Increase Space utilization and production capacity through process flow analysis - A case study of NIBE AB Master thesis in Supply Chain Management and Quality and Operations Management Alexander Nilsson Oliver Friman DEPARTMENT OF TECHNOLOGY MANAGEMENT AND ECONOMICS DIVISION OF SUPPLY AND OPERATIONS MANAGEMENT CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2021 www.chalmers.se Report No. E2021:116 Master’s thesis E2021:116 Increase space utilization and production capacity through process flow analysis - A case study of NIBE AB ALEXANDER NILSSON OLIVER FRIMAN Department of Technology Management and Economics Division of Supply and Operations Management Chalmers University of Technology Gothenburg, Sweden 2021 Increase space utilization and production capacity through process flow analysis - A case study of NIBE AB ALEXANDER NILSSON OLIVER FRIMAN © ALEXANDER NILSSON & OLIVER FRIMAN, 2021. Supervisor:Peter Almström,Department of Technology Management and Economics Examiner: Peter Almström, Department of Technology Management and Economics Master’s Thesis E2021:116 Department of Technology Management and Economics Division of Supply and Operations Management Chalmers University of Technology SE-412 96 Gothenburg Telephone +46 31 772 1000 iv Increase space utilization and production capacity through process flow analysis - A case study of NIBE AB ALEXANDER NILSSON OLIVER FRIMAN Department of Technology Management and Economics Chalmers University of Technology Abstract NIBE is a global provider of sustainable climate solutions. The strategy for the company is to increase the revenue by 20 % annually, both by acquisitions and or- ganic growth. Their product range includes various energy efficient climate solution such as water heaters, elements, and stoves. This thesis is focused on one of their product families within the water heater segment, Compact, at one of their facilities in Markraryd. Due to their philosophy to cover as large share of the market as possible, the com- pany produces a vast variety of products and components leading to an increased complexity of the production. And their high expansion rate causes problems since they experience that they need to utilize the available factory space more efficiently. As of today the production system is characterized by a push-based approach, lead- ing to inventory build ups that covers valuable factory space that could have been used to increase the current production capacity. This thesis therefore mainly fo- cus on the enhancement of space utilization and increase in capacity by providing suggestions and recommendations regarding reduction of inventories and work in progress (WIP) for the production flow of the Compact product family. To gain an overview of the production flow, value stream mapping was used as a tool. In or- der to understand the processes and where potential problems arise, the conducted current state map was analyzed and several improvement suggestions such as CON- stant work in process (CONWIP) loops and supermarkets have been recommended. With all improvement suggestions considered, a future state map was developed. This map visualized how the suggestions could lead to a reduction of inventories due to a more pull-based approach, hence increase the throughput according to Lit- tle’s Law. As can be seen in the future state map in section 7.3 the average waiting time is reduced, thereby both the space utilization and production capacity is believed to be increased. Although, it should be noted that it is needed to take a long-term strategic approach in order to evaluate which improvements that are feasible, and in which extent they are effecting the production flow. Crucial in order to achieve this is the communication between the different departments, as well as striving to continuously improve the flow further. Keywords: Value stream mapping, inventory reduction, space utilization, capacity, CONWIP, push and pull production v Acknowledgements This Master’s thesis is written at Chalmers University of Technology within the de- partment of Technology Management and Economics, in collaboration with NIBE AB and AFRY AB by students from master programs Supply Chain Management and Quality and Operations Management. The authors would like to thank our supervisor and examiner Peter Almström, who has been very helpful and provided valuable guidance throughout the thesis. Fur- ther, the authors would like to thank our supervisor at AFRY, Jessica Andersson, for her valuable insights and advises. Finally, the authors would like to thank Dan Ekener and Tomas Axelsson at NIBE AB that initiated the project and who has helped us during the project and our visits in Markaryd. Also, the authors feel great gratitude and appreciation towards all personnel at NIBE that have taken their time to let the authors conduct interviews and collect valuable production data. You have all contributed to making this thesis possible. Alexander Nilsson and Oliver Friman, Gothenburg, 2020 vii Contents List of Figures xiii List of Tables xv 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Problem statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.4 Delimitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 Methods 5 2.1 Quantitative research methods . . . . . . . . . . . . . . . . . . . . . . 5 2.2 Qualitative research methods . . . . . . . . . . . . . . . . . . . . . . 6 2.3 Data gathering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.3.1 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.3.2 Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.3.3 Site visits and observations . . . . . . . . . . . . . . . . . . . . 8 2.3.4 Quantitative data gathering . . . . . . . . . . . . . . . . . . . 8 2.4 Quality criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.5 Ethics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.6 Current state map . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.7 Future state map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.8 Analysis tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 3 Company description 13 3.1 NIBE Industrier AB . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.2 NIBE AB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.2.1 Product description . . . . . . . . . . . . . . . . . . . . . . . . 14 3.2.2 Production system . . . . . . . . . . . . . . . . . . . . . . . . 15 3.2.3 Quality and maintenance . . . . . . . . . . . . . . . . . . . . . 15 3.3 Factory description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 4 Theoretical framework 17 4.1 Order Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.1.1 Material Planning . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.1.2 Capacity Planning . . . . . . . . . . . . . . . . . . . . . . . . 18 4.2 Theory of constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 ix Contents 4.3 Variation and capacity utilization . . . . . . . . . . . . . . . . . . . . 20 4.3.1 Variation in quantity and time . . . . . . . . . . . . . . . . . . 20 4.3.2 Variation in quality . . . . . . . . . . . . . . . . . . . . . . . . 20 4.3.3 Variation and capacity . . . . . . . . . . . . . . . . . . . . . . 21 4.4 The bull-whip effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 4.5 Bottleneck identification methods . . . . . . . . . . . . . . . . . . . . 23 4.5.1 Average waiting time . . . . . . . . . . . . . . . . . . . . . . . 23 4.5.2 Utilization rate . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.6 Little’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 4.7 The Kingman formula . . . . . . . . . . . . . . . . . . . . . . . . . . 24 4.8 Push and pull production . . . . . . . . . . . . . . . . . . . . . . . . 25 4.9 CONWIP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.10 Lean production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 4.11 Waste in Lean production . . . . . . . . . . . . . . . . . . . . . . . . 29 4.11.1 Transportation . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.11.2 Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.11.3 Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.11.4 Waiting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.11.5 Overproduction . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.11.6 Over-processing . . . . . . . . . . . . . . . . . . . . . . . . . . 32 4.11.7 Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 4.11.8 Unused employee creativity . . . . . . . . . . . . . . . . . . . 33 4.12 Lead time reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 4.13 Value-stream mapping . . . . . . . . . . . . . . . . . . . . . . . . . . 34 4.13.1 Current state . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 4.13.2 Future state . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 4.14 The seven value stream tools . . . . . . . . . . . . . . . . . . . . . . . 40 4.14.1 Process activity mapping . . . . . . . . . . . . . . . . . . . . . 40 4.14.2 Supply chain response matrix . . . . . . . . . . . . . . . . . . 40 4.14.3 Production variety funnel . . . . . . . . . . . . . . . . . . . . 41 4.14.4 Quality filter mapping . . . . . . . . . . . . . . . . . . . . . . 41 4.14.5 Demand amplification mapping . . . . . . . . . . . . . . . . . 41 4.14.6 Decision point analysis . . . . . . . . . . . . . . . . . . . . . . 42 4.14.7 Physical structure mapping . . . . . . . . . . . . . . . . . . . 42 4.15 Spaghetti diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 4.16 RFID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 5 Current state 45 5.1 Current production flow . . . . . . . . . . . . . . . . . . . . . . . . . 45 5.1.1 Welding shop . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 5.1.1.1 Copper vessel . . . . . . . . . . . . . . . . . . . . . . 46 5.1.1.2 Stainless steel vessel . . . . . . . . . . . . . . . . . . 51 5.1.2 Metal sheet shop . . . . . . . . . . . . . . . . . . . . . . . . . 52 5.1.3 Painting shop . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 5.1.4 Final assembly . . . . . . . . . . . . . . . . . . . . . . . . . . 57 5.2 Current information flow . . . . . . . . . . . . . . . . . . . . . . . . . 58 x Contents 5.2.1 Current state map . . . . . . . . . . . . . . . . . . . . . . . . 59 6 Analysis 63 6.1 Utilization rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 6.1.1 Welding shop . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 6.1.2 Metal sheet and painting shop . . . . . . . . . . . . . . . . . . 65 6.2 Inventories and buffers . . . . . . . . . . . . . . . . . . . . . . . . . . 66 6.3 Transportation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 7 Future recommendations 71 7.1 Process level improvements . . . . . . . . . . . . . . . . . . . . . . . . 71 7.1.1 Inventories and buffers . . . . . . . . . . . . . . . . . . . . . . 71 7.1.2 Transportation . . . . . . . . . . . . . . . . . . . . . . . . . . 72 7.1.3 Pacemaker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 7.1.4 CONWIP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 7.1.5 Levelled production and optimized batch sizes . . . . . . . . . 76 7.2 System level improvements . . . . . . . . . . . . . . . . . . . . . . . . 77 7.2.1 More comprehensive data collection . . . . . . . . . . . . . . . 77 7.2.2 Frozen planning period . . . . . . . . . . . . . . . . . . . . . . 78 7.3 Future state map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 8 Discussion 83 9 Conclusion 87 Bibliography 89 A Appendix I A.1 Interview questionnaire planning department . . . . . . . . . . . . . . I A.2 Interview questionnaire production departments . . . . . . . . . . . . II B Appendix III B.1 Value stream map for the copper mantle . . . . . . . . . . . . . . . . III B.2 Value stream map for the steel mantle . . . . . . . . . . . . . . . . . IV B.3 Value stream map for the steel gable . . . . . . . . . . . . . . . . . . V B.4 Value stream map for the stainless steel vessel . . . . . . . . . . . . . VI B.5 Value stream map for the top plate . . . . . . . . . . . . . . . . . . . VII B.6 Value stream map for the side plate . . . . . . . . . . . . . . . . . . . VIII B.7 Value stream map for the back plate . . . . . . . . . . . . . . . . . . IX B.8 Value stream map for the front plate . . . . . . . . . . . . . . . . . . X C Appendix XI C.1 Flow Planner Report Current State . . . . . . . . . . . . . . . . . . . XI C.2 Flow Planner Report Future State . . . . . . . . . . . . . . . . . . . . XIV xi Contents xii List of Figures 3.1 Organizational structure water heater factory. . . . . . . . . . . . . . 14 3.2 Compact water heaters (NIBE, 2020). . . . . . . . . . . . . . . . . . . 14 3.3 An overview of the water heater factory (NIBE, 2020). . . . . . . . . 16 4.1 An illustration of the bull-whip effect. . . . . . . . . . . . . . . . . . . 22 4.2 The Kingman Formula (Holweg et al., 2018). . . . . . . . . . . . . . . 25 4.3 Illustration of Kanban and CONWIP system. . . . . . . . . . . . . . 28 4.4 14 principles of The Toyota Way (Liker, 2004). . . . . . . . . . . . . . 29 4.5 An example of a Value stream map (Rother & Shook, 2003). . . . . . 37 4.6 Example of common VSM icons (Braglia, Carmignani, & Zammori, 2006). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 5.1 Illustration of the production flow of the copper mantles. . . . . . . . 47 5.2 Illustration of the production flow of the steel mantles. . . . . . . . . 48 5.3 Illustration of the production flow of the copper gables. . . . . . . . . 49 5.4 Illustration of the production flow of the steel gables. . . . . . . . . . 50 5.5 Illustration of the production flow for the stainless steel vessel. . . . . 51 5.6 Production flow for the front plate. . . . . . . . . . . . . . . . . . . . 52 5.7 Production flow for the side and top plate. . . . . . . . . . . . . . . . 53 5.8 Production flow foot frame. . . . . . . . . . . . . . . . . . . . . . . . 54 5.9 Production of the back plates. . . . . . . . . . . . . . . . . . . . . . . 55 5.10 Layout of the painting line. . . . . . . . . . . . . . . . . . . . . . . . 56 5.11 Layout of the final assembly. . . . . . . . . . . . . . . . . . . . . . . . 57 5.12 Current state map for the welding shop and the final assembly. . . . . 60 5.13 Current state map for the metal sheet and painting shop and the final assembly. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 7.1 New suggested layout. . . . . . . . . . . . . . . . . . . . . . . . . . . 73 7.2 Future state map for the welding shop and the final assembly. . . . . 80 7.3 Future state map for the metal sheet and painting shop and the final assembly. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 B.1 VSM Copper Mantle. . . . . . . . . . . . . . . . . . . . . . . . . . . . III B.2 VSM Steel Mantle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV B.3 VSM Steel Gable. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V B.4 VSM Stainless Vessel. . . . . . . . . . . . . . . . . . . . . . . . . . . . VI B.5 VSM Top plate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII xiii List of Figures B.6 VSM Side Plate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII B.7 VSM Back Plate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX B.8 VSM Front Plate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X xiv List of Tables 4.1 Explanation of strategies used to counteract the bullwhip effect . . . 23 4.2 Explanation of data used for VSM . . . . . . . . . . . . . . . . . . . . 36 6.1 Planned average utilization for the operations at the welding shop. The standard deviation is measured in percentage points (pp). . . . . 64 6.2 Planned average utilization for the operations at the metal sheet and paint shop. The standard deviation is measured in percentage points (pp). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 6.3 Parameters for the as-is production situation at the welding shop. . . 67 6.4 Parameters for the as-is production situation at the metal sheet and painting shop. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 7.1 Current state vs Future state. . . . . . . . . . . . . . . . . . . . . . . 74 7.2 Comparison of the future state map and the current state map at the welding shop. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 7.3 Comparison of the future state map and the current state map at the metal sheet and painting shop. . . . . . . . . . . . . . . . . . . . . . . 79 xv List of Tables xvi 1 Introduction The following master thesis is conducted on behalf of NIBE AB and in collaboration with AFRY AB. AFRY AB is a consultancy firm with many areas of expertise. In this project, the department of production and logistics has been involved and pro- vided valuable insights and guidance. In the following section the background of the research is covered as well as the problem statement that initiated the research, the aim of the thesis, and delimitations. 1.1 Background NIBE Industrier AB is a global group that operates within the energy and heating industry. The group is divided into three different business areas - Climate Solutions, Element and Stoves - and produces and develops a wide range of energy efficient climate solutions. From the beginning the company has been driven by a strong cul- ture of entrepreneurship and responsible business operations with a strong focus on growth and the strive to grow has lead to investments in product development and acquisitions. Combined these factors have generated a turnover of over 25 billion SEK in 2019 and the goal is to reach 40 billion SEK by 2025 (NIBE, 2020). To reach this goal NIBE are planning to double their production capacity in Markaryd, both by increasing the efficiency of the current production but also through expansion. NIBE’s long-term strategy is to produce world-class energy solutions that contribute to a sustainable society. To reach this strategy they have four strategic focus ar- eas; growth with profitability, innovation, market oriented expansion, and long-term approach (NIBE, 2020). These strategic focus areas lead to a production that con- stantly has to face an increased demand, as well as a need to be able to produce innovative new products. This creates a demand for both higher volumes and an increased variety of products. According to Lewis and Slack (2017) higher volumes and increased variety often result in increased production costs due to the increased need for flexibility. To achieve profitable growth, NIBE has an average growth target of 20 percent annually. To achieve this, they strive to have a 10 percent organic growth and 10 percent growth through acquisition. The organic growth is done by continuous de- velopment of new products and regular product releases. To be able to develop and produce new products there is a continuous development of current operations and 1 1. Introduction a high investment rate. Moreover, NIBE emphasizes an effective production and joint purchasing of material for all companies within the NIBE group (NIBE, 2020). According to NIBE, an effective production is built around the idea that everything can be improved and that operations that are not measured cannot be improved. Time-measurements is therefore a commonly used tool to create correct calculations and production plans as well as sound investment plans. The measurements also lay ground for NIBE’s performance-based salary system, where employees get payed based on the amount of products they produce. According to production managers at NIBE, the performance-based salary system leads to operators wanting to produce as much as possible, thereby creating a push flow where components are pushed to the next activity, creating buffers between ac- tivities. Another consequence caused by this salary system is that all the operators mainly look after their own personal benefits which ends up in that the production flow can be said to be divided into different silos. Within each silo each activity is planned in isolation without taking other activities into consideration, leading to either big buffers or shortage of material between activities. In the long run this leads to transpositions in the production plan, thereby they are producing products that are not demanded downstream the production flow. Further, the buffers be- tween activities have increased the need for storage and has thereby taken valuable space in the factory that otherwise could have been used for production, harming the efficiency of the production. An additional reason to the build up of buffers is due to the many different products being produced since there are several different product families, each one contain- ing a number of different product variants. This entails that a large number of different products are in the production flow at the same time, hence creating a rather complex production system. Currently NIBE has no complete overview over the production systems, e.g. how long the lead-times are, what the bottlenecks are and where these are located, or how big the different buffers are. This indicates that NIBE lacks a system perspective over the production which makes it complicated to see how different improvements affect the whole production flow. Further, it makes it hard to know where to focus when investigating how to improve the production to meet the increasing production rate. According to Bellgran and Säfsten (2010), a system perspective is a holistic perspective that takes all the different parts of a production system into consideration and emphasise the the interplay between the different parts. Further, the authors explain that a system perspective is applicable when an organization wants to increase the understanding of a complex production system. When evaluating an existing production system, a system perspective is important since a comprehensive view of the system is only accomplished if the entire system is evaluated (Bellgran & Säfsten, 2010). According to the authors, a process flow analysis explains feasible improvements of a production system. As stated earlier, NIBE lacks a complete overview of thei production systems and it was mentioned by a manager from the logistics department that they suffer from sub-optimization 2 1. Introduction due to their tradition of using a functional layout in the water heater factory. To counteract that improvements are of local character, and instead take the whole system into consideration, value stream mapping is a commonly used tool (Rother & Shook, 2003). A value stream map (VSM) is a type of self-assessment tool, and is a visual representation of all the activities in the process - from material to informa- tion flow. By visualizing the whole flow it is possible to see how value is added and see the different sources of waste (Rother & Shook, 2003). Therefore, this tool seems suitable to use in order to increase their understanding of the existing production process, identify sources of waste, and suggest measures to eliminate these. 1.2 Problem statement As a result of NIBE’s goal to annually have a 10 percent organic growth the pro- duction constantly has to be able to scale up and increase the capacity. And, as previously mentioned, NIBE is planning to double the production volumes in their factories in Markaryd, both by increasing the efficiency of the current production but also through expansion. This has made NIBE realize that they need to analyze how they produce their products. As stated in section 1.1, NIBE lacks a com- plete overview over their production process in the water heater factory. They are therefore interested in evaluating these areas in order to get an overview over the production and get potential ways of removing waste and improving the production. Also, a more effective production can lead to sustainable benefits since a reduction of overproduction and inventories lead to reduced use of resources and therefore a more sustainable development of industry, innovation and infrastructure. This to be able to meet their objectives and achieve a cost-efficient and sustainable production. 1.3 Aim The aim of this thesis is to investigate the chosen production flow for the Compact family and provide NIBE with an overview of all activities and suggestions regarding how to increase the capacity of the production. The company and its products are further described in section 3. To guide the thesis and make sure the aim is fulfilled the following research questions will be answered: • What is the current state for Compact water heater production pro- cess regarding capacity, bottlenecks and variation? • What improvements can be made to cope with the identified im- provement potentials in the current production flow? 3 1. Introduction 1.4 Delimitations This master thesis project shall be finished within a time frame of 20 weeks, hence the scope of the project will be limited. The study will only focus on one of NIBE’s product families, the Compact water heater family, and the production flow related to these products in the water heater factory. The decision to solely study the flow of one product family has been taken in consultation with NIBE since it is considered that the chosen products’ flow are representative for NIBE’s other products. The analysis and conclusions can therefore in a larger extent be looked upon as relevant for other parts of their production facilities as well. 4 2 Methods This section contains information and explanation about the used research method, as well as a short description regarding how the value stream map was developed and how it was analysed. Holme and Solvang (1997) illustrate methodologies to describe and solve problems in order to generate new knowledge. The authors express that what methodology to use depends on what type of information that is going to be examined. Accord- ing to Williams (2007) quantitative and qualitative research methods consider the different claims to knowledge and are designed to address a specific type of research question. The author states that quantitative methods are aimed at providing an objective measure of the current state, while qualitative methods on the other hand are characterized by explore and better understand the complexity of a situation. 2.1 Quantitative research methods Quantitative research methods include collection of data in order to be able to quan- tify information and be subject to statistical treatment (Creswell, 2003). These re- search methods often involve collection of numeric data and require mathematical models and tools to analyse the data. According to Williams (2007) quantitative research begins with a problem statement and the research should include formation of a hypothesis or research questions, a literature study, and a quantitative analysis of the collected data. Leedy and Ormrod (2001) classify quantitative research methods into different clas- sifications; experimental, casual comparative, and descriptive. When applying an experimental research method Leedy and Ormrod (2001) bring up three different types of exploratory approaches; pre-experimental, true experimental, and quasi experimental. The true experimental approach contains a systematic procedure to quantitative data collection involving mathematical models in the analysis, whereas the pre-experimental and quasi-experimental designs include study participants that are non-random respectively randomly selected. Because of this the validity of the results may be lower for the quasi-experimental method due to the limited control. In a casual comparative research, the aim is to examine how independent variables are affected by dependent variables and therefore contains of relationships regarding 5 2. Methods cause and affect among the variables (Williams, 2007). This method allows the op- portunity to examine the interaction between independent variables and how they influence dependent variables. The authors describe descriptive research method as a basic method that examines the current situation and describe the as-is situation based on an observational ba- sis, or the investigation of correlation between two or more appearances. And since the aim of this project is to examine one of NIBE’s product flows one could argue that this is a suitable method to use when describing the as-is situation. 2.2 Qualitative research methods Qualitative research methods build their premises on inductive reasoning rather than deductive reasoning (Williams, 2007). What initiates qualitative research involves determined use for explaining, describing, and interpreting the data that have been collected. According to Creswell (2003) qualitative research enables the researcher to develop a high level of detail due to the possibility to be highly involved in the actual experiences. This is one distinct difference compared to quantitative research where the researcher is outside of the situation investigated. There are five different classifications of qualitative research; case studies, ethnogra- phy studies, phenomenological studies, grounded theory studies, and content anal- ysis (Williams, 2007). Creswell (2003) describes how these methods meet different needs, and in what situations they are more appropriate to use. A case study is defined by Creswell (2003) as "researcher explores in depth a pro- gram, an event, an activity, a process, or one or more individuals", and Leedy and Ormord (2001) state that case studies are used to increase the knowledge about a quite unknown or poorly understood situation. According to Yin (1999), case stud- ies are used to explain, describe, or explore causes in an everyday context in which they occur. Further Crowe et al. (2011) explain that case studies create a good understanding around the questions how, why, and what. Hence, one could argue that this is a suitable research methodology since the aim of the research topic is to increase the understanding of a product flow in order to evaluate its effectiveness and possibly suggest some improvements on how to improve the flow. It is also stated in literature that case study is a suitable method when applying a descriptive research method as stated in section 2.1 2.3 Data gathering The data collection for a case study is broad, hence Crowe et al. (2011) stress the importance of applying multiple sources of evidence, such as observations, on site visits, and interviews, in order to develop a deep understanding about the situation. 6 2. Methods 2.3.1 Literature review In order to gain relevant knowledge about the subjects that were investigated and to prepare for further data gathering a literature study was conducted. Related literature were collected primarily through Google Scholar, Chalmers Library, Uni- versity of Gothenburg’s library, and information from NIBE. Key words that were used during this process are value stream mapping, lean production, and production planning. By conducting a literature study not only will the researchers be better prepared for any future interviews and site visits, but it will also increase the thesis’ validity since the deeper understandings will lead to more thorough discussions and by that better conclusions could be drawn. 2.3.2 Interviews In the early stages of the research unstructured interviews were held with people within NIBE’s supply division to gain understandings about the organisation and the challenges that NIBE are facing within the research topic. Unstructured in- terviews are characterized by being more of a discussion rather than being carried out by prepared questions (Creswell, 2009). This leads to more flexibility since the interviewers does not have the control over the discussion and is preferable in early stages of projects when the researchers does not have enough knowledge and exper- tise within the research topic. Later on in the project semi-structured interviews were applied to enable larger data collection (Creswell, 2009). Semi-structured in- terviews are chosen since they allow the interviewers to direct the interviews to ensure that relevant aspects to the research topic are brought up, but it will also allow the interviewees to explain their meanings, and by that increase the crucial in-depth understanding (Yin, 1999). The questionnaire can be seen in Appendix A. Interviews were carried out with six persons with relevant knowledge regarding the product flow that was investigated. The interviews were held with production planning, managers and foremen from the different production departments, i.e. welding, metal sheet and painting shop, and final assembly. The focus during the interviews with the responsible personnel was on how well the production line is performing and what they experience as the strengths and weak- nesses of the current production. For the interviews with the production planning the focus was on the organisation’s planning processes and forecasting, and how it is related to the studied production flow. Lastly, when all the collected data from the interviews were compiled and analyzed it was discussed with experts to collaborate possible hypothesis and conclusions. All interviews except one was held face-to-face. When interviews are one of the 7 2. Methods main sources of information physical interviews are preferred over telephone inter- views since it provides better quality data (Knox & Burkard, 2009). All interviews were recorded upon permission since it enabled the interviewers to listen multiple times and therefore enable better analysis. This also reduced the risk of distorting the interviewees answer to question which can happen when the interviewer them- selves writes down the answers to questions (Bell, Bryman, & Harley, 2019). Besides the interviews a number of meetings were held with some of the persons that previously were interviewed and with employees with knowledge of production data. These meetings were held to clarify things and to get an insight on how the data for the production is gathered and measured. 2.3.3 Site visits and observations Site visits at NIBE’s facilities in Markaryd was conducted in connection with the scheduled interviews, and allowed the researchers to observe the current product flow that was investigated in order to get a deeper understanding and a more com- prehensive view of the situation. During the site visit a guided tour was given by a person with extensive experience from the different departments. This tour gave the researcher further understanding of the production flow. Site visits also offers the possibility to interact with people close to the production line to better understand the activities in the flow and how they are performed. 2.3.4 Quantitative data gathering To create a complete overview of a certain situation both quantitative and qualita- tive data can prove to be useful (Davidson, 2019). According to the author, relying on assumptions or precedence can lead to mistakes. Therefore, data-driven decision making is important if the organization strives to cut costs by understanding how processes work and how they perform. Quantitative data can come from a various amount of sources such as experiments, surveys, or observations (Davidson, 2019). During the project the researchers need to have access to NIBE’s production plans and follow-up systems to be able to calcu- late cycle times etc. Due to the performance-based salary system used at NIBE the researchers will ensure that the given data is accurate by comparing the data with fresh data measured by people at the line to ensure verifiable and testable results. 2.4 Quality criteria According to Bell, Bryman, and Harley (2019) quality criteria in business research makes it possible to evaluate if the results from the study are valid, reliable and 8 2. Methods replicable. Validity is related to whether the conclusions from the study are trustworthy and how well the study studies what it is supposed to study (Bell et al., 2019). Ac- cording to Yin (2003) validity relates to the integrity of the results generated from the study. He describes validity as how well a measuring measures the concept it should be measuring. Bell et al. (2019) explain that it exist four types of validity; measurement, internal, external and ecological. Measurement validity is related to whether a measurement entails what it is supposed to entail. Internal validity is re- lated to results about cause and effect from the study. External validity is connected to whether results can be applicable to other contexts. Lastly, ecological validity is related to whether the results of the study are applicable in real life (Bell et al., 2019). In NIBE’s factories the production system is rather complex, meaning that the pro- cess times for different models within the product family varies widely and this may effect the measurement validity of the research. In order to ensure the validity of the research, time-measurements that NIBE performs for the different processes will be used. These measurements are performed by specially trained personnel, are contin- ually updated and lays ground for the performance-based salary system. Thereby, the risk of bias from NIBE when performing the measurements is minimized. To secure the internal validity of the research, continuous dialogue and discussion with the factory management will be held. Since the study is conducted at NIBE, the analysis and conclusions are related to improving the current situation at NIBE. This might imply that the provided suggestions might not be useful and applicable in other contexts. There are however some theoretical and general areas like value stream mapping that can be used in other contexts and thereby ensuring the exter- nal validity. Since the aim of the research is to provide NIBE with an evaluation and improvement plan the study is applicable in a real life context. To further make sure the ecological validity is assured, interviews and discussions are performed in social and natural settings for the managers and operators, e.g. conference and meeting rooms within the factory. Reliability is connected to whether the findings from the study are repeatable, mean- ing that the next time the study is performed the outcome and the results will be the same (Bell et al., 2019). The reason for achieving reliability in a study is to make sure that errors and biases are minimized (Yin, 2003). According to Bell et al. (2019) the reliability of a study can be increased by using triangulation, which they describe as a process of using multiple sources of data when searching for in- formation and thereafter comparing and analyzing the different sources. For the following thesis the reliability was of specific focus during the creation of the value stream map. To ensure that it was reliable, the researchers started by performing observations and mapping their own process map, these were thereafter compared and any deviations was broken down in order to reach a consensus regarding the maps. Thereafter, a value stream map was jointly developed, and to be validated it was shown and discussed together with managers and operators ensuring that it was reliable. Further, the literature review also consisted of triangulation by using 9 2. Methods multiple sources of information. To make sure the interviews were reliable a standardized interview guide was de- veloped. The chosen participants were managers and foremen, and foremens were chosen together with the managers. To get at wider perspective, but also a more de- tailed overview of the current situation, interviews were held with participants from all the different departments. Discussions were also held with less experienced em- ployees, but which were considered to be able to give additional and valuable insights. All interviews were recorded so the researchers had the possibility to add valuable insights and facts to the notes that were taken during the interviews. Thereafter, the transcript and notes were analyzed and compared in order to reach consensus and form conclusions from the data. 2.5 Ethics When performing research there is always an aspect of ethics when involving people. According to Bell et al. (2019) there is four main areas of ethics in business research: • Harm to participants - Assuring that the participant is not harmed in any way by being involved in the research, e.g. physical harm and stress. • Lack of informed consent - Making sure the participants are given the correct information in order for them to evaluate if they want to participate or not. • Invasion of privacy - Assuring that the privacy of the participants are secured. • Deception - Making sure that the study is characterized in the correct way, e.g. not misleading participants. The data collection methods involving people in this research were related to obser- vations and interviews. To assure that the ethical aspects regarding the observations were fulfilled, the people working at the chosen production line were informed by their manager about the observations beforehand and the specific purpose for the observations. When a specific activity were observed the operator performing the activity were asked if they had any objections regarding the observations. Another aspect that had to be taken into consideration was the performance based salary system. To not restrict the integrity of the operators, no time measurements from the researchers were conducted during the project due to that this is very sensible for the operators since they are evaluated depending on how well they perform. For the interviews the ethical aspects were fulfilled by giving the interviewees cor- rect information about the purpose of the research. They were further asked if they were okay with the interview being recorded and told that their answers would be anonymized in order to assure their privacy, allowing the participants to speak freely. 10 2. Methods 2.6 Current state map Through the gathered data and information, a value stream map for the Compact was conducted. The authors used VSM as a tool to visualize the as-is situation for the production flow, but also to generate a improvement plan through a to-be situation with proposed improvements. The current state map was drawn with support from the collected data from inter- views, observations, and data from the ERP system. The data collection for the current state map was collected from the final assembly to the starting point of the production flow. During the data collection phase, current state maps were drawn by hand while walking along the flow. After this phase when all the empirical data was collected, the map was validated by people with great knowledge about the production flow. Thereafter, with the feedback from the validation, the map was further analyzed for improvements and the above mention procedure was repeated before the final current state map was completed. To gain further understanding of the production flow, and to induce more param- eters that could be analyzed and evaluated for the future state map a software called Flow Planner was used. Flow Planner is a tool for creating material flow charts within AutoCAD, and calculate material handling costs, distances, and time for different factory layouts. This provided the researchers with valuable insights about the current situation, and also gave an opportunity to evaluate future layouts. 2.7 Future state map The future state is mapped with the purpose of eliminating waste and to produce according to the customer requirements, as stated in section 4.13.2. The map was drawn after analysing the finalized current state map, together with the data col- lected through the literature study. The analysis of the current state map was conducted with the intention of increase the available floor space in the factory by reducing the inventory levels, and identify potential bottlenecks. To come up with potential improvements for the future state map, the map was drawn by consulting with people within NIBE that holds great knowledge about the current production flow and with great knowledge within the subject. Further, the questions presented in section 4.13.2 were answered. The finalized future state map is presented in section 7. 2.8 Analysis tool In order to find suitable improvement suggestions an analysis over the current state map was performed. The analysis was aimed at identifying problems within the 11 2. Methods current production facility. The problem identification was mainly done through interviews and observations. When the problem identification phase was done and a number of problems were identified, the researchers focused on how to cope with these identified problems. The search was done by analyzing literature and thereby finding potential ways to solve the problems. This resulted in potential solutions for the different problems and thereafter these solutions were discussed with employees within NIBE. This enabled the researchers to see if the solutions would be applicable and gave further insight on how the solution would affect the production system. Once appropriate solutions were chosen for the different problems, a future state map was conducted as well as recommendations on how to approach the problems. 12 3 Company description NIBE AB is a part of the NIBE group, which provides energy solutions for both private and commercial use. Accordning to NIBE (2020) their vision is to create world class sustainable energy solutions. NIBE AB is divided into three different business areas; NIBE Climate Solutions, NIBE Element, and NIBE Stoves (NIBE, 2020). In the following the chapter a more detailed description of the company will be presented. 3.1 NIBE Industrier AB NIBE Industrier AB acts as parent company within the NIBE group, a global group that consists of 135 different brands worldwide with 17.000 employees. In 2019 the group had a total turn over of 25 billion SEK (NIBE, 2020). To be able to handle all these brands the group is, as explained above, divided into three different business areas. The group has four financial objectives. These are for each year to reach a operating margin of 10 %, a growth rate of 20 % where the aim is that half of this growth should be organic and half through acquisitions. The two last objectives is to have a return on equity of 20 % and that the solidity not goes below 30 % (NIBE, 2020). These goals are the same for all the different brands within the group, in- cluding the company that are subject for this case study NIBE AB. 3.2 NIBE AB NIBE AB, from now on refereed to as just NIBE, is located in Markaryd in Sweden. The company operates three factories where they manufacture water heaters, heat pumps, and stoves. The product family that has been chosen is NIBE Compact, a water heater that belongs to NIBE Climate solutions. The organizational structure of the water heater factory is presented in Figure 3.1. It is the factory manager that has the overall responsibility that the production plan is followed. To support the factory manager, there are three department managers that have responsibility for sheet metal, welding, and assembly respectively. Below these department managers there are foremen that have the responsibility for the operators at each activity in the flow. The factory also has support by production technicians and quality technicians that belong to other departments within the organization. 13 3. Company description Figure 3.1: Organizational structure water heater factory. 3.2.1 Product description The chosen product family Compact is a group of water heaters aimed for household consumers and can be ordered in three different materials; stainless steel, copper, and enamelled. Each of these three variants occurs in different sizes. The copper heater can be chosen in the sizes 100, 200, or 300 litres. The stainless steel heater occurs in 150, 200, and 300 litres, and the enamelled heater in 200 or 300 litres. The copper and stainless steel models are all manufactured in-house while the enamelled models are purchased from a supplier. All models consist of a cabinet with the vessel inside, which consists of two gables, and one mantle. An illustration of a water heater is presented in Figure 3.2. Figure 3.2: Compact water heaters (NIBE, 2020). 14 3. Company description 3.2.2 Production system The production at NIBE is planned by the company’s sales department and their retailers. The company use a push-based production system where the majority of the orders are pushed out in the system without any customer order. It was stated by a manager at NIBE that "We always sell everything we produce". The company therefore use a make-to-stock (MTO) approach, and to plan the production NIBE use a Material requirements planning-system (MRP). NIBE provide the system with sales forecasts, then the MRP-system estimates when the production shall be ini- tiated and when material need to be ordered. The production plan is said to be locked three weeks before the production starts, but since NIBE see themselves as very flexible and always want high customer satisfaction changes could appear even after the set release time fence. As stated in section 1, NIBE use performance based salaries. According to the com- pany, this system allow the company to have full control over how the operators’ working hours are disposed, as well as it give the operators an incentive to develop existing and current methods and be as efficient as possible. But, as stated by NIBE, this approach also contributes to a lot of minor buffers between processes since every activity work independently and the operators want to produce as much as possible. The independecy further implies that they lack a holistic perspective over the production flow which cause a mismatch between the different departments, hence build ups of larger inventories between the departments is the major result. 3.2.3 Quality and maintenance The quality department at NIBE use three variables when measuring quality; qual- ity deficiency costs, the number of products that passes without any remarks, and the number of products that have defects that need to be remedied. The produc- tion quality is documented by on-site observations and controls at each production station. When defect products are observed they are moved to a control/repair station where the products are checked and repaired if possible, otherwise they are discarded. To ensure high quality, quality observations are conducted at several occasions throughout the production flow. It is stated "Quality checks are performed at those stations that if the problem is not identified and solved at this point, we won’t be able detect it and fix it later on" (Personal communication, 2020). The overall quality objectives that the company work towards is set by a business council and are the same for the whole NIBE group. These objectives are then decomposed to supplier quality, production quality, and customer quality. All quality areas are thereafter decomposed further down to the different department within each factory. The maintenance work at NIBE is performed by the maintenance department and takes place on an ongoing basis. The planning is governed by the manger at the department, and the work is then performed by maintenance technicians. The op- 15 3. Company description erators have personal responsibility to keep order at the work station, in agreement with 5 Nibevanor (5NV), and to report wear and occurred errors. 5NV are five principles that says that everyone has a responsibility over their work place so that it is clean, efficient, and safe. According to NIBE this create discipline and prevent that errors occur. The 5NV are listed below: • Sort out and recycle what is not needed. • Organize the workplace so it is easy to use. • Clean the workplace regularly. • Create and visualize routines. • Document and follow-up. 3.3 Factory description The water heater factory is divided into three main departments: metal sheet and painting shop, welding shop, and the final assembly. An illustration of the factory is presented in Figure 3.3. The metal sheet shop is illustrated in red, the paint shop in orange, the welding shop in blue, and the final assembly in green. Figure 3.3: An overview of the water heater factory (NIBE, 2020). At the metal sheet and painting shops, parts are manufactured and painted for the majority of NIBE’s facilities in Markaryd, it thereby produces parts for NIBE’s whole product portfolio. At the welding shop, the vessels for the water heaters are manufactured and welded together, and at the final assembly parts from the above mentioned departments are sourced and assembled together into the finished products. For a more detailed description of the current production flow, see section 5. 16 4 Theoretical framework This chapter presents the theoretical background for this thesis, and describes a num- ber of theories and models that have been used in order to execute a successful project. The chapter also aims to give a deeper understanding of the subject of the thesis, as well as creating a framework for the analysis and conclusions. 4.1 Order Planning Order planning is the planning level that relates to material supply to ensure that raw materials, purchased products, and semi-finished goods are purchased and man- ufactured in time and right quantities so that the production schedule can be held and carried out, and the result is material plans for manufacturing and purchasing in order to ensure the demand (Jonsson & Mattsson, 2009). According to the authors order planning executes plans established at a strategic and operational level within a company, and somewhat simplified one could say that the main purpose with or- der planning is to establish the quantities needed at the right time to ensure flow of material as efficiently as possible when considering tied up capital, delivery service, and resource utilization. Jonsson and Mattsson (2009) explain that considerations regarding current requirements of material and capacity must be taken into account in relation to supplies of material and capacity. Therefore, order planning must be executed from the perspective of both material and capacity. 4.1.1 Material Planning Material planning aims at balancing supply and demand of material as cost-efficient as possible (Jonsson & Mattsson, 2009). One commonly used method for material planning is Material Requirements Planning (MRP). A MRP system is a compo- sition of techniques that uses product structures, inventory data and a production schedule to calculate future requirements and ensure resource supply (Ptak et al., 2013). The fundamental principle of this method is to not schedule new orders be- fore a net requirement arise. When designing a MRP system Jonsson and Mattsson (2009) state that different parameters needs to be defined and established, i.e. plan- ning horizon, planning frequency, and types of orders. 17 4. Theoretical framework Graves (2011) discusses the consequences of changes in the material planning, and states that a change in order priority leads to additional costs since changing priori- ties certainly lead to inefficiencies in any production flow, as material gets put aside in order to facilitate the higher priority orders. To cope with the challenges that come with re-planning and the induced uncertainty, Graves (2011) suggests using time fences. Time fences are defined as the time intervals which specify the types of changes in the planning that are allowed. And as stated by Graves (2011), these time fences provide some stability, since short-term changes in the demand forecast get accumulated and then postponed beyond the the frozen time period. 4.1.2 Capacity Planning In order for a producing company to perform value adding activities different types of production resources are needed (Jonsson & Mattsson, 2016). The authors men- tions that capacity is a indication as to what extent the production resources are capable of performing value adding activities. Having a certain amount of pro- duction capacity is related to a cost, too much capacity leads to resources being underutilized and a higher cost (Lewis & Slack, 2017). Too little capacity leads to limit as to which extent the customer demand can be meet and therefore the revenue is affected (Lewis & Slack, 2017). As a result of this there is a need to balance the company’s access to capacity against the demand for capacity (Jonsson & Mattsson, 2016). The function in a company that involves balancing the access to capacity against the demand for capacity is called capacity planning (Jonsson & Mattsson, 2016). The authors further mentions that the capacity to produce is calculated or estimated for every specific group of production processes and represents a measure of how much each group can produce. The maximal capacity of a production group is the capacity that would be achieved if the production was producing non-stop, every day, the whole year (Jonsson & Mattsson, 2016). Since this is very uncom- mon the maximal capacity is often of less interest, instead the nominal capacity is often used when measuring the capacity (Jonsson & Mattsson, 2016). The authors mentions that the nominal capacity often is measured in four different variables: • Number of machines or production units within the group. • Number of shifts per day. • Number of hours per shift. • Number of working days per planing period. By using these variables the nominal capacity of the production can be calculated. The nominal capacity usually cannot be fully utilized since it almost always exist some type of lapse of capacity, e.g. breakdown of machinery, maintenance and ab- sence of personnel (Jonsson & Mattsson, 2016). By removing the lapse capacity the remaining capacity is called gross capacity. Included in the gross capacity is also different types of indirect times when production cannot be performed, e.g. waiting-time for material, time for meetings with management, etc. The remaining capacity after removing all these types activities is called net capacity and repre- 18 4. Theoretical framework sents the amount of capacity that will be able to use for value-adding activities in the planned production (Jonsson & Mattsson, 2016). The net capacity than has to be matched with the demanded capacity in order to be able to produce the right amount of products (Jonsson & Mattsson, 2016). 4.2 Theory of constraints Planning always involves a balance between what needs to be delivered and what is possible to be produced and taken from stock (Jonsson & Mattsson, 2009). This bal- ance imply that resource limitations that exist within the production system need to be taken into consideration. According to the authors, the primary limitation that has to be considered is the manufacturing capacity, but other limitations such as storage areas, transportation handling equipment, and supplier capacity are im- portant to consider as well. A well-known approach that considers capacity limitations when planning mate- rial flows is the the Theory of Constraints (TOC) approach. It originates from a method called optimized production technology (OPT) developed by Dr. Eli Goldratt (1984). This method is characterized by identifying and fully utilizing bottlenecks and subordinating the production system to these. This method was later developed into a more constraint-based method rather than just focusing on bottlenecks. According to Jonsson and Mattsson (2009), a constraint is defined as anything that harm and limits the performance of a system. Generally, a constraint take one of the following forms; physical, market, or policy. Physical constraints exists if the manufacturing capacity is lower than the demand. Market constraints occur when the demand is lower than manufacturing capacity, hence the system cannot be fully utilized. A policy constraint means, for example, that applied policies within the organization limit the capacity of the production system. The essence of TOC is that all systems have constraints. The existence of con- straints opens up the possibility of continuous improvements since, as stated above, a system always will contain at least one constraint. When a constraint has been identified one wants to synchronize the production and material flow with customer requirements (Jonsson & Mattsson, 2009). To be able to achieve this synchroniza- tion the following five steps presented by Goldratt and Cox (2004) are the core of the TOC approach: 1. Identify the constraint. 2. Exploit the constraint. 3. Subordinate everything else. 4. Evaluate the constraint. 5. Go back to the first step. 19 4. Theoretical framework To find the limiting resource in the flow it is important to first only consider the flow of material and understand what is demanded on the market disregarding the available capacity (Jonsson & Mattsson, 2009). According to the authors this is the only way to find out if any throughput-limiting resources exist and where they are located. The next step is then to focus on the capacity in order to maximize the utilization of the capacity at the bottleneck and adjust the flow of material to the extent capacity is available. According to Rahman (1998) the TOC approach can be controlled by the Drum-Buffer-Rope (DBR) methodology. This method coordinates the utilization of materials with the system’s resources. This is done by the pace of the constrain, the drum, which sets the pace for the whole flow. Then buffers are placed out strategically to prevent that the constraint never lack materials and create variations in the output of the system. Lastly the rope, e.g. a CONWIP card, handles the communication and makes sure that the products are pulled into the constraint at the right pace (Rahman, 1998). 4.3 Variation and capacity utilization Bottlenecks arise for two reasons: that the theoretical capacity is not sufficient, or that the actual capacity is not sufficient. One of the main causes why capacity is not always sufficient has to do with variation. According to Holweg et al. (2018) variation constrain any process, and they define variation as a measured deviation from an expected outcome. The ability to detect and reduce variation allowing com- panies to produce better products to their customers, hence give them a competitive advantage (Loose, Zhou & Ceglarek, 2008). Variation can appear in three different configurations: quality, quantity, and time. 4.3.1 Variation in quantity and time Variation in quantity and time influence both the supply and demand sides of the operation (Holweg et al., 2018). Seasonal customer demands can entail that quan- tities can vary substantially between periods with order releases or delivery dates that can be erratic. Moreover, the process itself can agonize variation in quantity and time since the output of the process may not be stable and produce expected quantities (Holweg et al., 2018). Variation in supply and availability of materials, and the work times for performing certain activities are further sources of variation explained by the authors. 4.3.2 Variation in quality Traditional quality control within manufacturing focuses on statistical process con- trol when detecting deviations based on product and process management (Loose, Zhou & Ceglarek, 2008). A drawback with thos method is that is that it not provides guidelines to identify the source of the variation. Holweg et al. (2018) highlight the 20 4. Theoretical framework importance of being able to distinguish random variation in the production from an transferable cause for a quality problem. They present a bunch of key quality tools that can be adopted by manufacturing processes, these are: Pareto diagram, Check sheet histogram, and Fish bone diagram. They have been useful when analyzing and controlling quality in many different settings. 4.3.3 Variation and capacity Holweg et al. (2018) state that it is essential to understand what variation does to a process’ output. Variation harms the capacity of any process, e.g. quality problems need time and resources to be fixed. Variation in quantities and time may starve or block stages within the process, and to cope with such variations the authors explain that more space, inventories, and labour may be needed. The search for capacity is fundamental in operations management, since capacity is a mean to ful- fill customers needs (Holweg et al., 2018). But the exact capacity needed is seldom known exactly, and the same applies to the capacity available since many aspects influence the calculation (e.g. bottlenecks, product mix, variability in quantity and time). The capacity of a process often varies between the different stages, and every stage’s capacity runs the risk of being the constraint of the entire process (Holweg et al., 2018). Further it is stated that not only is the capacity of the individual stages that can harm the total capacity, but also how the interaction between the stages works due to the variations that each stage is subject to. 4.4 The bull-whip effect The bull-whip effect is a phenomenon that is one of the most common problems in companies supply chains. Singh (2018) defines the bull-whip effect as a concept to describe fluctuations and inefficient asset allocation due to demand changes in the supply chain. Companies try to forecast demand by collecting a significant amount of raw materials and resources in order to satisfy the customer requirements (Sales- academy, 2018). However, while going up the supply chain variations tend to be amplified, causing issues regarding time, costs, and inventory (Sales-academy, 2018). According to Jonsson and Mattsson (2009) variations in demand tend to double for every step in the supply chain. The effects of the bull-whip effect are illustrated in Figure 4.1. 21 4. Theoretical framework Figure 4.1: An illustration of the bull-whip effect. To solve the problem companies need to understand what causes that lie behind so they can counteract them (Lee, Padmanabhan & Whang, 1997). They explain that companies in various industries have found that they can control the bullwhip-effect and improve the performance of their supply chains by coordinating information and planning along the chain. According to Jonsson and Mattsson (2009) the main reasons for the bull-whip effect are: • Large order quantities. • Few large customers. • Non-aligned planning and control. • Price fluctuations and promotion. • Lack of communication and information sharing. The causes of the bull-whip effect can lead to that companies either have lack of or an excess of inventory, which can be unfavourable (Sales-academy, 2018). Lack of inventory can lead to poor customer relations due to lower order fulfillment, and excessive inventories can result in higher costs due to tied-up capital and risk of obsolete products if the demand does not increase. To counteract the causes of the bull-whip effect, companies use different strategies. According to Lee et al. (1997) these strategies can be categorized into informa- tion sharing, channel alignment, and operational efficiency. These categories are described in table 4.1 22 4. Theoretical framework Table 4.1: Explanation of strategies used to counteract the bullwhip effect Strategy Explanation Information sharing Demand information at a downstream site is transmitted upstream in a timely fashion. Channel alignment Coordination of pricing, transportation, inventory planning, and ownership between sites in the supply chain. Operational efficiency Activities that improve the performance, eg. reduced costs and lead time. 4.5 Bottleneck identification methods As stated in section 4.2, all production systems suffer of different constraints and limitations. Quick and correct identification of the bottleneck and its location can lead to improvements in the operation management by increasing the throughput, and minimizing the total cost of production (Chang & Ni, 2007). It is therefore of vast importance that the bottlenecks are detected so the capacity utilization can be maximized. According to Law and Kelton (1991) there are two widely used meth- ods to detect bottlenecks in manufacturing systems, either by measuring the average waiting time in front of a machine, or by measuring the time a machine is active. These two methods will be presented more in detail in the following subsections. 4.5.1 Average waiting time When measuring the average waiting time, the machine where the products spend the most time waiting in front of the machine is considered to be the bottleneck in the system (Roser, Nakano & Tanaka, 2001). Waiting time cam be decided both by measuring the queue length or pure waiting time, hence both momentary and average bottlenecks can be found by comparing the queue lengths or waiting time (Roser et al., 2002). The method has received criticism since it only consid- ers the processing machines in the system, therefore aspects such as operators and automated guided vehicles (AGV) may not be considered at all or need additional considerations. Further, the available space to place the products in front of the ma- chine is often limited, and the product supply needs to be balanced comparing to the available machining capacity to ensure that the buffers are not permanently filled up. 4.5.2 Utilization rate When applying this method the machine with the highest workload is considered the bottleneck (Roser et al., 2001). However, many machines within the system may have very similar utilization rates which makes it hard to for sure determine which one that is the bottleneck (Roser et al., 2002). Furthermore, this method is only suitable for steady state systems, and needs data from a long period of time to determine average bottlenecks. On the other hand this gives the possibility to 23 4. Theoretical framework quickly locate the bottleneck (Lawrence & Buss, 1995). 4.6 Little’s Law Little’s Law deals with queues in a system and it is therefore a useful formula to see how the different variables affect the outcome of the system. Under steady state conditions Little’s Law says that the average number of WIP in a process is equal to the waiting/processing time for an item, multiplied with the throughput rate of the system (Little & Graves, 2008). Little’s Law can be seen below: L = λ ∗W L [units] λ [units/time] W [time] Where L is equal to the number of items in the queuing system (WIP), λ is equal to the throughput rate of the system, and W is equal to the waiting time. If there is less units than L in WIP, some processes will be starved and the throughput will decrease and if there is more units than L it contributes to an excessive inventory that is not needed to keep the processes running (Holweg et al., 2018). Little’s Law therefore provides a structure on how the different variables affect the conditions of the system, e.g. an increased capacity and output will lead to an increased WIP if the waiting time is not reduced (Little & Graves, 2008). 4.7 The Kingman formula Waiting is a common aspect in production systems and the time that customer spends waiting for the output of the system can be reduced (Holweg et al., 2018). In 1966 the British mathematician John Kingman published an article where he mentions that to reduce the waiting time, attention has to be put on the production rate, process variation, and the utilization of the capacity (Holweg et al., 2018). Kingman found that the waiting time in front of a process can be approximated accordingly: E(Wq) ≈ ( ρ 1 − ρ )( c2 a + c2 s 2 ) τ ρ = capacity utilization τ = mean service time (processing time) ca = variation of arrivals cs = variation of service times 24 4. Theoretical framework Simplified, the waiting time can be said to be equal to the product of the processing time, utilization of the process, and the effects of variation (production and demand) (Holweg et al., 2018). The authors further mentions that the faster the process, i.e. higher production rate and lower mean service time, the smaller the variability, and the lower the capacity utilization, the shorter the waiting time for customers. This relationships is illustrated in Figure 4.2 below. Figure 4.2: The Kingman Formula (Holweg et al., 2018). As can be seen in the figure, both curve A (moderate variation) and B (high vari- ation) illustrate how the waiting time can become exponential when the capacity utilization moves towards 100%. Curve A is more preferable than curve B since the waiting time is less for the different capacity utilization levels (Holweg et al., 2018). The difference between the curves is the variation, curve A has a moderate variation while curve B has a high variation, where the higher variation leads to a higher lead time in accordance with the formula. 4.8 Push and pull production In production and material planning there is often a distinction between push-based and pull-based planning. A push-based approach is defined as the manufacturing and material movement taking place without the authorization of the consuming unit, whilst a pull-based approach is defined as manufacturing and material move- ment taking place on the initiative of the consuming unit (Jonsson & Mattsson, 2009). Thereby, the main difference between the two approaches is related to the authorization of the next step in the process and movement of material (Jonsson & Mattsson, 2009). The pull-based approach is seen as the ideal state of the lean production princi- ple, Just-In-Time (JIT) manufacturing, which strives to give the customers exactly 25 4. Theoretical framework what they want, when they want it (Dennis, 2007). Dennis (2007) further mentions that the JIT production has a set of simple rules and that it is closely related to pull-based production. First, products should not be produced unless the customer orders it. Secondly, the demand should be levelled so that production is conducted evenly throughout the factory. The third principle says that all processes should be linked with customer demand through a Kanban-system to simplify the tracking of demand, and lastly, the flexibility of machines and people should be maximized. The pull system allows to control the WIP and the amount of Kanban-cards puts a maximum number of WIP in the system (Dennis, 2007). According to Dennis (2007) the control and upper limit of WIP in the system leads to: • Reduced throughput time (according to Little’s law). • Operating expenses are reduced since the finished good inventory and ordering of raw material are decreased. • The quality is improved since defect products are easier to find quickly and not produced in large batches. • Ergonomics is improved since the part bins are smaller and easier to lift. • Safety is improved when there is less trucks transporting goods in the factory. Lasa et al. (2008) mentions that a continuous one-piece flow of products should be established where possible. A one-piece flow means producing one product at a time and once the current process is done the product moves directly to the next step without any stagnation between activities. By having a continuous flow, the throughput time is reduced, the cost to cash period is shortened, and the quality can be improved (Liker & Meier, 2006). The push-based approach is when processes and activities are working on self-reliant schedules, and products are produced and pushed forward into inventory buffers between activities (Liker, 2004) The push production system has no declared agree- ment between customer and supplier in terms of the number of products that are to be supplied, and at what time that is supposed to happen. Products are therefore delivered to the customer whether he requested it or not, causing problems regard- ing the control of the process in terms of what and how to control it, e.g. are the production behind or ahead (Liker & Meier, 2006). In a traditional push-based system, the production scheduling of each department is managed individually (Rother & Shook, 2003). According to the authors, this in- creases the complexity of coordination of the production compared to only have one production unit to coordinate. By implementing a pacemaker in a bull-based pro- duction system, the upstream production can be controlled from this point, hence reducing the need of coordination. (Rother & Shook, 2003). Further, the usage of a pacemaker is beneficial when there is a mix of products in the production flow. A pacemaker controls the upstream production by sending a signal to the process in the beginning of the flow when there is a demand, and only then will the production be initiated, hence creating a pull-based production flow (Rother & Shook, 2003). 26 4. Theoretical framework One prerequisite for the downstream processes from the pacemaker need to be in a continuous flow all the way down to the finished product, according to the authors. 4.9 CONWIP CONWIP (Constant Work In Progress) is a system similar to Kanban that uses sig- nals/cards to communicate when production can be initiated (Spearman, Woodruff, & Hopp, 1990). According to the authors, CONWIP is a system that achieves all the benefits of a pull system while being suitable for production of a high variety of products. The basic principle of CONWIP is that each container is attached with a card at the start of the production, and when the container is used at the final station the card is removed and sent back to the beginning of the line (Spearman et al., 1990). At the beginning of the line the card waits in a queue until it is attached to another container based on the backlog list, where first come first served is used, thereafter the part goes through the production flow (Spearman et al., 1990). Ac- cording to Framinan, González, and Ruiz-Usano (2003), a production process cannot enter the system unless it is attached with a card. Jonsson and Mattsson (2009) further mentions that compared to a Kanban system, the CONWIP system sends the production signal from the last station directly to the first station, while Kan- ban sends the signal to the previous process. By doing this the CONWIP ensures a constant work in progress and the ability to handle a variety of products (Jonsson & Mattsson, 2009). The difference between a Kanban system and CONWIP system is illustrated in Figure 4.3. Spearman et al. (1990) mentions that one of the major advantages of CONWIP is that the flow times for products becomes fairly predictable since the WIP-levels are nearly constant. They further mentions that the coordination of the production becomes easier to coordinate when the WIP is constant and by keeping the WIP at a low and constant level the following benefits can be achieved: • Increased chances of detection of defective parts. • Less material on the shop floor, making it easier for operators to find WIP for the next process. • By reducing WIP problems becomes visible, e.g. machine failures, defects, yield losses. 27 4. Theoretical framework Figure 4.3: Illustration of Kanban and CONWIP system. 4.10 Lean production Lean production is characterized by an effective and efficient organization. The ap- proach encompasses a vast variety of practices such as (JIT), quality systems, work teams, supplier management etc. into an integrated system (Shah & Ward, 2002). The purpose of lean production is to create a synergy between all these practises in order to facilitate a streamlined system that produces exactly what the customers want at the right time and with as little waste as possible. Lean production orig- inates from Toyota Production System and builds upon the 14 principles founded by Liker (2004). The core of these principles is to eliminate waste. Waste is de- fined as all activities that does not add value to the product from the perspective of the customer. The principles are divided into four different classifications, which are illustrated in Figure 4.4. The Process classification is of particular interest for this thesis since the research doesn’t aims at questioning the company’s methods or philosophy, but to investigate the production processes and how to make these more efficient. 28 4. Theoretical framework Figure 4.4: 14 principles of The Toyota Way (Liker, 2004). Lean production focuses on the entire end-to-end value stream rather than sub- optimization of segmented activities in order to achieve high-value and efficient processes with as little waste as possible. By focusing on the entire value flow and eliminate waste, processes will be able to respond to changes in requirements and demands with high quality and agility, and low costs. Slack and Lewis (2015) made a study where they compared a conventional approach to operations and lean production, where the conventional approach assumes that each activity in the flow is independent of the activity downstream the flow. The findings from the study were that the in-dependency between the different activities caused build- ups of buffers in the flow, and the bigger the buffers, the more independent the different activities were. Lean production on the other hand allows activities to be more interdependent, hence the build-up of buffers can be reduced and the flow will become more efficient and agile (Slack & Lewis, 2015). 4.11 Waste in Lean production In lean production the main concept is to identify and eliminate waste in every pro- cess of the whole production system (Chiarini, 2012). In order to understand what waste is, it is important to understand and focus on value (Chiarini, 2012). Value is what the customers are willing to pay for a product, e.g. sheet metal being bent and welded. The opposite of value is waste, which is what the customers are not willing to pay for. One example of this is excess inventory and waiting time (Dennis, 2007). In the following section wastes in lean production will be briefly described. 29 4. Theoretical framework 4.11.1 Transportation Transportation is a waste linked to transportation of parts between processes, e.g. moving WIP from one place to another in order to process it. This also involves moving material and finished goods in and out of inventories and between processes (Liker & Meier 2006). This type of waste is often related to poor factory layout, large machinery, or ordinary batch production, creating a need for transporting goods between activities (Dennis, 2007). Another factor increasing the need for transportation is excessive inventories which cause a need to move products from one warehouse to another, or moving products from a warehouse to a production pro- cess (Chiarini, 2012). To reduce the number of transports, Chiarini (2012) suggests using VSM to redesign the layout, implement U-cells or use multi-skilled workers that can perform several activities. 4.11.2 Inventory Inventory is related to excessive inventory, finished goods and WIP (Liker & Meier 2006). The authors explain that this leads to longer lead times, cost for trans- portation and storage, obsolescence and damaged goods. Rother and Shook (2003) further mentions that excessive inventory is one of the main reasons for longer lead times. Inventory is described as products or raw material being stored within the company boundaries for a certain amount of time (Chiarini, 2012). The author further mentions that inventory is a waste connected to producing more than what is actually demanded by the customers. Liker and Meier (2006) mentions that in- ventory has a tendency to hide problems related to production, e.g. defects and unbalanced activities. It causes slow deliveries from suppliers and leads to longer setup times. Chiarini (2012) states that the best way to identify this type of waste is to observe where there is an accumulation of products and then try to understand why the inventory is stocking up so much at that place. He further mentions some of the main reasons for inventory to be: • Time consuming changeover time. • Producing according to large economic lots. • Production starts early. • Processes operating at different speed. • Processes operating inefficiently or creating defects. • Accepting excessive inventory since it means instant delivery to customers. Chiarini (2012) mentions that the last reason is specially important since accepting excessive inventory has a tendency to hide problems instead of fixing them and he therefore mentions that it is important for the company and its staff to realize that it is possible to eliminate the excessive inventory. 30 4. Theoretical framework 4.11.3 Motion Motion is related to the idea of unnecessary movement (Chiarini, 2012). Motion has two components; human motion and machine motion (Dennis, 2007). Human motion is connected to the idea of wasted time when workers has to do unneces- sary movement, e.g. look for a tool that is not close to the operation where it is needed (Chiarini, 2012). In a poorly designed ergonomic workplace Dennis (2007) states that the productivity is affected when there is unnecessary motion in terms of walking and reaching. Further, the quality is affected when the worker has to twist or reach to check a work piece and lastly the safety of the worker is affected when the workplace has a poor ergonomics design and can lead to work related injuries (Dennis, 2007). The second component of waste related to machine motion, for example if a ma- chine is placed unnecessarily far from the next machine making it an unnecessary motion for the work piece to move from one machine to the next one (Dennis, 2007). Further, Chiarini (2012) explains that some of the reasons for the unnecessary mo- tion is caused by a poor factory layout, low involvement from staff and failure to keep the workplace clean and in order. Changes that can be made to decrease the unnecessary motion is to move towards a production flow, implement 5S (Sort, set in order, shine, standardize, sustain), and design U-shaped cells (Chiarini, 2012). 4.11.4 Waiting Waiting is related to the concept of workers and machines having to wait before conducting a new activity. It can for example be a worker waiting for a part to be processed in a machine before moving it to the next step (Dennis, 2007). Dennis (2007) mentions that waiting often occurs when there is large batch production, problems with machinery downstream in the process, or when parts needs rework- ing caused by defects. Waiting also leads to increased lead times, which is the time between customer order and delivery of order (Dennis, 2007). There are several reasons for waiting, but some of the main causes mentioned by Chiarini (2012) are uneven balance between operations, lack of preventive maintenance and large batch production. The author further mentions some potential ways of removing the main causes of waiting, e.g. having a balanced and leveled production and performing preventive maintenance of machinery. 4.11.5 Overproduction According to Taiichi Ohno, by many seen as the father of Toyota Production System (TPS), overproduction is the root cause of poor manufacturing and means producing products that exceeds customer demand (Dennis, 2007). Chiarini (2012) mentions that overproduction is known as producing products even if there is no customer order. Many firms producing according to a MRP, think that eventually all the produced products in the warehouses will be bought. But there is no guarantee of 31 4. Theoretical framework this and in the meantime, the inventory ties up capital and storage area and the products faces the risk of being obsolete, stolen or damaged during the time in the warehouse (Chiarini, 2012). Overproduction comes with a number of negative effects, Chiarini (2012) mentions inventory (which is another type of waste), it decreases the flexibility of the produc- tion planning, slows down the production process and has an increase in cost caused by transportation, storage and inspection. He further mentions some potential rea- sons for overproduction, e.g. producing according to oversized economical batches, creating inventory to cope with defects and machines operating to fast. 4.11.6 Over-processing Over-processing is a profound form of waste in lean production related to the idea of performing more activities to the product than the customer actually requires (Dennis, 2007). Liker and Meier (2006) explain that this involves processing parts with steps that are not necessary. This type of waste is often found in engineering departments when the connection to the customer is lost and the product gets at- tributes that the customer do not desire. These extra attributes then causes a need for extra processing in the production leading to waste (Dennis, 2007). Some of the main reasons for over-processing is poor analysis of activities and process design, lack of standardization and inadequate material and equipment (Chiarini, 2012). He further mentions some ways of removing this type of waste, e.g. redesigning processes and activities, updating instructions and procedures and using tools such as value engineering. 4.11.7 Defects Defects are a type of waste related to damage caused during manufacturing, hence a need for repairs and fixing products occurs. This waste involves all the extra time, material and energy needed to repair the defects (Dennis, 2007). Defects is related to quality which is to do the right thing directly, and is therefore a result of bad quality and can result in dissatisfaction among customers as well as damaging the companies reputation (Domingo, 2015). Chiarini (2012) describes that some of the reasons for defects to occur is caused by: • Specification and instructions not clearly stated. • Lack of skill and knowledge among employees. • Processes operating ineffective and lacks control. • Incapable material and products. The author further mentions that to eliminate these defects companies can increase the employees’ knowledge and skills in quality work, design processes and machines with automation to detect defects or implement poka-yoke, which is a process that 32 4. Theoretical framework avoids mistakes being made. All these things can help to eliminate defects, but it is important to identify and find the root causes for every defect occurring in order to fully reduce these and increase the efficiency (Chiarini, 2012). 4.11.8 Unused employee creativity Originally when talking about waste in lean production there has been a focus on the seven waste identified by Taiichi Ohno, but lately an extra waste has been added to the list, which is about "underutilized people" (Wahab, Mukhtar & Sulaiman, 2013). This waste is about not utilizing the knowledge that the employees possess (Liker & Meier 2006). Liker and Meier (2006) state that by not involving the employees, time, skills, improvements, and development is lost. The authors further mentions that according to Taiichi Ohno the seven wastes mentioned above all have an impact on this last waste, since reducing waste displays problems within the factory and creates a need for the employees to use their knowledge and skills to solve the newly exposed problems. 4.12 Lead time reduction Lead time is defined as the time from receiving a customer order until the order is shipped (Hopp, Spearman & Woodruff, 1990). Reducing the lead time is therefore a way to increase the competitive advantage (Tersine & Hummingbird, 1995). Accord- ing to Hopp et al. (1990) shorter lead times can reduce the inventories of in-process material, reduce the frozen zones in production planning and thereby minimize the dependence on forecasts from sales. Tersine and Hummingbird (1995) argue that a major consequence of excessive lead times are problems with planning and schedul- ing, resulting in longer planning horizons and magnified inventories. They further mentions that reduced lead times can improve the quality management since it en- ables products to leave the factory quicker and thereby minimize the opportunity for products to be damaged. Hopp et al. (1990) discuss three main areas in reducing the lead time. The first one is related to throughput time and how long time products spend waiting and in queues. Secondly, the lead time and WIP are related to each other and large inven- tories leads to excessive lead times. The third reason the bring up is that lead time is affected by the variance of throughput time. The authors thereafter continues by presenting five general methods for lead time reduction: • Review WIP. An increase in WIP results in an increased lead time and an analyze is therefore necessary to understand which WIP that is necessary to cope with bottlenecks and which WIP that can be reduced. 33 4. Theoretical framework • Make products continuously move towards completion. The general idea is that if the product continuously move towards completion both the lead time and inventories will be reduced. This is largely related to the fact that products spends 90-95% of their time waiting, so by enabling a continuing movement of the product the lead time will be reduced. • Synchronize production. Since part assembly cannot be be finalized until all components are accessible, it is important to synchronize between manufacturing and assembly. By syn- chronizing it will be possible to produce according to what is needed instead of what is available. • Level the work flow. By enabling a smooth work flow the inventories and lead times can be reduced. Some easily implemented methods to achieve this are to establish a constant work flow, level the release of orders, and justify line balancing. • Eliminate variability. The main reason for variability in processes are related to downtime, rework and production methods with a lack of consistency. Some possible strategies to reduce variability are to minimize rework, analyze the variability in supplier lead time, improve the reliability of machines and processes, and plan for yield losses. Another aspect related to reducing the lead times is to not only focus on the pro- duction and operation areas, but also to understand how constraints related to the supply and demand-side affect the lead time (Tersine & Hummingbird, 1995). They explain that a complete strategy for reducing lead time should focus on all bottlenecks in the system and start with the most inhibiting one for the throughput. 4.13 Value-stream mapping As briefly mentioned in section 1.1 a value stream is the collection of activities, both value and non-value adding, that are needed to bring a product through the flows of the production (Abdulmalek & Rajgopal, 2007). By looking at the entire value stream, and not only eliminate waste at isolated points, could lead to pro- cesses that need less space, capital, and time. In addition, information management becomes more accurate and simpler to mange. One tool that are widely used to map flows, identify waste and redesign the products way through the production is Value Stream Mapping (VSM) (Lasa, Labure & de Castro Vila, 2008). VSM helps companies to understand where they are (Current state), where they want to go (Future state), and map a way to get there (Implementation plan) (Chen & Meng, 2010). According to the authors, this create a high-level perspective and look of the total efficiency, and not independent efficiencies of individual departments or production processes. 34 4. Theoretical framework The main objective while using VSM is to find and take steps towards eliminating all the wastes found in the value stream in order to minimize the amount of non- value adding activities and increase the amount of value-adding activities (Rother & Shook, 1999). Hines and Rich (1997) mentions that in order to make improvements in the value-stream it is important to have an understanding of different wastes in the flow, and VSM is therefore a suitable tool. Rother and Shook (2003) state that "Value-stream mapping is a pencil and paper tool that helps you to see and under- stand the flow of material and information as a product makes its way through the value stream". They further state that a VSM is a visual representation of the prod- ucts path through the production, from supplier to customer, involving all processes in the flow of material and information. When looking at the production flow, the flow of material is usually what is focused on, but the flow of information is just as important to look at since it express the next step for each process (Rother & Shook, 2003). Rother and Shook (2003) present four steps for value stream mapping: 1. Select a product or product family that the VSM will focus on. 2. Create a current state map. 3. Create a future state map. 4. Develop a plan on how to implement the future state map. By starting with selecting a product or product family the map becomes more fo- cused and this is important since the customers only care about the specific product that they are buying and not all the products that the company produce (Rother & Shook, 2003). This product narrowing also avoids making the map too complicated. The authors explain that a product family is a series of products that goes through the same type of processes in the downstream flow of products but with slightly different attributes and components. Both step two and three involve mapping dif- ferent states, and although the current state map is conducted first it is important to know that these two activities are not strictly after each other but has some over- lapping tendency. This means that while conducting the current state map, future state ideas might arise and likewise drawing the future state map will often result in new information that has been overlooked in the current state, and therefore there are some overlapping between these two activities (Rother & Shook, 2003). The last step is to establish an implementation plan on how to minimize the gap between the current and future state. The plan should tell what actions that are needed to move from current state to future state (Bicheno & Holweg, 2000). They further state that after the improvements from the implementation plan has been made and has achieved stable results, the process starts over again, by generating new current and future state maps and this continuous process continues to progressively move towards the vision of lean processes. On a products way through the factory it passes through several different production processes and in the VSM the product processes are visualized by product process boxes. Rother and Shook (2003) mentions that if the components for the chosen product family goes through different pr