional Layouts Different processes can be realized by using different arrangements of machinery and equipment and the physical positioning of the production systems components is called layout (Bellgran & Säfsten, 2010). There are four traditional layouts that are further described in the following sections. 2.5.1.1 Fixed Position Layout In the fixed position layout all of the value-adding activities are performed at the same place (Bellgran & Säfsten, 2010). The layout is suitable for very large prod- ucts that are produced in a small quantity where the material and personnel are transported to the product instead of having the product travel between different stations (Bellgran & Säfsten, 2010). In Figure 12 an example of the fixed position layout is shown. 27 2. Theoretical Framework Figure 12: Example of Fixed Position Layout 2.5.1.2 Functional Layout In the functional layout all equipment of the same type are located at the same place (Bellgran & Säfsten, 2010). It is usually used when there is a large number of products produced in small volumes (Bellgran & Säfsten, 2010). The flow between the functional areas depends on which operations are needed for making a certain product which leads to flexibility as it is possible to choose the machine in the specific area that is available at that particular time (Bellgran & Säfsten, 2010). On the other hand the layout can lead to high throughput times and waiting times due to the transportations between machine groups which also increases the need of planning (Bellgran & Säfsten, 2010). In Figure 13 an example of the functional layout is shown. Figure 13: Example of Functional Layout 2.5.1.3 Cell Layout In the cell layout the different equipment and processes that are needed for making a product are located at the same place (Bellgran & Säfsten, 2010). The layout is usually used when products are produced in large volumes and to some extent 28 2. Theoretical Framework in many variants (Bellgran & Säfsten, 2010). The machines are placed in the di- rection of the flow instead of the functional similarity between them which leads to a product-oriented layout which can lead to short throughput times (Bellgran & Säfsten, 2010). The aim of using the cell layout is to create a sequential flow for as many products/parts as possible while having short set-up times to increase the flexibility (Bellgran & Säfsten, 2010). In Figure 14 an example of a cell layout is shown. Figure 14: Example of Cell Layout 2.5.1.4 Line Flow Layout In a line flow layout the machines are placed in the operation sequence for the product produced (Singh & Khanduja, 2019). The line flow layout is optimal to use when the production volume is large as it is efficient and reduces the cost per each item to produce (Singh & Khanduja, 2019). If more than one product is produced within the layout the sequence of the operations might differ which means that an alternative routing must be in place (Singh & Khanduja, 2019). The traditional assembly line can be categorized into different groups depending on their shape and number of products that are produced on that line (Kara, Gökçen, & Atasagun, 2010). The different shapes can be the traditional straight line or a U-shaped line where the traditional straight line has been most common for mass production industries while the U-shaped is a newer shape that has provided more flexibility, productivity and quality as well as it has been proven to adapt to de- manded changes quickly in comparison to the traditional shape (Kara et al., 2010). In Figure 15 an example of a line flow layout is shown. Figure 15: Example of Line Flow Layout The number of products that are produced on the line depends on if the assembly line 29 2. Theoretical Framework is a single-model or a mixed-model (Kara et al., 2010). The single-model assembly line is arranged to only produce one product type while the mixed-model produces different models of a certain product type (Kara et al., 2010). The advantages with line flow layout is the short throughput times, the simple material flow, high degree of exchangeability as well as high resource utilization (Bellgran & Säfsten, 2010). On the other hand the line flow layout is sensitive to disturbances, is inflexible as well as hard to balance (Bellgran & Säfsten, 2010). 2.5.2 Dynamic Layout Problem Manufacturing companies needs to be able to respond to changes in the requirements in order to meet the cost, time and quality and therefore the factory layout need to be planned from the beginning in a way that uncertain but significantly influential factors can be included within the system (Rogalski, 2012). Therefore there is a need of factory layouts that are flexible, modular and easy to rearrange to be able to avoid having to redesign the layout when the production requirements and system changes (Benjaafar, Heragu, & Irani, 2002). In environments where the product demand and mix are varying the layout must either be easily re-configurable or robust enough to be able to meet the demand (Lahmar & Benjaafar, 2005). According to Pillai, Hunagund, and Krishnan, 2011 there are two major ways of solving the Dynamic Layout Problem, either by using an re-configurable approach or by using an robust approach. In industries where the re-layout costs are high it can be preferable to have a layout which is robust under multiple scenarios, and even though it might not be optimal for any, it is still suitable for all of them (Benjaafar et al., 2002). The re-configurable approach is more suitable when the re-layout costs are low but the uncertainty high (Benjaafar et al., 2002). 2.5.2.1 Re-configurable approach The re-configurable approach regards the increase of flexibility, modularity or re- configurability in the layouts with the aim of reducing the cost of relayouts when the production requirements are changed (Benjaafar et al., 2002). The re-configurable approach assumes that the layout will have to be reconfigured after each time period with minimized cost while at the same time guarantee good material handling for the new time period (Benjaafar et al., 2002). Modular Layout Modular layouts can be seen as hybrid layouts that include a complex material flow that isn’t similar to one of the traditional layouts etc. functional or line flow (Benjaafar et al., 2002). Within modular layouts there are modules that consist of a group of machines that are connected to the material flow illustrated in Figure 16 (Benjaafar et al., 2002). 30 2. Theoretical Framework Figure 16: Example of a Modular Layout Design For the re-configurable approach the modular layout is a suitable choice as modular- ity can be seen as one of the most important aspects for changeability due to that if there is a need for change the modular structure will allow existing modules to be expanded or exchanged which leads to easy and cost efficient changes (Burggräf et al., 2021). Modularity in production process and layout design is characterized by organizing production processes into standardized groups that have a few or multiple strong or- ganizational ties, permitting the sequencing of machines and tooling with little loss in functionality due to each production module working as a fairly autonomous unit. Thus, if each production operation is independent of prior operations, the process can be viewed as modular (Kubota, Hsuan, & Cauchick-Miguel, 2017). Modularity is a determining factor in creating re-configurable manufacturing systems. By de- veloping a modular production system design parts of the production system can be exchanged or added without a negative effect on the whole system. This means that defined modules can be developed relatively independently of the additional system and be used at different locations or be reused in other systems like in a sister factory (Rösiö et al., 2020). As stated in Rösiö et al., (2020) there are several benefits in developing modular production systems which allows for added agility: • Increased re-configurability: The modules of the production system can rapidly be configured and modified to meet altered manufacturing requirements. • Handling complexity in the system: The technical equipment is broken down in smaller parts which allow an increased transparency and reduces the complex- ity. The modules can be developed, produced, tested and validated relatively independent from each other. • Reduces the risk of unwanted modifications: Since parts of the production- system are more independent of each other in a modular system the risk of changes in one part of the system does not alter the other modules. • Improved maintainability: The system becomes easier to diagnose, maintain and repair. 31 2. Theoretical Framework • Economical scalability: The modular system can be scaled since the develop- ment and modification can be added gradually per module. • Increased control and organization: Parallel development since multiple mod- ules can be developed simultaneously which decreases lead-time for develop- ment. The challenges with modularity is that it initially requires a large effort and capital in building the module system and developing standardized interfaces like facility planning, information system and material handling between the modules (Rösiö et al., 2020). In addition, to create a production system that is re-configurable in nature there are other aspects except modularity that is important. Other important factors include integrability, diagnosability and mobility. Integrability includes standardized interfaces so modules can be easily and cost efficiently integrated in the system. Diagnoseability which handles quality and equipment status in order to achieve a low variation of quality and an efficient setup of the machine or module. Mobility however handles in that equipment and tools are able to be transferred in a preferred process layout to optimize production or to handle new product variants (Rösiö et al., 2020). Purpose of reconfigurable approach Through the specified properties reconfigurable manufacturing system aims to con- tribute to a number of benefits in order to handle disturbances and changes in the production system according to Rösiö et al., (2020): • Capacity changes: This handles the scalability of the system in order to adjust for the capacity of changes in production volume. • Product variant changes: Re-configurable manufacturing system aims to adapt the production system and its subsystems to new variants or products. This through utilizing and re-configuring existing equipment and integrating new equipment in the system in an efficient and cost effective way. • Automation variation: A re-configurable manufacturing system can handle the degree of automation in the system by replacing manual stations to automated cells if needed or removing automation to increase flexibility. • Handling transport of equipment between departments and factories: Through modular and mobile solutions production equipment can be transported to different departments and factories. 2.5.2.2 Robust Approach When using a robust layout the assumption is that re-layouts will lead to too high costs and therefore the aim is to minimise the total material handling cost that can occur in all different periods with a single layout (Pillai et al., 2011). The layout is therefore developed with multiple scenarios in aspect when creating the layout that will be suitable for all of them (Pillai et al., 2011). 32 2. Theoretical Framework For the robust approach production data for multiple periods are needed at the initial design stage so that the layout can be confirmed to be robust over multiple periods which can be impossible in dynamic environments (Benjaafar et al., 2002). Instead the layout can include inherent features that will lead to reasonable material handling efficiency throughout all of the time periods (Benjaafar et al., 2002). The advantage of using a robust layout is that there is no need for interruptions if there are changes within the production system as it has already been accounted for. Robust Functional Layout As the aim of the robust layout is to be able fulfil a wide range of product require- ments the most common robust layout is the functional layout (Lahmar & Benjaafar, 2005). Within the functional layout resources of the same time are grouped together into functional departments and then placed depending on the material flow cost for all future planning periods within a specific planning period (Lahmar & Benjaafar, 2005). The benefits of using a functional layout is the limited commitment to having a strict flow pattern as well as providing a capacity pooling for each of the resource types (Lahmar & Benjaafar, 2005). On the other hand the material flow often becomes inefficient and the scheduling within the layout becomes complicated due to the fact that the layout is not optimized for a particular product (Lahmar & Benjaafar, 2005). In comparison to the traditional functional layout the robust layout is constructed for multiple planning periods, while the traditional is created for a certain time-frame. Distributed layout A distributed layout consists of functional departments that are divided into smaller sub departments that are distributed in strategic places (Lahmar & Benjaafar, 2005). The aim of the division and distribution is to be able to meet future changes in the product mix, the product routing and the demanded volume (Lahmar & Benjaafar, 2005). As the distribution of the sub departments into different places throughout the plant could lead to increased accessibility to the different departments and that could result in more efficient flows which in turn leads to that need of re-layout could be extinguished even if production requirements change (Lahmar & Benjaafar, 2005). In Figure 17 two different distributed layouts are shown. 33 2. Theoretical Framework Figure 17: Distributed layouts based on Lahmar & Benjaafar (2005). The use of distributed layouts are most suitable in production systems where the demand variability is high or where the product variety is low (Lahmar & Benjaafar, 2005). One of the problems when creating a distributed layout is how the sub departments should be created and how many of each that are needed (Benjaafar et al., 2002). The duplication of departments might lead to increased flexibility within the lay- out but it comes with the cost of needing duplications of operators and auxiliary resources (Benjaafar et al., 2002). 34 3 Methodology This chapter regards the project’s structure and the methodology that has been used in order to conduct this project. Further, it aimed to justify the research method that was used and why it was suitable for this project. 3.1 Research Design The research was designed to be able to fulfill the aim of the study, which was to find an optimized factory layout for battery cell factories and help pace up the industry. The research was carried out with an inductive work process where the research had its basis in developing a general standardized workflow in enabling battery factor cell design for unknown parameters (Patel & Davidson, 2019). As the research questions were explorative, the research process needed to be iterative as shown in Figure 18. Furthermore, the structure of the project was a combination of five different phases as illustrated in Figure 18 at the next page. 35 3. Methodology Figure 18: Research Design 36 3. Methodology 3.2 Research tools The research tools that were used in this thesis was a qualitative study based on interviews with relevant people within the field and literature studies. A qualitative approach was chosen as it provides a deep insight to the subject which helped to answer the research question (Tenny, Brannan, & Brannan, 2023; Venkatesh, Brown, & Sullivan, 2023). The decision to use a qualitative study was further supported by the consideration of the amount of available numerical data. Given that the battery cell industry is relatively new, there was a limited amount of published numerical data. This scarcity of quantitative data and the lack of measurable phenomena made a qualitative approach more appropriate (Patel & Davidson, 2019). 3.2.1 Interviews As the study was to an extent relying on interviews the main access to interviewees were through NOVO Energy, either employees of the company or other contacts they had. The decision to have interviews rather than surveys was because the project aims to solve a complex problem and a detailed understanding of the collected data were necessary which surveys could not support (Denscombe, 2014). The interviews were in most cases semi-structured to let the interviewee speak more widely and develop new ideas (Denscombe, 2014). Therefore the interviews were conducted with a low level of standardization and semi-level of structure (Patel & Davidson, 2019). The use of semi-structured interviews also let the interviews de- velop as more knowledge was gained and to continuously develop the data collection to become more detailed. In some cases an in depth interview was used when there was a specific topic that needed further investigation (Kvale & Brinkmann, 2014). The transcribed interviews were also checked with the interviewee to make sure that he or she was comfortable with his or her answers. The data that was gathered from the interviews were used as the base for the project and helped with the idea generation part. As the industry is relatively new and published information was limited, the knowledge of the interviewees was of crucial importance to gain a deeper understanding to be able to solve the complex problem. 3.2.2 Literature review In addition, a literature review was carried out in advance and alongside with the interviews in order to gain valuable knowledge and provide validation for the inter- views. The main purpose of the literature review was to gain a general understanding of different project models, and different processes as well as how flexibility can be increased in factory layout design within battery cell production. The gained in- formation from the literature review was used to form a theoretical framework and support the creation of interview questions. The data gained from the literature was also used to make sure that the data that was gathered from the interviews were reliable by using triangulation with other interviews and the collected data 37 3. Methodology (Denscombe, 2014). In order to ensure a high standard of the literature review the researched papers had to be peer-reviewed (Patel & Davidson, 2019). The literature review was carried out by searching academic databases (Google Scholar, Chalmers Library and Scopus) in order to find relevant articles, publications, industry journals, and case studies on topics related to the scope. Printed books were also used in cases where it was deemed necessary. The main literature that was searched for regarded the different production processes, factory layouts, project models, and battery production. Keywords that were used in the search was Li-Ion battery production, Gigafactory, Factory layout, Cell Finishing, Project models etc. 3.2.3 Qualitative analysis Further, a qualitative analysis was performed in order to evaluate results of the re- search design. An important aspect of the literature review as well as the interviews was to analyze the gathered data ongoingly, since it provided useful insights that were important for further steps (Patel & Davidson, 2019). The main idea was that the iterative qualitative analysis of the material would provide a pathway for further investigation and ideas and in extent enrich the iterative process in order to design the solution. The qualitative analysis also aimed to problematize variations and homogeneity in the collected data as well as the relations between the parts in the material that is aimed to constitute the entirety of the project (Patel & Davidson, 2019). 3.3 Phase 1 The first phase regarded the research introduction, and the main focus was on the planning report as well as onboarding within the company and a basic understanding of the processes and the product. 3.4 Phase 2 The second phase was a pre-study of other mature industries and investigated how their project models looked. This was done by a literature study that regarded project models as a general topic and then focusing on how they work in different industries. After the literature study, interviews were conducted with employees at NOVO Energy that had previously been working at other companies and had experiences with different project models. The interviews took place during a 30 minute period and recorded on Microsoft Teams to ensure that the results could be transcribed and validated after the interviews. The interviewees and their previous work industry is presented in Table 3 but due to confidentially their real names are masked. The interview questions can be seen in Appendix A. This phase was a way to understand how other industries had been working with projects and helped generate ideas for the battery industry. The generated results 38 3. Methodology Table 3: interviewee in the Pre-Study interviewee Previous Industry A Automotive B Process C Battery D Automotive from the interviews were analyzed and compared to existing literature to make sure that the information was correct and validated. From this phase, three different project models were discovered and further investi- gated with the help of a literature study to create three models that could be used to generate factory layouts. 3.5 Phase 3 The third phase regarded the different production processes within Li-ion battery manufacturing and investigated how the different processes worked in-depth and how they could become more flexible. This was done by first doing a literature study as well as reading technical reports regarding the processes before holding semi-structured interviews with process engineers within the specific area. The interviewees are presented in Table 4 together with their current role at NOVO Energy. Due to confidentiality, the interviewees’ real names are masked. Table 4: interviewee in the third phase interviewee Current Role A Senior Technical Project Manager - Formation and Ageing B Process Engineer - Formation and Ageing C Manager Engineer - Factory Layout & Simulation D Material Flow Engineer The interviews were conducted during an hour each time and recorded on Microsoft Teams to ensure that the answers could be transcribed and validated after the interviews. The information that was gathered from the interviews were analyzed in comparison to the found literature to make sure that the information was validated. The interview questions can be seen in Appendix B. 3.6 Phase 4 The fourth phase regarded answering the research questions with help from the pre- vious phases. When knowledge of how the different production steps could become more flexible and how other industries were working with projects, the main focus 39 3. Methodology was on how the factory layout could become more flexible and how that would affect the other areas within the project. As this master thesis aimed at finding a solution to how to optimize the layout of a battery cell factory for the unknown