Quality Assurance in Light Incontinence Products Investigating Super Absorbent Polymers and Pulp Across the Phases of the Development Chain Master Thesis Project in Product Development 2025 Ida Kalentun Louise Selin Department of Industrial and Material Science Division of Product Development Chalmers University of Technology Gothenburg, June 13, 2025 Master Thesis 2025 Quality Assurance in Light Incontinence Products Investigating Super Absorbent Polymers and Pulp Across the Phases of the Development Chain Ida Kalentun Louise Selin Department of Industrial and Materials Science Division of Product Development CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2025 Quality Assurance in Light Incontinence Products Investigating Super Absorbent Polymers and Pulp Across the Phases of the Develop- ment Chain ©Ida Kalentun 2025 ©Louise Selin 2025 Examiner and supervisor: Peter Hammersberg, Department of Industrial and Mate- rial Science Master thesis 2025 Department of Industrial and Materials Science Division of Product Development Chalmers University of Technology SE-412 96 Gothenburg Phone +46 31 772 1000 Cover Page Image Source: Allegro Medical Website, 2025. Licensed under Creative Commons. Retrieved from: https://www.allegromedical.com/incontinence/ Gothenburg, Sweden 2025 iii Quality Assurance in Light Incontinence Products Investigating Super Absorbent Polymers and Pulp Across the Phases of the Develop- ment Chain IDA KALENTUN LOUISE SELIN Department of Industrial and Material Science Chalmers University of Technology Abstract The master’s thesis investigates the procedure of specifying and monitoring the key materials Super Absorbent Polymers (SAP) and pulp in light incontinence products at Essity. The project focuses on four phases within the value chain from product development, design verification, process validation to running production. The aim was to evaluate existing procedures to identify potential gaps and opportunities for improvement. The research approach was based on the DMAIC framework originating from the Six Sigma philosophy, integrating qualitative methods and data from semi-structured interviews, workshops with quantitative data from laboratory testing on prototype products. Statistical tools were also implemented to analyze variation and correlation and to perform capability studies. The study identified that variation in SAP and pulp amount and distribution affects the product performance. Due to process-related influences, such as environmental conditions and technical factors, variation will always be present in the production system. Therefore, the importance of robust design strategies was highlighted together with the value of proper monitoring of SAP and pulp amount and distribution across the value chain. Control measures based on gathered information were proposed to enhance process control and stability, emphasizing the importance of centering processes around target values rather than only specification limits. While the limited sample size and test scope restrict broad generalization, the results provide a solid foundation for further investigation and practical improvements when it comes to quality assurance. Additionally, the research highlights the value statistical tools can bring within the organization to sustain quality improvements. Overall, this thesis contributes to a deeper understanding of SAP and pulp management in hygiene products and supports Essity’s efforts to optimize product consistency and customer satisfaction through data-driven quality control. Keywords: Quality Assurance, Super Absorbent Polymers (SAP), Pulp, Statistical Tools, Light Incontinence Products, Requirement Setting iv Acknowledgements We would like to sincerely thank Essity, for the opportunity to learn and collaborate during our time at the company. A special thanks to our supervisors Katarina Brantin and Gunilla Kindberg for their continous support and valuable contributions throughout the project. We also extend our gratitude to Peter Hammersberg, our supervisor from Chalmers University of Technology, for his guidance that has greatly supported or work. June 2025 Ida Kalentun Louise Selin v Contents List of Figures ix List of Tables xi 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Problem Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.3 Purpose and Research Questions . . . . . . . . . . . . . . . . . . . . . . . . 2 1.4 Delimitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 Theoretical Framework 4 2.1 Urine Incontinence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.2 Key Product Functions for Incontinence Products . . . . . . . . . . . . . . 5 2.3 Key Materials in Incontinence Products . . . . . . . . . . . . . . . . . . . . 5 2.3.1 Super Absorbent Polymers . . . . . . . . . . . . . . . . . . . . . . . 5 2.3.2 Fluff Pulp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.4 Quality Assurance Theories . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.4.1 Deming’s Philosophy of Quality Improvement . . . . . . . . . . . . 7 2.4.2 Robust Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.4.3 Shewhart Control Charts . . . . . . . . . . . . . . . . . . . . . . . . 8 2.4.4 Process Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3 Methodology 10 3.1 Research Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 3.1.1 System Perspective on the Value Chain . . . . . . . . . . . . . . . . 10 3.1.2 Application of the DMAIC Framework . . . . . . . . . . . . . . . . 10 3.1.3 Process and Project Layout . . . . . . . . . . . . . . . . . . . . . . 11 3.2 Research Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3.2.1 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3.2.1.1 Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3.2.1.2 Workshop . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.2.1.3 Internal and External Searches . . . . . . . . . . . . . . . 13 3.2.2 Laboratory Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 3.2.3 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 3.2.3.1 Qualitative Analysis . . . . . . . . . . . . . . . . . . . . . 15 3.2.3.2 Quantitative Analysis . . . . . . . . . . . . . . . . . . . . 16 4 Result 18 4.1 Current Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 4.1.1 The Product Development Phase . . . . . . . . . . . . . . . . . . . 19 4.1.2 The Design Verification Phase . . . . . . . . . . . . . . . . . . . . . 20 4.1.3 The Process Validation Phase . . . . . . . . . . . . . . . . . . . . . 21 4.1.4 Running Production . . . . . . . . . . . . . . . . . . . . . . . . . . 22 4.2 The Product Development Process . . . . . . . . . . . . . . . . . . . . . . 24 4.2.1 Identification of Key Parameters for Incontinence Products . . . . . 24 4.2.2 Quality function deployment . . . . . . . . . . . . . . . . . . . . . . 25 vi 4.2.3 Correlations Between Design Parameters and Technical Requirements 27 4.2.3.1 SAP Amount Correlations . . . . . . . . . . . . . . . . . . 27 4.2.3.2 Thickness Correlations . . . . . . . . . . . . . . . . . . . . 28 4.2.3.3 Interaction Effects Between SAP Amount and Thickness . 29 4.2.4 Precision of Measurement Systems . . . . . . . . . . . . . . . . . . 32 4.2.4.1 Weight and Thickness . . . . . . . . . . . . . . . . . . . . 32 4.2.4.2 SAP Amount Methods . . . . . . . . . . . . . . . . . . . . 34 4.2.4.3 Material Distribution Methods . . . . . . . . . . . . . . . 35 4.2.5 Target and Tolerance Design in Specifications . . . . . . . . . . . . 36 4.2.5.1 Variation Requirements . . . . . . . . . . . . . . . . . . . 37 4.2.5.2 Centering Requirements . . . . . . . . . . . . . . . . . . . 37 4.2.6 Design FMEA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 4.3 The Design Verification Process . . . . . . . . . . . . . . . . . . . . . . . . 39 4.3.1 Overview of the Evaluated Parameters . . . . . . . . . . . . . . . . 39 4.3.2 Evaluation of Variation and Centering Requirements for Product Concept A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.3.2.1 Product Weight for Product Concept A . . . . . . . . . . 40 4.3.2.2 SAP Amount for Product Concept A . . . . . . . . . . . . 41 4.3.2.3 Thickness for Product Concept A . . . . . . . . . . . . . . 44 4.3.2.4 Material Distribution for Product Concept A . . . . . . . 45 4.3.2.5 Rewet for Product Concept A . . . . . . . . . . . . . . . . 47 4.3.3 Evaluation of Variation and Centering Requirements for Concept B 48 4.3.3.1 Product Weight for Product Concept B . . . . . . . . . . 49 4.3.3.2 SAP Amount for Product Concept B . . . . . . . . . . . . 50 4.3.3.3 Thickness for Product Concept B . . . . . . . . . . . . . . 52 4.3.3.4 Material Distribution for Product Concept B . . . . . . . 53 4.4 The Process Validation Process . . . . . . . . . . . . . . . . . . . . . . . . 55 4.4.1 Current Measurement Method for SAP and Pulp . . . . . . . . . . 55 4.4.2 Root-Cause Analysis Using a Fishbone Diagram . . . . . . . . . . . 56 4.5 Running Production and Quality Control . . . . . . . . . . . . . . . . . . . 57 4.5.1 The Manufacturing Process of Incontinence Products . . . . . . . . 57 4.5.2 Current Quality Control Methodologies in Running Production . . 58 4.5.3 Production Data Overview . . . . . . . . . . . . . . . . . . . . . . . 59 5 Analysis 62 5.1 Analysis of the Current Situation . . . . . . . . . . . . . . . . . . . . . . . 62 5.2 Evaluation of the Product Development Process . . . . . . . . . . . . . . . 63 5.3 Assessment of the Design Verification Process . . . . . . . . . . . . . . . . 64 5.4 Assessment of the Process Validation Process . . . . . . . . . . . . . . . . . 66 5.5 Assessment of Running Production Process . . . . . . . . . . . . . . . . . . 66 5.6 Opportunities for Improvement . . . . . . . . . . . . . . . . . . . . . . . . 67 5.7 Quality Control Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 5.7.1 Laboratory-based Product Testing . . . . . . . . . . . . . . . . . . 68 5.7.2 Integrating and Training in Statistical Methods . . . . . . . . . . . 69 5.7.3 Accelerating the Implementation of Centering Requirements . . . . 69 vii 6 Discussion 71 6.1 Discussion of Result Compared to Research Questions . . . . . . . . . . . . 71 6.2 Discussion of Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 6.3 Ethical, ecological and social aspects . . . . . . . . . . . . . . . . . . . . . 72 6.4 Further investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 7 Conclusion 75 viii List of Figures 2.1 An example of the material SAP after it has absorbed liquid [13] . . . . . . 6 2.2 Visualization of Control Chart (Own Illustration) . . . . . . . . . . . . . . 8 3.1 An overview of the report structure in relation to the projects process (Own Illustration) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 4.1 Schematic overview of the value chain (Own Illustration) . . . . . . . . . . 19 4.2 Process map of the product development phase (Own Illustration) . . . . . 20 4.3 Process map of the design verification phase (Own Illustration) . . . . . . . 21 4.4 Process map of the process validation phase (Own Illustration) . . . . . . . 22 4.5 Process map of running production (Own Illustration) . . . . . . . . . . . 23 4.6 Key parameters for incontinence products . . . . . . . . . . . . . . . . . . 25 4.7 Importance rating for the QFD . . . . . . . . . . . . . . . . . . . . . . . . 25 4.8 QFD 1 - Customer needs and Technical requirements . . . . . . . . . . . . 26 4.9 QFD 2 - Technical requirements and Design parameters . . . . . . . . . . . 27 4.10 Correlation between SAP amount and Rewet . . . . . . . . . . . . . . . . . 28 4.11 Correlation between SAP amount and Inlet time . . . . . . . . . . . . . . . 28 4.12 Correlation between Thickness and Rewet . . . . . . . . . . . . . . . . . . 29 4.13 Correlation between Thickness and Inlet time . . . . . . . . . . . . . . . . 29 4.14 Interaction effects for Rewet . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.15 Interaction effects for Inlet 1 . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.16 Interaction effects for Inlet 2 . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.17 Measurement System Analysis for Weight . . . . . . . . . . . . . . . . . . . 33 4.18 Measurement System Analysis for Thickness . . . . . . . . . . . . . . . . . 33 4.19 Time series plots for Thickness . . . . . . . . . . . . . . . . . . . . . . . . 34 4.20 Individual value plot for SAP amount methods . . . . . . . . . . . . . . . . 35 4.21 Schematic figure of variation and centering requirements (Own Illustration) 37 4.22 DFMEA (Design Failure Mode and Effects Analysis) . . . . . . . . . . . . 38 4.23 An overview of the evaluated parameters . . . . . . . . . . . . . . . . . . . 39 4.24 Specification for Concept A . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.25 A process capability report of the Product Weight for Product Concept A . 41 4.26 A histogram of the SAP Amount for Product Concept A . . . . . . . . . . 42 4.27 Equivalence test of the SAP amount for Product Concept A . . . . . . . . 42 4.28 Correlation between SAP amount and Product Weight for Product Concept A 43 4.29 A histogram of Thickness for Product Concept A . . . . . . . . . . . . . . 44 4.30 Capability analysis of Thickness for Concept A . . . . . . . . . . . . . . . 45 4.31 Visual imaging of three samples from Product Concept A . . . . . . . . . . 46 4.32 Weight of a smaller area of Concept A . . . . . . . . . . . . . . . . . . . . 46 4.33 SAP amount in a smaller area of Concept A . . . . . . . . . . . . . . . . . 47 4.34 A histogram of Rewet for Product Concept A . . . . . . . . . . . . . . . . 47 4.35 Capability analysis of Rewet for Concept A . . . . . . . . . . . . . . . . . . 48 4.36 Specification for Concept B . . . . . . . . . . . . . . . . . . . . . . . . . . 49 4.37 Process capability report of Weight for Product Concept B . . . . . . . . . 50 4.38 Process capability report of the SAP amount for Product Concept B . . . 51 4.39 A correlation study between SAP amount and Product Weight for Product Concept B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.40 A histogram of Thickness for Product Concept B . . . . . . . . . . . . . . 52 ix 4.41 Visual imaging of three samples from Product Concept B . . . . . . . . . . 53 4.42 Weight of a smaller area for Product Concept B . . . . . . . . . . . . . . . 54 4.43 SAP Amount in a smaller area for Product Concept B . . . . . . . . . . . 54 4.44 Root-cause analysis in relation to process validation (Own Illustration) . . 56 4.45 An overview of the manufacturing process (Own Illustration) . . . . . . . . 58 4.46 A histogram presenting the Weight measured in production for a span of two years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 4.47 A histogram presenting the Thickness measured in production for Year 1 and Year 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 4.48 A control chart presenting the Thickness measured in production for Year 1 and Year 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 5.1 Table of control measure related to laboratory-based product testing . . . . 68 5.2 Table of control measure related to integrating and training in statistical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.3 Table of control measure related to accelerating the use of centering require- ments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 x List of Tables 3.1 GR&R Acceptance Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . 16 4.1 Evaluation of Distribution method 1 . . . . . . . . . . . . . . . . . . . . . 36 4.2 Tolerance intervals for SAP Amount - Concept A . . . . . . . . . . . . . . 43 4.3 Tolerance Intervals for Thickness - Concept A . . . . . . . . . . . . . . . . 45 4.4 Tolerance intervals for Rewet - Concept A . . . . . . . . . . . . . . . . . . 48 4.5 Tolerance intervals for SAP Amount - Concept B . . . . . . . . . . . . . . 51 xi 1 Introduction This chapter provides an overview of the context of the project, outlining the background and the problem definition. It introduces the purpose and research questions for the project together with the delimitation. 1.1 Background Essity is a company focused on hygiene and health, active in three different business categories [1]. These categories are Health and Medical, Consumer Goods, and Professional Hygiene. Essity develops, produces, markets and sells products such as incontinence protection, baby diapers, sanitary pads and toilet paper. Well-known brands are TENA, Libero, Libresse and Tork. Essity’s purpose is expressed as “Breaking Barriers to Well- being”, through creating increased awareness of the importance of hygiene and health and how it links to well-being [2]. Furthermore, Essity strives to improve care and break silence around areas such as menstruation and incontinence. Incontinence is defined by the International Continence Society as ”the complaint of any involuntary leakage of urine” [3]. Moreover, incontinence is one of the greatest public health disorders, primarily due to an aging population. On the other hand, it is not only the elderly who can be affected by the problem, it can affect anybody at any age. In addition, the severity of the problem can vary, from small amounts to large amounts of urine leakage. Product quality is critical for ensuring the functionality, user experience, and sustainability of incontinence products [4]. A key factor affecting this is the performance of the absorbent core, which is primarily made up of two materials: Super Absorbent Polymers (SAP) and Pulp [5]. SAP is a material that is capable of absorbing and retaining large amounts of liquid, while pulp helps to distribute the liquid throughout the core. These materials have a direct impact on many of the product’s essential functions, such as quick absorption, leakage security overall discretion during use. Since the amount and distribution of SAP and pulp are known to influence product performance, monitoring these parameters throughout the value chain may contribute to maintaining consistent quality. 1.2 Problem Formulation In this project, the focus is placed on products meant for light incontinence, typically involving smaller volumes of leakage. When developing light incontinence products, it is important to understand how SAP and pulp, the two main materials of the absorbent core, are specified, verified and controlled throughout the value chain, from requirement setting, design verification, process validation to production. These materials are closely connected in the way they interact to fulfill the core´s functional requirements. Understanding how 1 variations in their amounts and distribution may impact product quality is important. moreover, maintaining control over these parameters throughout the value chain is crucial to ensure consistent functionality and performance for light incontinence products. 1.3 Purpose and Research Questions The purpose of this master thesis was to investigate four phases of the value chain and understand how two of the key materials in light incontinence products, SAP and pulp, are specified, verified and controlled along this chain with regards to amounts and distribution. By investigating existing quality assurance procedures, the thesis aimed to identify potential gaps and opportunities for improving product consistency. Furthermore, the project investigated the impact of key design parameters on product functions, with the aim of understanding how to develop reliable and robust specifications. The study also explored how statistical analysis could support more robust target setting and process control along the value chain. The purpose form the basis for the following research questions that the project aimed to answer: • What impact do the amount and distribution of SAP and pulp have on the product functions and overall quality? • What are the main challenges associated with controlling SAP and pulp throughout the value chain? • How do variation and centering requirements influence process control and product performance? • In what ways could the integration of statistical tools and methods influence and support development and quality control processes? 1.4 Delimitations The project had several delimitations that should be considered when interpreting the results. The collection of customer needs and the process of translating these into requirements and specifications was not included in this project. However, existing customer needs were evaluated to understand how they relate to the current requirements, with focus placed on how these requirements are maintained and controlled in the current processes. Emphasis was given to examining whether the amounts and variations of SAP and pulp ensured the required product performance and compliance with established initial specifications. Improving production systems and machines was not within the scope of the project. Instead, the focus was mainly on understanding and evaluating variations that arise from the production process. The project specifically focused on material parameters related to SAP and pulp. Other components of incontinence products, such as the topsheet, backsheet, and adhesives, were excluded from the analysis. Therefore, any potential variations or interactions involving these elements fell outside the scope of this work. 2 Due to time constraints, only a limited number of laboratory tests were performed. Moreover, testing was conducted on conceptual or simplified prototypes rather than on fully developed products available on the market. As a result, the findings may not fully reflect real-use conditions but can instead be regarded as indicative. Additionally, the work was conducted over five months during the spring of 2025, and completed by mid-June. The research was carried out at Essity, under supervision from both Essity and Chalmers University of Technology. The study was also subject to confidentiality constraints, specific data and exact numerical values are not disclosed. 3 2 Theoretical Framework This chapter presents the theoretical foundation of the project, addressing key aspects of incontinence as a condition and the structure and functions of incontinence products. Additionally, it includes relevant quality assurance theories to support the evaluation of a products overall quality. 2.1 Urine Incontinence Urine incontinence is defined as the loss of bladder control that results in unintentional urine leakage [3]. Considering that the population trends to becoming older, issues with urinary incontinence will continue to grow. However, while incontinence may be associated with aging, it is a condition that can affect anybody across all stages of life. The problems can range from occasionally leaking urine when sneezing or coughing to having an urge to urinate that is so sudden that the person doesn’t make it to the toilet in time [6]. There is also a large proportion of elderly people in nursing homes who are unable to go to the toilet at all. Urinary incontinence can be divided into different categories and types depending on the nature of the problem. Some of these are stress incontinence, urge incontinence, overflow incontinence, functional incontinence, and mixed incontinence. Below, these types are described and examples of individuals with these problems are presented. • Stress incontinence. Urine leakage due to pressure on the bladder. This can occur when an individual sneezes, coughs, laughs, or lifts something heavy. This type of incontinence can affect young women active in sports, women after childbirth, or men with prostate issues. • Urge incontinence. A sudden and intense urge to urinate that leads to leakage. A person with this type often needs to urinate frequently, including during the night. • Overflow incontinence. Urine leakage through continuous dripping due to a bladder that never fully empties. • Functional incontinence. A physical or mental condition that prevents an individual from reaching the toilet in time. • Mixed incontinence. A person experiencing multiple types of incontinence, often related to both stress and urge incontinence. Generally, urinary incontinence is a condition that affects quality of life. There are several ways to get help with these problems, with both medical treatments but also with other solutions such as protections [3]. There are products such as liners, pads, pants and heavy incontinence products that can be used in daily life according to needs. The problem is often considered taboo, which means that the proportion of the population with the 4 problem is underestimated. Among adult women in the United States, over 60% reported some form of incontinence, with stress incontinence being the most common. For men, the percentage was around 30%. 2.2 Key Product Functions for Incontinence Products Urinary incontinence, as previously explained, poses numerous challenges for individuals living with it, affecting their quality of life [3]. Therefore, incontinence products play a crucial role in supporting people’s daily lives. To fulfill their purpose, these products must meet certain functions. Leakage security and absorption of liquid is the most crucial function of an incontinence product [7]. This process can be divided into several sub-functions, such as Inlet time, to capture the liquid fast and absorption capacity, to keep the liquid inside the product. When these functions are met, the product is considered leakage secure, ensuring that no liquid will soil the user´s clothes or the surroundings. The dryness of a product is important for ensuring comfort and skin health [7]. Dryness means that liquid is stored and retained inside the product, creating a dry surface against the user´s skin. One key function is Rewet, which ensures that liquid remains safely in the structure even when pressure is applied, such as during body movements or when the user sits down on the product. Discretion is a key concern for users, as many wish to avoid revealing that they are wearing an incontinence product [8]. This leads consumers to prefer thinner, less bulky products that are less noticeable under clothing. 2.3 Key Materials in Incontinence Products Products designed for incontinence typically consist of several layers, a top sheet (the layer in contact with the skin), an acquisition and distribution layer (ADL), an absorbent core and a back sheet (which acts as a barrier layer) [9]. Usually, an incontinence product is composed of the following materials: fluff pulp, SAP, nonwoven material, adhesives and elastics [10]. The absorbent core consists of the two materials SAP and pulp, and plays a vital role in ensuring that the liquid is absorbed, distributed and retained within the product. These materials are further explained in the following sections. 2.3.1 Super Absorbent Polymers One of the main benefits of SAP is its capability to absorb and retain saline solution, up to 100 times their own weight [11]. The material is a cross-linked polymer, typically made from sodium polyacrylate. In its dry state, the material exists as small granules with tightly coiled polymer chains. Upon contact with liquid, they chains uncoil, expanding the molecular network, allowing liquid to be stored within the structure [12]. This reaction turns the material into a gel, effectively locking the liquid in place, providing security 5 against leakage. Figure 2.1 shows the gel-like state that SAP transforms into after absorbing liquid[13]. Figure 2.1: An example of the material SAP after it has absorbed liquid [13] SAP is a material that offers several important benefits and is widely used in hygiene products such as diapers, feminine pads and incontinence products [12]. Not only does it absorb and draw the liquid away from the surface but it can also retain the fluid under pressure, such as when sitting down [5]. This helps to ensure comfort for the use but also to prevent skin irritation and infection caused by long exposure to moisture [12]. The SAP-to-pulp ratio in the absorbent core is an important factor for product performance. Pulp typically makes up the majority of the mixture in the core, however, as incontinence products trend toward becoming thinner, the proportion of SAP is gradually increasing, helping to maintain high absorption performance in a more compact form [5]. 2.3.2 Fluff Pulp The other major component in the absorbent core is pulp, also known as fluff pulp, which is made from cellulose fibers made from wood or other fibrous materials [14]. Pulp has the ability to quickly absorb liquid, enabling fast fluid intake before the SAP captures and locks the liquid in. It is highly hydrophilic, meaning it attracts and absorbs liquid, and helps distribute it throughout the core via wicking and capillary forces along the fiber network. Additionally, pulp contributes to the structural stability of the core, helping the core to stay in place and prevent cracking or shifting during use [5]. Pulp is composed of a network of cellulose fibers with airspace between them, giving the material a low density [15]. This structure contributes to a soft cushioning feel as well as adding Thickness to the product [16]. However, this means that the material is highly sensitive to moisture. When exposed to humidity, it can begin to swell, affecting its absorption capacity [17]. Therefore, maintaining appropriate levels of moisture both during manufacturing and storage is important to preserve the pulps absorbent performance. 6 2.4 Quality Assurance Theories Quality assurance as concept refers to a systematic approach to reduce defects and to address faults across the whole value chain before they occur to improve overall product quality [18]. This theoretical framework integrates quality assurance theories to establish a foundation for analyzing and improving product development systems. 2.4.1 Deming’s Philosophy of Quality Improvement Dr. W. Edward Deming is known for his theories related to quality thinking and especially his systematic point of view [19]. He emphasizes the importance of optimizing the entire system instead of isolated parts, since individual components interact within a larger system. A critical element in achieving alignment across all parts of a system is the use of operational definitions. Deming emphasizes that these definitions, such as “acceptable” or “on time” need to be clear and well understood. Without this, data collected from different facilities or parts of the system may be incomparable or misleading, leading to inconsistent outcomes. Another key aspect for improving a system is understanding the causes of variation, some are common, meaning they are inherent in the system and some are special, meaning they are unusual or unexpected [19] . Common causes of variation are usually stable and predictable and can be due to inadequate testing, faulty design, wrong specifications, misunderstandings and uninformed workers. A possible misstep is treating a special cause of variation as a common one. For instance, replacing machinery or a staff member due to a special cause may introduce new disruption to the system, further decreasing the quality outcome. To improve the process Deming also emphasizes the need for precise definitions of specifications. The definitions needs to increase communication between involved parties and have the same meaning for all. Deming also identified barriers that can prevent effective quality improvement[19]. One challenge is the misunderstanding or lack of use of statistics. Managers may try to improve a process without knowing what they don’t know, trying to interpret data without understanding the context. He argues that statistical expertise is essential and that decisions must be guided through data and structured experimentation. There are also human factors to consider, such as the resilience from managers to let go of control over the process, as well as the reluctance among workers to adapt to a new system. 2.4.2 Robust Design Robust design focuses on designing systems, processes and products that are insensitive to variation [20]. The idea is to build in quality thinking into the design stage, rather than relying on correction after production. Robust design is structured into three main stages, system design, parameter design and tolerance design. System design serves as the foundation and includes customers requirements and evaluates the products intended func- tions, considering possible trade-offs. Parameter design focuses on identifying controllable factors that can be adjusted in order to improve quality. It’s about determining parameter settings that can make the product insensitive to noise factors, such as environmental factors, machine error or production variation [21]. The next step is tolerance design and 7 focuses on specifying how much variation is accepted in production without compromising product quality [20]. Too tight tolerance intervals can drive up costs, while loose tolerance intervals may cause decreased quality. In robust design, causes of variation can be due to factors including equipment differences, material variation and environmental conditions [20]. These are factors that might be hard to control during manufacturing. In contrast, controllable factors are variables that can be adjusted during design or production, such as material type or machine settings. A key concept in robust design stems from Taguchi’s method which is structured experimentation. This involves a planned series of experiments to analyze how different combinations of design parameters and noise factors influence the product´s performance. By seeing what parameters have the biggest impact on quality and determining what levels result in the most stable performance, engineers can adjust the design to achieve a more reliable product. 2.4.3 Shewhart Control Charts Control chart is a common tool applied in statistical process control, developed by Walter A . Shewhart in the 1920s [22]. It is used to monitor production data over time and therefore the stability of a process. As presented in Figure 2.2, a control charts typically consists of a time-order plot with a center line and an upper (UCL) and lower (LCL) control limit, typically falling within ±3 standard deviations from the process mean. The theory is based on differentiating between common causes of variation and special causes of variation. Common causes of variation are expected by the system and are the variation of the data points that still falls within the established control limits. Points falling outside the control limits indicate more special causes of variation, which should be investigated. A special cause of variation could indicate a process change, which is possibly because of assignable causes, meaning that corrective actions may be taken. Figure 2.2: Visualization of Control Chart (Own Illustration) 8 2.4.4 Process Capability Process capability is a statistical measure that is used to assess how well a process meets specified limits by evaluating its variability [23]. It is used to help organizations determine if a process can consistently produce acceptable outputs within its upper and lower specification limits, (USL) and (LSL) [24]. Two commonly used metrics to evaluate this are Process Performance (Pp) and Process Performance Index (Ppk). The Pp measures the potential capability for the process by comparing the width of the specification limits to the spread of the process. It does however not take into account whether the process mean is centered between these limits [23]. The Ppk value, on the other hand, does consider the location of the process mean relative to the specification limits, as can be seen in Equation 1[23]. Meaning that the Ppk value considers both how well the process is centered and the variation requirements when evaluating the process capability. If the process mean shifts towards one of the specification limits, the Ppk value will decrease, meaning that there is a decrease in process capability [24]. Ppk = min ( x̄− LSL 3S , USL− x̄ 3S ) (1) x̄ = Mean S = Standard deviation In practice, Ppk, together with control charts are used to monitor, control and improve processes [24]. Control charts evaluate the process stability (predictability) and Ppk evaluates the process capability, i.e, the ability of the process to produce within set speciation limits. These indices help track process performance and potential areas for improvement. For the Ppk value to be valid, it is important to ensure that the data are representative of the process [25]. 9 3 Methodology In this chapter, the methodology applied in this project is presented. It provides an overview of the research approach together with a description of how Six Sigma methodologies were applied during the project. Finally, the specific methods used for data collection, laboratory testing and data analysis are outlined. 3.1 Research Approach This section outlines the research design used to carry out the project. It describes the steps followed during the process of the project. The structure of the report is presented to guide the reader through the different stages of the study. 3.1.1 System Perspective on the Value Chain The project was conducted from a value-chain perspective, to include all relevant stages of the chain, from product development to production. The main purpose was to examine the characteristics of SAP and pulp such as their amounts, distribution and how they are controlled, and to capture it in a holistic view. This included understanding the product development stage where amounts and material details are specified based on customer needs. It involved gaining an understanding of how the specifications are tested and verified and how the specifications are later transferred into the production facilities. The project investigates potential noise factors and gaps at all phases of the chain and their impact on the expected outcome. This comprehensive approach allows for a broader understanding of inter-dependencies and cause and effect relationships, helping to identify possible bottlenecks and gaps. By optimizing several steps of the value chain, rather than isolated segments, the project aims to find opportunities for improvement to enhance overall quality throughout the chain. 3.1.2 Application of the DMAIC Framework The DMAIC framework, which stands for Define, Measure, Analyze, Improve, Control, has served as a foundational framework for the projects structure [26]. Originating from the philosophy Six Sigma, DMAIC provides a systematic approach to identifying inefficiencies and implementing solutions for achieving better quality and customer satisfaction. The method is divided into five stages, however, throughout the project, these stages have been revisited and revised for an iterative and comprehensive understanding. 1. Define The initial step involved creating a problem statement that clearly defined the objective of the project and the processes that need improvement. Several meetings were held with relevant people that had knowledge that related to the scope of the project in order to gather input on the specific area that was to be investigated. An 10 workshop was conducted together with semi-structured interviews. Process mapping was also done to better understand the company’s way of working. 2. Measure To get a comprehensive understanding of the process and the specific areas to be investigated, data were collected from various sources. This data was used to analyze cause and effect relationships and to identify key metrics of interest. Laboratory tests were applied to identify key indicators and methods used to assess product quality. 3. Analyze In this stage, the data collected during the measurement phase was analyzed to uncover variation and to identify underlying gaps in current control measures. Correlation and capability studies were conducted to identify patterns or dependencies in parameters or methods. Additionally, the findings from semi-structured interviews were analyzed providing further insights. 4. Improve Based on the insights gathered from the previous steps, possible improve- ment for more optimal and robust parameter settings and tolerance design has been suggested. The findings from the earlier stages have been used to gain a broader understanding of the causes of variation and errors, enabling the creation of recommended control measures. 5. Control An essential step in this framework is the control phase, which ensures accountability for sustaining the process improvement identified. During this phase, the findings and results were communicated to relevant stakeholders and product owners to ensure that opportunities for improvement are understood and implemented. Preliminary suggestions for practical control measures were also presented, along with their benefits. 3.1.3 Process and Project Layout In this section, the process of the entire project is explained in relation to the structure of the report. The initial stages of the project focused on defining the problem and gaining a broad understanding of the company and its operations. The methodology chapter outlines the actions taken and the data collection method used throughout the process. Data were gathered mainly from lab testing but also through workshops, literature reviews and semi-structured interviews. This is summarized and presented in the result chapter. The result and findings were analyzed using statistical evaluations and qualitative insights from interviews. The analyze chapter also included proposed improvements and practical control measures with the goal of enhancing quality assurance efforts. Final conclusions and further recommendations are presented in the concluding chapter. An overview of the practical process connected to the structure of the report is presented in Figure 3.1. 11 Figure 3.1: An overview of the report structure in relation to the projects process (Own Illustration) 3.2 Research Methods In this section, the applied methods during the project will be presented. The three overarching phases of the methodology are divided into data collection, lab testing and data analysis. The details of the methods will be outlined together with an explanation of how they were conducted and implemented. 3.2.1 Data Collection In order to ensure a broad and deep understanding of the project’s problem definition, a structured data collection was carried out in the first stage. The data collection took place in several steps with the aim of gathering both qualitative and quantitative data. During this stage, interviews and workshops were held, and both internal and external searches were conducted. The methods used are presented below. 3.2.1.1 Interviews As part of the data collection, semi-structured interviews were conducted with selected employees at the company. People from different parts of the organization were interviewed to create a broad picture, including product developers, lab engineers and those working closer to production. Interviews were used as a method in several parts of the project to capture existing processes, experiences, and challenges. Semi-structured interviews are a method based on predetermined open questions, but where it is room to discuss certain topics freely [27]. This form of interview was chosen to enable dialogue on interesting topics and to adapt the questions according to the direction the interviews took. Additionally, semi-structured interviews are a good option when a mix of qualitative and quantitative data is desired [27]. The interviews held in the beginning of the project were mainly with product developers and focused on understanding the products function and how they relate to the process of setting requirements. Interviews were also held with quality engineers at the company to understand the role of statistical tools within the company and to gain knowledge in quality practices. An interview was held with a production engineer at a prototype facility 12 to gain better understanding of the production process. The interviews were documented through notes, and the material was analyzed thematically to identify recurring patterns regarding needs and challenges. The questions were tailored for each interview to address the unique experience of the interviewee and to focus on the most relevant areas in relation to their role. The duration of the interviews was approximately one hour for each. 3.2.1.2 Workshop Another method used during the data collection was a workshop, also known as the Affinity Interrelationship Method (AIM). An AIM workshop is a method for understanding and solving complex problems by breaking down the problem into several smaller parts [27]. It is a step-by-step process that ensures that qualitative data is collected. The workshop is based on a predetermined open question, where all members answer this question on their own post-it notes before the answers are grouped and further explained in several steps. The question that the AIM workshop revolved around was: ”What potential challenges may prevent the company from taking control over SAP and pulp amounts and distribution within their products?” The purpose of the question was to get participants to not only assess current issues but also reflect on the underlying reasons with the difficulty of addressing them. This helped to understand the company’s perspectives on their own processes and quality operations. The question was formulated in an open-ended manner to encourage participants to give detailed responses. The objective of the AIM workshop was to discover a shared understanding of the root cause of a complex issue. The workshop lasted for one hour and involved seven members of the light incontinence team at Essity, including product developers and quality engineers. Each participant was provided with a notepad and pencil to actively participate in the exercise. For 15 minutes, the participants wrote answers to the questions on notepads and then posted them on the wall for everyone to see. Following this, the responses were explained and then categorized and titled based on common themes. When the AIM workshop was completed, a follow-up session was conducted to compile everything that was discussed and revealed during the workshop. 3.2.1.3 Internal and External Searches As a complement to the other data collection methods, a more targeted information search was conducted both internally within the company and externally in relevant databases and sources. The aim was to gather information regarding existing methods and documentation that could serve as a basis for further analysis and to ensure that the project was based on established information and expanded the knowledge base. External and internal searches are established methods for identifying relevant and existing information within a specific subject [27]. By combining internal and external sources, a wide range of perspectives is enabled. The internal search focused on the company’s documentation, including process documents and previously conducted projects. The external search targeted scientific literature, with searches conducted via databases such as Google Scholar and Scopus. The external search formed the basis for the theoretical framework and statistical evaluation. The information 13 searches were conducted iteratively throughout the whole project. Search words used were: quality assurance, incontinence, super absorbent polymers, robust design, together with similar words related to the project. 3.2.2 Laboratory Testing Lab testing is another method that has been used continuously throughout the project. The lab method used has been linked with the parameters SAP and pulp, as well as relevant methods connected to these materials. These tests have been used a way to verify amounts and distribution and product performance, in order to link lab results to specifications. Both design parameters and technical functions have been analyzed to identify potential opportunities. Two product concepts were created with the purpose of testing their performance, referred to as product concept A and B in the results and analysis chapters. The two product concepts were prototype products designed to resemble incontinence products. They vary in both Thickness, Weight and SAP and pulp amount, enabling a broader perspective when evaluating the concepts. The concepts were manufactured on a prototype machine that closely replicates regular production. The goal of this analysis has been to evaluate how various quality requirements such as target, specification limits and equivalence limits can be applied and controlled. To facilitate a comparison, a simplified specification has been created from which the concept will be evaluated. The values have been set based on dialogue with product developers to find reasonable values. No exact values will be provided, instead, everything will be described as a percentage of the target. A test plan was developed for the concepts, defining test parameters, test methods, and the number of variants to be tested in each method. The tests were carried out in a lab environment using established test methods within the company. Before the methods were carried out, training in the methods was conducted to ensure they were performed properly and consistently. All tests were documented, and photos were taken to capture interesting findings. The methods used were related to material amounts, material distribution, and functional tests. The two different sampling strategies used were random sampling and stratified sampling with a focus on extreme values. For product concept A, a total of 240 samples were received. To capture the entire population’s spread of high and low SAP amounts, stratified sampling was chosen as the method. Stratified sampling involves selecting variants from subgroups based on relevant parameters, ensuring that extreme values and thus critical cases are included in the testing[28]. This involves dividing the population, in this case based on Weight. Thereafter, products were randomly picked from each group. This sampling strategy ensured that extreme values such as low and high weights, but also intermediate weights were included in the testing. Through this, a good conclusion about the current state and variations could be drawn. For product concept B, random sampling was used, where products were randomly chosen from the bag. Random sampling means that a population is represented by a random selection of test variants [28]. By distributing the selection randomly, it is ensured that the result is not affected by conscious preferences and systematic errors. However, there is a risk that some extreme values are missed. 14 The number of concepts tested for each method was considered for both concepts. For both concepts, a sample size of 24 was used as the standard for most lab methods, however, when time constraints limited testing, a reduced sample size of 10 was applied instead. Exceptions were when measuring Product Weight where 240 samples were tested for concept A and 100 for concept B. Additionally, the measurement for Thickness had a smaller sample size for both concepts due to time constraints. The laboratory procedures will be described in detail in the result section, alongside the execution and the outcomes, to better connect the results in relation to how the tests were performed. 3.2.3 Data Analysis The collected data from interviews, workshops, internal/external searches and lab tests were analyzed using a combination of qualitative and quantitative methods. The purpose of the analysis was to transform the collected data into structured insights. 3.2.3.1 Qualitative Analysis Qualitative data, such as collected data from semi-structured interviews, results from the AIM workshop, and information from internal and external searches, were analyzed using qualitative methods. Initially, thematic analysis was used, where recurring themes were discovered and listed. These themes were further broken down into various cause-and-effect analyses, where Quality Function Deployment (QFD), Process mapping, P-diagrams, Fishbone diagrams and Design Failure Mode and Effects Analysis (DFMEA) were used as tools. QFD was used as a tool to identify relationships between customer needs, technical functions, and design parameters[29]. QFD is a method where, for instance, customer needs are listed on one axis and technical requirements on the other. The relationship between individual parameters is assessed and weighted, making it possible to prioritize which technical functions have the greatest impact on customer satisfaction. The grading was based on a numbering system of 1, 3, or 9, where 9 indicates the strongest impact. The grading criteria are further detailed in the results, reflecting the knowledge gained throughout the process and the insights from interviews and workshops. Furthermore, process mapping and P-diagrams were used as tools to identify and visualize the process and potential causes of variation [26]. Process mapping was used to describe the development flow, where an overall picture described the different phases of the value chain. A process mapping is a good tool to identify the flow of events in a process as well as inputs and outputs in each step of the process. Based on this comprehensive process map, separate P-diagrams were established for each individual phase of development. These diagrams are used to systematically identify factors affecting each step, including controllable parameters, uncontrollable parameters, failure modes, and desired outcomes. Fishbone diagrams were used to structure and visualize cause-and-effect relationships between identified problems and their possible causes[26]. Fishbone diagrams are a method in which everything starts from the main problem, and the causes of the problem are divided into categories such as people, machine, material, method, environment, and measurement. Each category of causes is then broken down into underlying causes. 15 Another method that was used during the qualitative analysis was a DFMEA. A DFMEA is a risk analysis method used to identify and analyze potential failure modes in a products design by evaluating potential failures in individual components or materials, their causes, and consequences [30]. Each potential failure was graded based on severity, occurrence and detectability. Severity describes the impact on product quality and customer satisfaction should the failure mode occur, rated from 1 (not severe) up to 10 (very severe). Occurrence explains the likelihood of the failure mode happening, with 1 being unlikely and 10 being very likely. Detection indicates the probability that the current control system will identify the failure before it reaches the next process step, with a lower value indicating a higher chance of detecting the issue. Together these form the Risk Priority Number (RPN), where a higher value indicates the need for investigations and improvement. 3.2.3.2 Quantitative Analysis Quantitative data from lab testing and existing data were processed using Excel and statistical tools. The software Minitab was used as a statistical tool to analyze the collected data. The purpose of the quantitative analysis was to identify patterns, variations, correlations and significant differences between the different product concepts. Depending on the data type and objective of the analysis, various tools in Minitab were used. The following section outlines the main statistical tools used in this project. Gage Repeatability and Reproducibility (Gage R&R) was a method used to perform a Measurement System Analysis (MSA) on specific methods, such as Thickness measurement and Weight measurement. Gage R&R is used to assess the reliability of a measuring instrument [26]. The analysis shows the proportion of total variation due to measurement uncertainty and whether this uncertainty comes from random causes (Repeatibility) or special causes (Reproducibility). When conducting a Gage R&R study, it is essential to compare the results against established acceptance criteria. Table 3.1 below outlines these criteria, their threshold values, and their interpretation [26]. Contribution(%) Total Variation(%) Marginally Acceptable < 9% < 30% Good < 4% < 20% Excellent < 1% < 10% Table 3.1: GR&R Acceptance Criteria Histograms were used to visualize the distribution of measured values. The graph in the histogram can show the mean and variation of a data set [26]. Vertical lines can be used to indicate target values, set specification limits for maximum and minimum values, and equivalence limits if these exist. The histograms provided a good overview of the data distribution and facilitated the assessment of normality, spread, and deviations from specifications. Analysis of Variance (ANOVA analysis) was another tool used. ANOVA analysis is a method used to examine differences in mean values between groups and to determine whether a specific factor has a significant impact on the mean value of a response or not [26]. ANOVA enables an objective comparison and reduces the risk of visual misinterpretations. 16 Regression analysis was also used to examine relationships between. Regression analysis p This method provides a predictive model that helps identify and understand significant factors influencing the results [26]. To complement this analysis, factorial plots were employed to visualize interactions and main effects, making it easier to interpret complex relationships in the data. Furthermore, capability analysis was used to assess of the lab results performed in relation to the defined specification limits. As part of this analysis, Ppk was a statistical measure applied to evaluate how well the process performs in relation to its specification limits[26]. This helped to support a more data-driven understanding of the process behavior in relation to quality expectations. 17 4 Result This chapter presents the results of the thesis, structured according to the various phases of the value chain highlighted in this study. It begins with an overview of the current state, describing each phase, from product development and design verification to process validation and ongoing production, and their current practices. The chapter then follows the sequence of these phases, detailing the activities done within each part in relation to the project. A cause-and-effect analysis is presented to provide insights into the product development phase. Next the design verification process is explained, including the results from laboratory tests conducted on test concepts. The process validation phase is then presented, focusing on the relevant validation measures involved together with a root-cause analysis. Finally, the ongoing production phase is described, highlighting the current process and the quality practices employed. 4.1 Current Situation This section provides a comprehensive overview of the current situation, detailing the existing procedure from product development, design verification, process validation and running production. This serves as a foundation for identifying potential areas for improvement. The information presented in this chapter is based on insights from semi-structured interviews, workshop and internal research. Several mentioned similar challenges and insights, contributing to a broad understanding of the current situation. An overview of the value chain, from a high-level perspective is illustrated in Figure 4.1. This project will primarily focus on the product development, design verification, process validation and running production, as highlighted by the yellow boxes. Meaning, the process of translating customer needs into a Requirement specification will not be a central focus in this report and is seen as input to the phases analyzed in this study. 18 Figure 4.1: Schematic overview of the value chain (Own Illustration) 4.1.1 The Product Development Phase In the product development phase, the Requirement specification serves as a foundation and input when starting to develop the design and Product specification. During the design stage, initial target values, centering and variation requirements for design parameters, like SAP and pulp levels are determined. A common theme in the interviews was the emphasis on setting clear and specific re- quirements in the Requirement specification. These requirements form the basis for the Requirement specification which is handed over to production. This helps to not only ensure that customer needs are properly addressed, but also that these are accurately communicated to production for correct implementation. Anything not clearly specified in the Requirement specifications, Product specifications, and other documents will not be adhered to and implemented by the production. Therefore, using proper specifications as a tool to improve control and quality was something the interviewees highlighted. The design solution developed is assessed to identify potential failures throughout the development process. The interviewees highlighted that the process is complex, making it crucial to analyze and consider all steps. They emphasized the importance of clear and precise communication at all parts of the development process, in order to reach high quality. Additionally, the interviewees explained that it is important to clarify and establish clear areas of responsibility early on. This is particularly important for ensuring that areas such as following up on the correct SAP and pulp levels are properly managed, preventing any omissions. Responsibilities are distributed across several teams working together. The Innovation team is responsible for generating concept ideas. These concepts are then developed into producible products by the product developers in the R&D department. The Technology team ensures that the necessary machinery is in place to produce the developed products, after which the Manufacturing team takes over for full-scale production. This means that understanding the cross-functional perspective can help R&D and the developers to make better-informed decisions by considering both upstream and downstream interfaces. 19 In Figure 4.2, the primary process steps are outlined by the blue boxes in the middle, together with the input and output marked by the dark blue boxes. The green boxes represent the control measures that are applied to manage these steps. The pink boxes highlight noise factors, influences that may introduce variation or uncertainty into the process. The orange box represents a potential failure mode that demonstrates what could go wrong if the process steps are not properly monitored. Figure 4.2: Process map of the product development phase (Own Illustration) 4.1.2 The Design Verification Phase The Requirement specification is received as a design input in this phase. During this stage, a DFMEA is conducted to identify potential failure modes and areas of concern with the design. Based on this, a design verification plan is developed and implemented. The activities outlined in the plan depend on the specific areas intended for verification, such as changed design parameters, impacted technical requirements and claims made about the product. When the design solution is verified against the technical requirements, the output from this stage is a verified Product specification. During the workshop, the risk of focusing more on functional parameters and less on design parameters was highlighted. For instance, when evaluating absorption properties it is important to also ensure that the correct amount of SAP and pulp are specified as well. The product is currently viewed and tested as a whole, potentially making it difficult to know what exactly is causing variations in the product. The process map for the design verification phase is presented in Figure 4.3. 20 Figure 4.3: Process map of the design verification phase (Own Illustration) 4.1.3 The Process Validation Phase An important part of the development chain is the transition from product design to running production, where process validation is a key intermediary step for this. The objective of this stage is to evaluate whether the production system can produce according to the verified Product specification. The process step for this phase is outlined in the process map in Figure 4.4. The extent of the validation activities depends on the risks and potential failure modes associated with producing the product. The process validation stage includes creating a clear validation plan, outlining what will be tested and how. Secondly, all equipment and systems are validated to ensure that they are installed correctly and meet the standards. Before executing the process validation plan, all objectives related to quality are set. The plan is then executed and evaluated to see if the process can meet quality and performance standards. In this phase, a validation method (referred to as SAP Validation Method 1) is implemented to calibrate the machines and ensuring that correct amount of SAP is applied to the product. Interviewees have indicated that if variations in SAP amount are present, it is believed to be reflected on variation in the overall product weight. They also expressed that there may be certain challenges in controlling and ensuring the accuracy of SAP and pulp amounts and distribution. 21 Figure 4.4: Process map of the process validation phase (Own Illustration) 4.1.4 Running Production Once the design verification and process validation phases are successfully completed, the product is ready for production. A Device Master Record (DMR) is handed over to production, that includes a comprehensive compilation of all documentation and specifications required to manufacture a product. Included in the DMR are Standard Operating Procedures (SOP) that outlines the specific procedures and tasks that must be performed. During production, products are periodically tested to ensure they meet the Product specifications. The product developers receive reports monthly for measured parameters during running production. The interviewees explained the importance of having a proper control plan in place with relevant sample size and frequency for testing. The interviewee also highlighted potential risks, particularly related to technical challenges during running production. SAP and pulp are mixed simultaneously in a mill using turbulent airflow and later sprayed onto the product. A full explanation of the manufacturing process is presented in Chapter 4.5.1. The interviewees explained that there is always an ongoing effort to develop and build better machines. However, variations will always be present, making it essential to identifying the underlying causes for manufacturing variation and, in parallel, seek robust design solutions that are less sensitive to manufacturing noise. During the workshop, the cost of continuous monitoring of SAP and pulp amounts and distribution was discussed, as well as the importance of an aligned view of how to interpret Product specifications. The process map for running production is presented in Figure 4.5 22 Figure 4.5: Process map of running production (Own Illustration) 23 4.2 The Product Development Process This chapter presents a systematic review of the parameters used in this project to support product development. It begins with a fundamental description of some of the most relevant customer needs related to SAP and pulp. Followed by a derivation of product functions and finally the design parameters, which constitute the actual factors for the design output. To clarify the relationships between needs, technical requirements, and design parameters, a QFD is presented. Finally, lab results are analyzed to explain and demonstrate the actual impact of the design parameters on product functions. It is of interest to concretely see the consequences of variation in SAP and pulp amounts and distribution. Furthermore, measurement systems used within the company to evaluate design parameters and technical requirements, as well as the precision of the measurement systems, are presented. 4.2.1 Identification of Key Parameters for Inconti- nence Products During the development of a product, the functionality and design are always based on the user’s needs. It is from the user’s needs that the technical requirements on product functions are determined. The technical requirements are then used as design input to determine a design, where the design parameters become the design output. For an incontinence product, there is a wide range of needs, however, for this project, three overarching needs related to SAP and pulp have been selected for further examination. The needs included were based on both interviews and the theoretical framework. The technical requirements are in the next step a translation of the customer’s and user’s perceived needs into measurable and verifiable properties. Based on the technical requirements, the design parameters are determined. These parameters are the design choices of the product developer. It relates to construction and often involves material choices and amounts. In Figure 4.6 below, the dimensions are listed. 24 Figure 4.6: Key parameters for incontinence products 4.2.2 Quality function deployment In order to understand the relationships that exist between the three dimensions described in the previous chapter, two QFDs were developed based on information gathered through interviews and the theoretical framework. One describes the relationships between the customer needs as a function of technical requirements. Furthermore, the technical requirements are broken down as a function of product design parameters. The functions can be described as follows, N stands for needs, T for technical requirements and D for design parameters. Nx = f(Tx) (2) Tx = f(Dx) (3) In the QFD diagrams below, the strength of the correlation between different elements is rated. In this case, the numbers 1, 3, and 9 are used to indicate the strength of the correlation, where 1 indicates a low, 3 represents a medium, and 9 means a strong correlation. There is also a case where the box is left empty, indicating that no correlation between the parameters can be drawn. The relationship ratings are presented in Figure 4.7 below. Figure 4.7: Importance rating for the QFD 25 The QFD diagram explaining the relationships in Equation 2, how the technical require- ments contribute to customer needs, is presented in Figure 4.8 below. The importance rating describes from 1-5 which user need is the most important, where 1 is the least important and 5 is the most important. The rating is based on the previously presented theory as well as knowledge that has been gained during the project. Of the three presented needs, Leakage security is the most important need to fulfill, followed by Dryness and then Discreetness. On the other hand, it should be kept in mind that the rating can vary for different products and purposes. As can be seen in Figure 4.8, Leakage security is strongly related to Inlet time and Absorption capacity. A fast Inlet time ensures that the liquid is removed from the surface quickly which reduces the risk of leakage. A high Absorption capacity is essential for capturing and retaining liquid which prevents leakage. Dryness is mostly correlated with Rewet, since low Rewet keeps the surface dry. Discreetness is mostly correlated with Thickness, since a thicker product may be more visible through clothing. The Table shows that Absorption capacity has the highest technical importance score. However, all the technical requirements serve a critical role in the product and are often interrelated, highlighting the need to consider them collectively. Figure 4.8: QFD 1 - Customer needs and Technical requirements A second QFD diagram, which presents the relationships in Equation 3, how the design parameters contribute to the technical requirements, is presented in Figure 4.9 below. In this diagram, the technical requirements are also rated based on importance, where Inlet time and Absorption capacity are considered the most important, which also emerged from the previously conducted QFD. Figure 4.8 presents that SAP amount is a prominent driver of Rewet and Absorption capacity since it is the main material that can retain large volumes of liquid. The Pulp amount poses a correlation between Thickness since the pulp is the main material that adds Thickness to the product, therefore, product Thickness will also be further investigated as the amount and distribution control of the pulp is to some extent encompassed by Thickness. The Distribution of SAP is correlated with Absorption capacity and Rewet and helps to ensure that the liquid is captured. The 26 diagram also shows a strong correlation between Thickness and Inlet time, as a thicker product allows the liquid to pass through and spread within the product more quickly. SAP amount, Product Weight and Thickness receive the highest importance scores due to their correlation with different technical requirements. Figure 4.9: QFD 2 - Technical requirements and Design parameters 4.2.3 Correlations Between Design Parameters and Technical Requirements In this chapter, the impact of SAP amount and Thickness on the product functions Inlet time and Rewet, will be presented in more detail. Both main effects and potential interaction will be analyzed to identify their impact on the product’s functionality. The two design parameters, SAP amount and Thickness, were chosen for further analysis as they had a high technical importance score in the QFD diagram, but also because they are interesting to study from a specification perspective. The decision to examine only these was based on the strong correlation shown between the selected design parameters and technical requirements in the QFD diagram but also because of existing data available. When studying Inlet time, two different times are measured. The first one involves applying liquid to a concept and measuring how long it takes for the liquid to disappear from the surface of the product. Then, the product is dosed a second to determine the Inlet time for a second dose. 4.2.3.1 SAP Amount Correlations In this section, the correlation between SAP amount and the two technical functions, Rewet and Inlet time, will be presented. In Figure 4.10, the relationship between SAP amount and Rewet is shown, the higher the amount of SAP in a concept, the lower Rewet. With an SAP amount of 400 grams per square meter, the Rewet is as low as 2g, while for a concept without SAP, the Rewet is up to 25g. 27 Figure 4.10: Correlation between SAP amount and Rewet The relationship between SAP amount and the Inlet times can be seen in Figure 4.11, where the correlation shows that the higher the SAP amount, the faster the Inlet time. It is especially true for Inlet time 2, seen on the right side. Where an even steeper curve can be observed for the significance of SAP amount, indicating that the SAP amount plays an even greater role for the second Inlet time. Figure 4.11: Correlation between SAP amount and Inlet time 4.2.3.2 Thickness Correlations For Thickness, the relationships with Rewet and Inlet time are presented below. In Figure 4.12, the relationship between Thickness and Rewet is shown. It can be seen that the thinner the product, the lower the Rewet. 28 Figure 4.12: Correlation between Thickness and Rewet Figure 4.13 describes the relationship with the Inlet times, showing that the thicker the product, the faster the Inlet time. The same relationships applies for the two different Inlet times. Inlet time and Rewet are therefore in conflict, one function improves with a thin product while the other improves with a thick product. To optimize these functions, careful consideration is required to find a balanced solution. Figure 4.13: Correlation between Thickness and Inlet time 4.2.3.3 Interaction Effects Between SAP Amount and Thickness In addition to the main effects of SAP amount and Thickness on Rewet and Inlet time, interaction effects can also be observed. The interaction effect of SAP amount and Thickness on Rewet can be seen in Figure 4.14 below. In the lower left quadrant, the results show that the higher the amount of SAP in a concept, the less impact Thickness 29 has. With a lower amount of SAP, Thickness plays a larger and more decisive role in Rewet. The red curve in that quadrant represents a Thickness of 10.33mm, while the blue curve represents 3.82mm. The upper quadrant describes the same relationship, showing that for a concept with no SAP at all, Thickness plays a significant role. When there are high amounts of SAP in a concept, Thickness has less impact. In this figure, the blue line represents no SAP, and the red line represents 400 grams per square meter of SAP. Figure 4.14: Interaction effects for Rewet The interaction effects for Inlet time 1 can be seen in Figure 4.15. It is evident that when the SAP amount is low, Thickness does not play a role in improving the Inlet time. Thickness has the most impact when the SAP amount is simultaneously high, then a thicker concept performs better. For a thin concept, the SAP amount does not significantly affect a faster Inlet time. 30 Figure 4.15: Interaction effects for Inlet 1 The interaction effects for Inlet 2 are presented in Figure 4.16. For a concept with low SAP amount the effect of Thickness is low, and the Inlet time remains on a low level even with increased Thickness. In contrast for a concept with high SAP amount, the effect of Thickness is high with a fast inlet for a higher Thickness. Somewhere in between, with medium SAP amount, Thickness does not play a role. Figure 4.16: Interaction effects for Inlet 2 31 4.2.4 Precision of Measurement Systems In this chapter, methods related to the previously presented design parameters and technical requirements will be presented, as well as an evaluation of their precision. Methods are used during laboratory execution to evaluate and measure the performance of a product. During product development, design verification, and process validation, laboratory testing is used to make various decisions. The most basic methods to ensure the correct amount of material is present in the products and in the right place are Weight and Thickness. In addition to these, there are more complex and specific methods, where individual material amounts can be determined, such as SAP amount methods. Furthermore, there are methods related to determining the material distribution and methods linked to test product functions. Due to the fact that this project will be based on several laboratory tests, it was considered relevant to evaluate some of the most used methods. 4.2.4.1 Weight and Thickness Weight and Thickness, as previously described, are two of the most basic methods, which are easy to perform and providing a quick overview. The methods are related to two parameters that, from a product perspective, can affect many of the desirable functions of an incontinence product, such as Inlet time and Rewet. Weight is measured using a calibrated scale placed on a stable, level surface. The product is positioned in the center of the scale and once stabilized the Weight is recorded. Thickness is measured using a Thickness gauge that is placed on a stable surface. The product is positioned between a measuring foot and the surface ensuring the product is centered and flat. An MSA was conducted on Thickness and Weight to determine its precision. 5 samples to each methods were used. The results from the MSA are presented below. The Weight measurement system proved to be precise based on the Gage R&R analysis conducted. See Figure 4.17 for a summary of contribution and study variation for the Weight measurement. According to the established acceptance criteria (see Table 3.1 in Chapter 3.2.3.2) , the reproducibility and repeatability have a contribution value of less than 1%. In this case, the reproducibility consists of variation between operators. Together, they have a value of 1.2% for the total contribution of variation. With a result under 1%, the acceptance criteria indicate that it is excellent, and under 4% is considered a good result. Additionally, the study variation is within the acceptable range, with a percentage between excellent (under 10%) and good (less than 20%). This demonstrates that the measurement system is precise. It is also noted that the greatest source of variation is part-to-part, meaning that the measurement system variation is small in relation to the different parts. In Figure 4.17, the relationships for the different parameters can be read through the graphs, where it can be seen that the operators follow the same pattern but also how the variation for the different components looks. 32 Figure 4.17: Measurement System Analysis for Weight The Thickness measurement, however, showed a different result. The Gage R&R analysis indicated that reproducibility was excellent, but repeatability had a contribution of variation of 25.5%, which is above the threshold for marginally acceptable. The same relationship applied to study variation for the different parameters, which can be seen in Figure 4.18. Through the presented graphs, it can be seen that the operators follow the same pattern. Additionally, it can be observed that there are differences in the various measurements for the same product and how it ranges. Figure 4.18: Measurement System Analysis for Thickness 33 Due to this result, with a high percentage of repeatability, it was deemed relevant to further investigate and determine the underlying cause. Initially, a hypothesis was formed that the products might be compressed during testing, becoming thinner and thinner, which would result in a decreasing Thickness. This would mean the testing was destructive in that way. Therefore, a time series plot was created to study the testing over time, as shown in the Figure 4.19 below. The graphs indicate that the Thickness generally increases over time and with the number of measurements, which contradicts the initial hypothesis. In fact, the opposite is true, the products become thicker and thicker. Figure 4.19: Time series plots for Thickness 4.2.4.2 SAP Amount Methods To determine the amount of SAP present in different products, established internal methods provided by the company were used. There is always uncertainty regarding the measured values and how well they reflect reality. In this project, three internal methods were evaluated and applied, referred to as SAP Method 1, SAP Method 2, and SAP Method 3. • SAP Method 1 is relatively simple and requires minimal preparation, allowing for spontaneous execution. Additionally, this method is quick to perform. • SAP Method 2 is more complicated and requires several steps to be prepared before testing. Consequently, the time required to execute this method is somewhat longer compared to Method 1. • SAP Method 3 is even more complex, and it is believed that this increases accuracy. Despite its complexity, it usually does not take longer to perform than Method 2. However, preparation can make it feel more time consuming. The three methods were compared to see the difference in average values. Additionally, the intended SAP amount in the concepts was compared with the methods measured 34 value, to conclude which method was the most correct. The results are presented in Figure 4.20 below. It can be seen that the SAP Method 3 has the highest SAP amount mean and it is closest to the intended SAP amount in the product concept. It can also be seen that the SAP amount method 2 and 3 has some overlapping values. However, an ANOVA analysis shows that there is a significant difference between the methods. Therefore, only SAP Method 3 will be further used to evaluate the SAP amount in the concepts. Figure 4.20: Individual value plot for SAP amount methods 4.2.4.3 Material Distribution Methods In this project, two distinct methods were used to analyze the material distribution in product concepts. This section presents both approaches, referred to as Distribution Method 1 and Distribution Method 2, and provides a comparison between them. • Distribution Method 1 uses visual imaging to provide a good estimate of material distribution. Additionally, this method provides a Coefficient of Variance (CV) value, which is a statistical measure that can be used to draw conclusions about the uniformity of material distribution in the concept. A lower CV suggests a more uniform distribution, while a higher CV indicates higher variation. The methods also provides an estimated surface weight of the studied concepts. • Distribution Method 2 is another method for studying material distribution. This method is somewhat simpler and can be used to study different parts of a product or concept by punching out and analyze these. The two distribution methods were also evaluated in regard to their accuracy. Distribution Method 1 was used to analyze both entire products and a smaller area on a product. It was known that the smaller area studied contained more material per square meter than the entire product. The basis weight obtained can only be seen as an approximate estimate as it is known that this is not particularly accurate. However, the basis weight for the 35 entire product and the small, analyzed area with more material than the whole product was the same, which indicates that distribution method 1 cannot identify these differences in material amount. This can be seen in Table 4.1 below. Variant Basis Weight (gsm) Whole Product 632 Smaller Area (7x7cm) 629 Table 4.1: Evaluation of Distribution method 1 The purpose of this analysis was to see if Distribution Method 1 could be used similarly to Distribution method 2, to study the CV and basis weight of different areas. By doing so, conclusions could be drawn about how much material that is present in these areas, as a complement to the visual image of the distribution. The result showed that Distribution method 1 is a good method for visual study, as it provides a quick overview of how the material is spread across the product. It also generates a CV value that quantifies the material distribution. However, the basis weight cannot be applied to study material amounts in specific points based on analyzed data. 4.2.5 Target and Tolerance Design in Specifications Based on interviews and meetings with quality engineers at the company, together with internal searches, the process for setting target values and tolerance limits has been outlined. Specifications serve as a crucial link between the voice of the customer and the tangible characteristics of the product. The process involves translating customer needs into technical requirements and design parameters. As previously presented in the QFD, these parameters are closely interconnected. Two key types of specifications used by the company are Technical Requirements and Product specifications. The Technical Requirements define the essential functions and performance levels that the product must achieve. These requirements guide the product developers in designing solutions that fulfill the set criteria. Following the product development process, the Requirement specification is created, detailing specific material amounts, Thicknesses, and other attributes necessary to meet the technical requirements. The insights presented in this chapter are based on interviews with product developers and a review of existing specification documents. Setting the specification involves finding optimal target values for each parameter. However, in reality, due to variation in manufacturing and systems, it is almost impossible to hit these targets with perfect precision. Therefore, robust design principles are valuable to try to make the design insensitive to incoming variation. Two commonly used tools are variation and centering requirements, which will be presented in more detail below to understand the significance of using these requirements. In Figure 4.21 below, a schematic picture of the difference between the two can be seen. 36 Figure 4.21: Schematic figure of variation and centering requirements (Own Illustration) 4.2.5.1 Variation Requirements Variation requirements are defined by upper and lower specification limits, referred to as USL and LSL. Specification limits define when a product is considered to be defective. For instance, if a product has a target weight of 15 grams, an acceptable specification range may be 15 ± 2 grams. Products falling outside of this range are classified as defectives, as they may not meet performance or quality expectations. Variation requirements are used in both Requirement specifications and Product specifications and have been used for a long time. It is important to analyze both process variation and Ppk demand in order to set a producible specification. Ppk defines the proportion of products allowed to fall outside the specification limits. Essity follows industry standards and has a requirement for Ppk ≥ 1.33 for design parameters and Ppk ≥ 1.0 for technical requirements. A Ppk value of 1.33 indicates that 99.9937% of all values will fall within the specified limits, while a Ppk of 1.0 means that 99.73% of all values will fall within the limits. 4.2.5.2 Centering Requirements Centering requirements are defined by equivalence limits, which can be seen in Figure 4.21 above where LEL and UEL describe the lower and upper equivalence limits. Centering requirements are used to center a parameter around the mean value. This means defining the acceptable range within which the mean value of a given parameter may differ from the target. Unlike variation requirements, which control the variation of individual values, centering requirements focus on the mean of the population. Centering requirements are used in the Requirements Specification but are currently being implemented in the Product specification. The reason centering has not previously been applied in the Requirement specification is that the same product may be manufactured on different machines with varying conditions, making it difficult to define appropriate levels. 37 4.2.6 Design FMEA A DFMEA was conducted in order to systematically identify design related causes to failure modes of critical functions such as Inlet time, Rewet and Thickness, and assessing their impact. This risk assessment helps to gain a holistic understanding of SAP and pulp, from design input to final product output and to identify potential mitigating actions. Figure 4.22 , presents the key parameters assessed in the DFMEA including severity, occurrence and detection and their rating. These three factors are multiplied to obtain a risk priority number, which is used to prioritize actions. Figure 4.22: DFMEA (Design Failure Mode and Effects Analysis) The highest RPN value are seen for Inlet time and Rewet, both of which share a common potential cause of failure, insufficient SAP amount. This highlights the need to gain better insight into these parameters and the importance of setting clear specifications based on variation. Although Absorption capacity has a lower RPN of 72, due to potential failures being easily detected by performing absorbency lab tests, it is still a critical function that needs to be considered. The potential cause of failure related to Thickness showed a RPN of 112 but has the highest severity and occurrence scores. This underscores the need for further understanding of environmental influences like moisture and compression during production when setting Product specifications. These potential failure modes highlight critical areas where increased control and understanding may be needed to ensure product consistency throughout the value chain. 38 4.3 The Design Verification Process In the design verification process, product concepts are tested to evaluate whether the proposed design solution will meet the Technical Requirements. When testing product performance, it is equally important to check the compliance to Product specification. In this stage, two different test concepts, named A and B, that were produced in a prototype facility at Essity were analyzed. This section presents the results and findings from all the laboratory tests that have been conducted on the two product concepts. The purpose has been to understand the importance of design verification and identify any existing gaps. 4.3.1 Overview of the Evaluated Parameters The selection of parameters for the laboratory testing was based on the relevant design parameters and technical requirements presented. An overview of the evaluated parameters together with their purpose is presented in Figure 4.23 . Figure 4.23: An overview of the evaluated parameters 4.3.2 Evaluation of Variation and Centering Require- ments for Product Concept A This chapter presents the results from testing and evaluating product concept A. In Figure 4.24 below, the limits against which the concept will be compared can be read. For concept A, centering requirements were set for SAP and pulp amounts. For the Product Weight, variation requirements are used. Thickness has previously been shown to be a critical parameter related to both SAP and pulp, and is therefore specified with variation requirements and centering requirements. The same applies to the technical requirement Rewet, which is strongly linked to Thickness. 39 Figure 4.24: Specification for Concept A In the following chapter, the results from testing Product Concept A using the previously described methods are presented. In addition to these methods, a standardized Rewet measurement method used internally by the company was also included. 4.3.2.1 Product Weight for Product Concept A Product Weight was one of the first tests conducted as it is a good way to get a quick overview. The total Product Weight can be used as a measure to understand if the correct amount of material is present in the product. However, it is difficult to distinguish specific material types. In Figure 4.25 below, an overview of the 240 samples that were weighed is shown. As shown in Figure 4.24, the Weight parameter is defined by both a target value and a variation requirement. The measured mean is very close to the target, although slightly skewed toward the higher side. From a visual inspection of the distribution graph, the majority of the values appear to fall within the specified variation limits. The process capability was evaluated, resulting in a Ppk value of 1.18. However, due to the limited sample size, there is a degree of uncertainty associated with the estimate. A lower bound of the Ppk was calculated at 1.08 with 95% confidence. Since the required Ppk for this parameter is 1.33, neither the estimated value nor its lower confidence bound meets the Ppk demand. Furthermore, no centering requirement is currently defined for this parameter, which allows the process mean to drift within the variation limits. 40 Figure 4.25: A process capability report of the Product Weight for Product Concept A 4.3.2.2 SAP Amount for Product Concept A Of the 240 samples that were weighed, 24 of them, both high and low weight products, were used to determine the amount of SAP and compare it with the specification. The subset was chosen to ensure a representative sample size while also balancing the time required to perform the method. In Figure 4.26 below, the results are presented through a histogram. The SAP amount is specified with target and equivalence limits. The average SAP amount was slightly shifted towards the lower end but remains within the lower equivalence limit. The equivalence limits thus enable control over the average shift relative to the target. It can also be noted that there was a large spread between the measured maximum and minimum values. Since there are no variation requirements for SAP amount, there is no requirement limiting the variation of individual values. 41 Figure 4.26: A histogram of the SAP Amount for Product Concept A Visually, it can be observed that the SAP amount falls within the set centering requirements. To study this further and with the small sample size in mind, an equivalence test was conducted. The equivalence test is presented in Figure 4.27 below. In the equivalence test, it can be seen that equivalence cannot be claimed and the lower bound for 90% confidence interval goes below the LEL. Figure 4.27: Equivalence test of the SAP amount for Product Concept A Since the SAP amount in this case lacks variation requirements in the set specification list, it was deemed relevant to analyze what appropriate limits could be based on the gathered data. Using variation limits in the form of upper and lower specification limits can restrict SAP variation in the products. The tolerance intervals were obtained through the measured standard deviation and the mean set to the target, as the tolerance limits 42 should be symmetrical around the target. In Figure 4.2 the interval is presented with a set requirement on Ppk of 1.33 and a sample size of 24 measured values. With a sample size of 24 and a tolerance factor of 5.4, the corresponding tolerance limits were ±35%. There are a significance impact of sample size and therefore, the same analysis was conducted for an increased sample size. With the increased sample size (N=100), the calculated tolerances become somewhat narrower. The suggested tolerance limit was ±30%, with a tolerance factor of 4.56. A tolerance factor of 4.56 corresponds to a two-sided tolerance interval that extends ±4.56 standard deviations from the sample mean, capturing 99,9937% of the population. However, while the analysis was repeated using a larger sample size (N = 100), the standard deviation used in calculating the tolerance interval was still based on the smaller sample size. As a result, the estimated tolerance interval may not fully reflect the true variability of a larger sample size. For a more accurate and representative interval, the standard deviation should ideally be recalculated based on the increased sample size, as it provides a more robust estimate of population variability. In Table 4.2 below, a quick overview of the tolerance intervals corresponding to different sample sizes and tolerance factors can be seen. SAP amount Sample size Tolerance factor Tolerance interval 24 5.4 ±35% 100 4.56 ±30% Table 4.2: Tolerance intervals for SAP Amount - Concept A Furthermore, a correlation analysis was performed to examine the relationship between Product Weight and SAP amount. The objec