Production System Improvement A case study of Segezha Nørresundby

Second semester project group B2-14 Global business engineering The faculty of engineering and science aalborg university 21st May 2014

The Faculty of Engineering and Science Global Business Engineering 2. Semester Niels Jernes vej 19 9220 Aalborg Ø Phone 99 40 99 40 Web: www.tek-nat.aau.dk

Title: Production System Improvement Theme: Description, Analysis, Solution Development and Assessment of a Business System Project period: P2, Spring semester 2014 Project group: GBE B2-14 Group members: Dan Bach Olesen Henrik Vinther Færch Jesper Thoftgaard Tran Line Lund Ypkendanz Magnus Nørlem Patrick Hoffmann Mørk

Synopsis: The purpose of this project was to describe and analyse a queuing system. The project took basis in Segezha A/S, located in Nørresundby. Through an analysis of the queues in the production company, the focus point became a specific production line which manufactures paper sacks. The purpose of this project became developing and applying a sequencing method to reduce the total amount of changeovers in the SANA department, and as a result reduce the order queues. Through sequencing and simulation a method was concluded to reduce the total amount of changeovers, which sequence the orders according to specifications and material numbers. It showed that the amount of changeovers could be reduced by extending the planning period.

Supervisor: Stig B. Taps Herman Lyhne Knudsen Claus Christian Monrad Spliid

Number of pages: 95 Appendix hereof: 17 Finished: 21th of May 2014

The content of the report is freely available, but publication (with source reference) may only take place in agreement with the authors.

Preface This project report is the work of group B2-14 as a second semester project of Global Business Engineering at Aalborg University. The project has been written in the period from February 2014 to May 2014. The topic for the project is queuing systems. The purpose of the project has been to study how to improve a production system by analysing waiting lines, customers and planning in an industrial context. The analysis and results of this project were developed in cooperation with the manufacturing firm Segezha Nørresundby. The target group of this project report is people with an interest in queuing systems in a production system. More precise the order process flow, and potential problems or possible improvement related to an order process flow within a production system. Acknowledgements The group would like to thank the employees and managers at Segezha Nørresundby for their collaboration and participation during the whole project. Special thanks to production manager Flemming Nørgaard, process manager Lise-Lotte Bach, and account coordinator Bente Christensen. Last but not least special thanks to Stig B. Taps, Hermann Lyhne Knudsen and Claus Christian Monrad Spliid who supervised and inspired our work with the project.

Dan Bach Olesen

Henrik Vinther Færch

Jesper Thoftgaard Tran

Line Lund Ypkendanz

Magnus Nørlem

Patrick Hoffmann Mørk

v

Contents Chapter 1 Introduction

1

Chapter 2 Empirical Background 2.1 Segezha group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Segezha Nørresundby . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 3 4 7

Chapter 3 Pre-analysis 3.1 Distribution of time 3.2 Order flow . . . . . . 3.3 Changeovers . . . . . 3.4 Sub conclusion . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

11 11 12 15 16

Chapter 4 Problem statement

17

Chapter 5 Methodology 5.1 Data Selection . . . . 5.2 Data Collection . . . . 5.3 Data Analysis . . . . . 5.4 Reliability and validity

19 19 19 20 22

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

Chapter 6 Theoretical approach 25 6.1 The theory of waiting line management . . . . . . . . . . . . . . . . . . . . 25 6.2 Planning and control activities . . . . . . . . . . . . . . . . . . . . . . . . . 27 Chapter 7 Case studies and data presentation 31 7.1 Components of the queuing system . . . . . . . . . . . . . . . . . . . . . . . 31 7.2 Sub conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Chapter 8 Solutions development 8.1 Data analysis . . . . . . . . . . 8.2 Simulation . . . . . . . . . . . . 8.3 Statistics . . . . . . . . . . . . 8.4 Capacity calculations . . . . . . 8.5 Implementation . . . . . . . . . 8.6 Improvements . . . . . . . . . . Chapter 9 Discussion 9.1 The solutions . . . . . . . . . 9.2 Consequences for the external 9.3 Comparison of the solutions . 9.4 Sub conclusion . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . . . and internal . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . . . . environment . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . .

37 37 46 50 52 53 55

. . . .

57 57 58 59 61 vii

Group B2-14

Contents

Chapter 10 Perspective

63

Chapter 11 Conclusion

65

Bibliography

67

Appendix A CD 69 A.1 Unprocessed data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 A.2 Processed data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Appendix B Segezha Organisational Structure

71

Appendix C Data processing - Changeover

73

Appendix D Presentation of queuing system data

75

Appendix E The Different formats in the SANA department

81

Appendix F Distribution of products

83

Appendix G Amount of material numbers

85

Appendix H ANOVA

87

viii

Introduction

1

This project is a part of the second semester at Global Business Engineering (GBE). The purpose of the project is to improve a production system by analysing waiting lines, customers and planning in an industrial context. Segazha Nørresundby has agreed to be a part of this project and to collaborate in order to find potential problems in their production system, which can be improved. The information regarding Segezha Nørresundby has been obtained from interviews with the production manager, account coordinator and process manager. The firm is divided into two production departments, the industrial department and the SANA department. The industrial department produces paper sacks for the building industry, the food industry, and the chemical industry. The SANA department produces paper sacks purposed for waste and recycling. The industrial department has been undergoing several improvement processes. LEAN techniques have been implemented and the technology of the production lines are of greater standards compared to the SANA department, with new machinery purchase, and therefore it is assessed that there currently is a minimal need for improvements. The SANA department has not been undergoing the same improvement processes as the industrial department. LEAN techniques have not been implemented in the same extent as in the industrial department. However, a few elements from LEAN, such as employees working across the hierarchical structure in teams, have been implemented. The technology of the production lines are old-fashioned with machines from the 1960´s, which requires manual changeovers. Nevertheless, since SANA is profitable year after year and the investor does not allow new investments, it is not possible to purchase new machines. The planning of the orders in the SANA department is being made primarily by one person, with a minimum use of planning theory. Instead of using theoretical methods the planning of the orders are based on experience, and organised in size, materials etc. Overall, Segezha Nørresundby is satisfied with how the firm and the two departments are running, but as the introduction implies, there are potential for improvements in the SANA department. Based on the above the following problem thesis was asked: How can the amount of changeovers in the SANA department be reduced? Two different approaches were suggested by the firm and the study coordinator: • To reduce the amount of changeovers by looking at the planning department and how the planning of orders were conducted.

1

Group B2-14

1. Introduction

• To reduce the amount of changeovers by looking at the actual changeovers in the manufacturing department. Since it has been set as a criteria, that the project should include theory about queuing and simulation methods, a delimitation has been made based upon, which approach was the most tangible and where the theory about queuing, planning, and simulation methods were seen as the most applicable. In continuation of the delimitation, the following initiating problem has been derived: How are the interaction between the planning process and the changeovers in the production lines within the SANA department, and which elements can be identified in order to improve the production?

2

Empirical Background

2

The data and facts presented in the following sections have been obtained from interviews with Production manager [2014] and Process manager [2014], unless otherwise is stated. The Danish manufacturing plant Segezha Nørresundby is a part of the Segezha group, which is a large Russian company. The Segezha group will be described in order to understand the structure which Segezha Nørresundby is a part of. Moreover, this will enhance the understanding of the internal and external factors, which influence Segezha Nørresundby.

2.1

Segezha group

The name Segezha originates from a city in Russia, where the firm was founded. The Segezha group is owned by InvestLesProm, which is an investment bank founded in 2006. The focus of the Segezha group is the wood industry with the aim of covering the entire value chain, from forest to finished products. The different types of products are newspaper, paper sacks, packing materials & cardboard, sawn timber and wooden houses[Appendix A.2.2, slide 3-6]. To understand the Segezha group, it is important to know the structure of the organisation. The Segezha group’s structure is illustrated in the organisation chart, figure 2.1. The group is divided in the four main areas: forestry, pulp & paper, packaging and fibre board. In this project the main focus is going to be on Segezha Packaging and more specific the factory in Nørresundby. A part of the Segezha group is Segezha Packaging. As a leading global paper sack supplier, Segezha Packaging provides integrated paper sack and filling solutions to five key sectors: building materials, human food, chemicals & minerals, agriproducts and waste & recycling [Segezha Packaging A/S]. Segezha packing has 10 Figure 2.1. Organisation chart of the Segezha factories spread across Europe and Russia. group They produce 1.2 billion sacks annually, and have sales in more than 60 countries. Segezha Packaging is a firm with 1,150 employees and a annually revenue of €250 million [Segezha Packaging A/S] [Appendix A.2.2, slide 7-17]. 3

Group B2-14

2.2

2. Empirical Background

Segezha Nørresundby

Segezha Nørresundby is located in Northern Jutland. The factory is a manufacturing firm that produces paper sacks and has more than 130 employees. In addition to the production of paper sacks, the firm is responsible for marketing and development of refuse and industrial sacks for the Segezha Packaging departments in Europe. They provide companies like Arla, Aalborg Portland and Danisco with paper sacks of all kinds [Appendix A.2.2, slide 24-28]. As mentioned in the introduction, the firm is divided into two departments, the industrial department and the SANA department. The industrial department produces paper sacks for the building industry, the food industry, and the chemical industry. The SANA department produces paper sacks purposed for waste and recycling.

1929

Plant founded under the name Bates Ventil Sække Co. ApS

1959

Starts producing paper sacks

1967

SANA department is being built

1995

Merger between Bates and DAC

1997

Bates sold to Kornäs

1999

New factory is built

2000

Production in Sweden closed and moved to DK

2002

Production in Finland closed and moved to DK

The technology at Segezha Nørresundby is different 2006 Korsnäs sold to ILP/Segezha depending on, which of the two departments are being analysed. The SANA department consists 2014 TODAY of machines from the 1960’s. Improvements and modernisation have been minimal, because the investment costs for new machines are too high Figure 2.2. Time line Segezha Nørrewhich outweigh the advantages that an automated sundby machine can provide. All the changeovers have to be made manually and as a result of this the changeover becomes time consuming. The technology in the industrial department is very different. The department has been undergoing modernisation, and production machines within automation and paper solutions have been acquired. Figure 2.2 show a time line, which visualise the history of Segezha Nørresundby from the firm was founded in 1929 and until present day [Appendix A.2.2]. The factory was founded in 1929, as Bates Ventil Sække Co. A/S. The firm started producing paper sacks used to store cement, which previously where stored in barrels. In 1959 the production of refuse sacks started. In 1967, the SANA department was build, because a need for larger machines occurred. According to the Production manager [2014], the refuse sacks market has gone through a lot of acquisitions, mergers and factory closures, which also has been the case for Segezha Nørresundby. A merger between Bates Ventil Sække Co. and DAC Emballage occurred in 1995, and the name was changed to Bates Emballage A/S. In 1997 Bates A/S was sold to Korsnäs AB, a Swedish factory under the Kinnevik Group, and the name was then changed to Korsnäs Packing A/S. In 1999 a new factory was built for the purpose of centralising the production in Scandinavia. In 2000 and 2002 the firm underwent improvement processes and the productions in Sweden and 4

2.2. Segezha Nørresundby

Aalborg University

Finland were closed and moved to Denmark. Finally in 2006, Korsnäs Packaging A/S was sold to InvestLessProm and the name was changed to Segezha Packaging A/S [Appendix A.2.2].

2.2.1

Mission

The mission is a clear way to set goals for the firm and base a strategy on[Daft, 2007,p. 58]. The mission of Segezha Nørresundby is the same as Segezha Packing, which is: “Segezha Packaging will provide our shareholders with an above average return on investments and aims to be the fastest growing paper sack company in the world. We will achieve this by providing reliable service to our customers, pursuing zero defect products, operating at the lowest possible cost and creating a safe, challenging and rewarding work environment for our employees” [Appendix A.2.2, slide 3] The mission stated, implies values such as cost efficiency, growth, precision, quality and a rich working environment for employees. Values such as precision, quality, rich working environment emphasize cost efficiency, whereas the growth is the only value which distinctively stands out. These values are opposites and can create a conflict with each other, which can result in difficulties in the decision making process[Daft, 2007].

2.2.2

Strategy

The reason for defining a strategy is to design a way to accomplish a set of goals[Daft, 2007,p. 62]. Based on the goals, which Segezha Nørresundby has listed, the strategy that fits the organisation best, is the defender strategy. The defender strategy attaches importance to such things as centralized authority, tight cost control and emphasis on production efficiency. The strategy tries to defend its current market, and do not seek to expand. This carries the advantages of being in a familiar market and not moving out on uncharted territory. It keeps investment costs down, and secures a steady flow of income[Miles and Snow, 1978]. Since the firm was acquired by the Segezha group, it has been a goal to reduce investments in new equipment, and the costs related to the LEAN concept. LEAN is a tool that can improve internal efficiency, and was used at the acquisition. Reductions of resources to the LEAN concept, does not correspond to the way the defender strategy is defined. Segezha Nørresundby has currently divided their production into two different departments. This is due to the departments operating with different technology, but also on two entirely different markets. Additionally with the defender strategy characteristics, which emphasize on not expanding their operation, the mission statement deviates from the defender strategy, since it gives a message of becoming the fastest growing paper sack firm in the world.

2.2.3

Organisational structure

Segezha Nørresundby´s current strategy affects the way the organisation is designed. Segezha Nørresundby has been analysed as a defender strategy, and the organisation structure which is most complimentary towards the defender strategy is mechanistic[Burns and Stalker, 1961]. It emphasises vertical communication, and high control span, since 5

Group B2-14

2. Empirical Background

the variety of problems which occurs are at a minimum, and already well known. All of these characteristics imply low cost, and it means that Segezha Nørresundby will be able to control their outcome more efficiently, but it does not encourage change and innovation in the workers[Burns and Stalker, 1961]. Their mission states a goal of creating a challenging and rewarding environment for the employee, which does not correspond to a mechanistic organisation structure. A section of their current organisation structure in Segezha Nørresundby is depicted in figure 2.3. The full organisation structure is shown in appendix B. Segezha Nørresundby organise the organisation by a functional departmentalisation, and a mechanistic structure. Functional departmentalisation means that Segezha Nørresundby has chosen to divide their functions in departments, so that the planning management are responsible for their own field in their respective department, while purchasing has their own department. The communication between the departments are vertical and the decision making is centralized, which means that the employees reports to their superiors, and it is the management which makes all the decisions [Burns and Stalker, 1961].

Figure 2.3. Section of the Segezha Nørresundby Organization Chart B

2.2.4

Effectiveness

A departmentalisation also occurs in the production level. As previously mentioned in this chapter, Segezha Nørresundby is divided up in the industrial department and the SANA department[Figure2.3]. The technology improvements in the industrial department have been significant, which has increased the output and automated the whole production. This has clear linkages to their efficiency goals and accomplishes to fulfil them. The SANA department is not able to acquire any new technology without investments exceeding the budget, and therefore not able to accomplish the mission statement this way. Since the investment costs outweighs the advantages of getting an automated machine, the only way to improve it is by increasing the total throughput time of the current machines, according to the Process manager [2014]. Even though their strategy and organisation design does not live up to their mission, and deviates from the theoretical characteristics the SANA department still has continuous profit.

6

2.3. Environment

2.3

Aalborg University

Environment

The external environment is every element outside an organisation [Daft, 2007,p.138]. This section will analyse the elements which are outside Segezha Nørresundby and are affecting the firm. This will be done to improve the understanding of the conditions, which Segezha Nørresundby is operating in. Segezha Nørresundbys external environment consist of four main elements, which are customers, competition, suppliers and investors. These four eleFigure 2.4. Environment ments are outside the organisation and each element are affecting Segezha Nørresundby [Figure 2.4]. Inside the organisation, is the internal environment, which in Segezha Nørresundby’s case consists of the different departments in the firm. The relations between the elements will not be analysed in this section, because some of the internal factors are describe in the section 2.2 and an analysis of the internal factors, which are related to this project will be made in chapter 3.

Figure 2.5. Environment complexity[Daft, 2007,Fig:4.2]

The environment of an organisation can be described as simple or complex and stable or unstable dimensions. The simple-complex dimension refers to the various numbers of external factors which affects the organisation. The fewer factors in the external environment, the more simple the dimension is [Daft, 2007,p.143]. The stable-unstable dimension refers to the dynamic of the external factors. External factors which are not changing over a period of time are less dynamic and therefore more stable[Daft, 2007,p.144]. A simple and stable environment will indicate a low uncertainty for the organisation, which means that the company has to focus on being flexible in order to adapt to change. The 7

Group B2-14

2. Empirical Background

higher complexity or unsteadily the environment is, the more flexible the organisation has to be in order to adapt to change. Each combination of the two dimensions has some characteristic, see figure 2.5, which can be used to determine the type of environment. The complexity is determined by this analysis, and is a combination between the two dimensions.

2.3.1

Customers

Segezha Nørresundby operates with two kinds of customers. One type of customer places orders at a weekly or monthly basis. These are often distributors and incineration plants. These customers are requesting a short delivery time. Segezha Nørresundby is in general operating with a delivery time on a month, but some of these customers are requesting a delivery within a week. This means that Segezha Nørresundby has to produce some standard products to stock in order to meet the demand for a short delivery time. It is not cost-efficient for Segezha Nørresundby to produce different kinds of customised products to stock. The other type of customers, which often are local authorities, are ordering one year in advance and on a monthly basis, requesting a delivery of products adjusted to their current stock. This means that Segezha Nørresundby has to handle the orders differently according to the type of customer. The order received at a weekly/monthly basis has to be processed fast in order to minimise the delivery time and maintain a good service for the customers. The orders, which are planned a year ahead does not have the same priority, and these orders are produced according to the free capacity of the machines. Segezha Nørresundby is supplying customers all over Europe. Two third of the products is, according to the Process manager [2014], sold within Denmark. The other third of the products is exported, mainly to France and Germany. The firm is providing the customers with custom made packaging solutions in different; sizes, colours, shapes, prints, types of paper, types of gluing etc. Segezha Nørresundby is focusing on providing above average service for the customers. For instance, this is done by providing the customers with technicians, who are specialised in the machines which can be modified for Segezha Nørresundby’s products. The technicians can help the customers adjust their machines to fit Segezha Nørresundby’s products and hereby the customers avoid continuous problems with the products, which could result in the customers not buying Segezha Nørresundby’s products. Segezha Nørresundby provides Aalborg Portland, the local cement manufacture, with a remote storage, which is a special service. The remote storage is a warehouse located and owned by the cement manufacture. It is a warehouse stored with paper sacks from Segezha Nørresundby. Aalborg Portland pay for the paper sacks continually, as the paper sacks are consumed. This benefits both the cement manufacture and Segezha Nørresundby. The cement manufacture does not have to tie up capital in inventory and Segezha Nørresundby maintains an important customer. Since Aalborg Portland has a stable consumption of paper sacks, Segezha Nørresundby can produce the products for Aalborg Portland when idle time appears in the planning process, and hereby Segezha Nørresundby can utilise the production capacity.

8

2.3. Environment

2.3.2

Aalborg University

Competition

An element of the external environment, which influences Segezha Nørresundby, is the competitors on the market, where one of them is the swedish firm Svingo. There are several other competitors in the market, which influence the sale of the Segezha Nørresundby´s products, but Svingo is one of the few competitors that offer the same product in terms of quality and price. At Segezha Nørresundby the order winning factor [Hill, 2000] is quality, since they are not able to compete on prices in some cases. This was apparent, when they lost a deal to supply the local authorities in Aalborg with garbage sacks, since they could not match the prices of their competitor. Segezha tries to differentiate themselves on service and quality, and this is why the customers sometimes are willing to pay extra. Internally the Segezha group has factories positioned around Europe. The factories have focus on different kinds of products in the paper and packaging business. Even though it is not a desirable situation, it does happen that the factories compete with each other. The Segezha group is conducting factory comparisons, by looking both at their balance sheet over a given period, and comparing it with the data from sister factories placed around Europe. Factories, which are not performing at the same level as the others, are the focus of the organisation at the given time. By comparing, you can receive a lot of knowledge about the current situation of the different factories, but it can be argued that it is not a viable way of assessing the individual numbers on the balance sheet. These factories are making different products, with different technology and they are located in different environments. Some firms might need to allocate more expenses in maintenance of their equipment, e.g. Segezha Nørresundby, since some of their machinery is older than the machinery at sister factories. Due to this, Segezha Nørresundby has experienced restrictions to their budget, since the budget was higher than budgets of sister firms placed in Eastern Europe. Comparing the factories overall performance is viable, since this is based on the individual environment of the firms, and comparing individual numbers on their respective balance sheet is not valid.

2.3.3

Suppliers

The Segezha group controls the whole decision making process, since the Segezha group owns the whole chain of supply, from the woods and paper mills in Russia to the factories throughout Europe. As a result of this, the Segezha group has freedom of action, which is advantageous and gives the group more possibilities. The Segezha group does not produce white paper and therefore, it has to be delivered by external suppliers. Segezha Nørresundby is imposed to buy brown paper from the paper mills in the Segezha group, which is another branch of the organisation. On one hand, this secures the supply of brown paper to the production of garbage sacks, but on the other hand, the factory is not allowed to negotiate with other suppliers in order to get cheaper raw materials and hereby increasing their profit by minimising costs. Furthermore the factory has experienced problems with on-time deliveries. The issues with on-time deliveries has occurred due to the different work culture of the Russian workers compared to Danish workers, and the long distance from Russia to Denmark also plays a 9

Group B2-14

2. Empirical Background

role in the delays. Sometimes the trucks, which deliver the brown paper, can be delayed up to two weeks. Due to the unstable deliveries from Russia, Segezha Nørresunby has increased their “buffer” inventory, which has resulted in unnecessary tied up capital. At the time where the Swedish group Korsnäs owned Segezha Nørresunby they never, experienced delays of the same extend. The paper mills in the Segezha group only produce brown paper, so the white paper, which primarily is being used in the industrial production line, is currently being bought and delivered by a Swedish supplier, who is able deliver within days. This is a big advantage, since the products, which are being produced with white paper can be very customised and because of this, a sudden demand for white paper can suddenly arise. Since they are not imposed to buy white paper within the Segezha group, they have the possibility to search for different suppliers and hereby pressuring their current supplier on price and delivery time.

2.3.4

Investors

In 2006 Segezha Nørresundby was acquired by the Russian investment bank InvestLesProm. Since the firm now is owned by an investment bank, it implies that the Segezha group is managed by investment controllers. Due to this, the main goal for the firm is to make instantaneous profit. This means that there is a certain risk that investments, which would make the firm more profitable in the long-term, will not be exploited because it does not show a short-term profit. As previously mentioned, Segezha Nørresundby already started implementing LEAN in the firm but since this, is an ongoing process, which requires some form of financial investment it does not fit with the strategy of investment controllers. Another situation, which has occurred since the firm takeover, is the comparative analysis within the Segezha group, which compares the budget for each factory, and then compares it to the other factories in the group. This means if Segezha Nørresundby chooses to invest more in one department and then gain this investment back somewhere else in the factory, this will be a problem due to that InvetLesProm will through the analysis compare each departments budget separately and not look at the overall profit of all the departments joined together.

2.3.5

Environmental Conclusion

Based on the description of the different environmental elements mention in figure 2.5 on page 7, it is possible to determine which environmental characteristics that fit Segezha Nørresundby in proportion to the stable-unstable and simple-complex dimensions. Segezha Nørresundby is first of all evaluated to be a simple environment. The organisation is only influenced by competitors and suppliers, and can therefore be categorized as a few external elements. Furthermore, Segezha Nørresundby is considered to be a stable environment, since the frequency of the change in the elements is low. The complexity in the environment can therefore be concluded to be low. To conclude this section, these characteristics results in low uncertainty. Low uncertainty means that there are few departments. Segezha Nørresundby is not required to be fast in responding to changes in the market and it seems to be the best fit for Segezha Nørresundby.

10

Pre-analysis

3

The data and facts presented in the following chapter have been obtained after an interview with the Account coordinator [2014], unless otherwise is stated. This section will analyse the interaction between the planning process and the changeovers in the production lines within the SANA department, to identify elements which can be affected to improve the production. First, by establishing that there is a possibility of reducing the changeovers in the SANA department, which will be done by analysing the distribution of total time used in the SANA manufacturing department. Second, by analysing the order flow to determine which process has a potential to be adjusted in order to improve the production. Third, by calculating the distribution of the total time in each of the five production lines in order to determine the amount of changeovers in each production line.

3.1

Distribution of time

The SANA manufacturing department has five production lines, which are called 7151, Total time Stop Changeover time 7155, 7156, 7157 and 7158. Each of the pro8% 15% duction lines are configured to produce different products, see appendix E. By finding the distribution of time used in the five producRuntime tion lines, the extent of the changeover time, 77% runtime and stop time can be established, in order to determine which element or elements have a possibility of being improved. The percentage distribution of the total runtime, Figure 3.1. The percentage distribution of changeover time, and stop time of the 5 prothe total time use in the five duction lines is illustrated in figure 3.1. As production lines [Appendix C] the figure indicates, the runtime amounts to 77 percent, the changeover time to 8 percent, whereas the stop time amounts to 15 percent. Figure 3.1 does not include the scheduled maintenance in the calculations. Run time is the actual time the five production lines are producing paper sacks. The runtime on the different production lines have continuously been improved, by implementing new technologies such as electronic equipment, which have improved the capacity. An overall runtime on 77 percent indicates possibilities for improvement, which could be done by reducing the amount of stop time or and changeovers. The stop time is unforeseen breakdowns in the machines. It can be everything from paper 11

Group B2-14

3. Pre-analysis

stuck inside the machines to actual breakdowns of the machinery. The stop time amounts to 15 percent of the total time. As mentioned in section 2.2, the machines in the SANA department are from the 1960´s and gradually getting older and more worn. The stop time could be reduced by improving the maintenance or upgrading the machines. According to the Process manager [2014] it is not possible to improve the maintenance, or upgrading the machines due to the budget. The changeover time is the time it takes to set up the machines, from producing one product to another. The changeover time amounts to 8 percent of the total time. Since the stop time is assessed to have minimal options for improvements, it can be established that the changeover time in the SANA department can be an element for improving the runtime, and hereby improving the production.

3.2

Order flow

The order flow illustrates the processes, from the customers are requesting products, to the process where the products are delivered. By analysing the order flow it should be possible to determined, which process or processes have a potential to be adjusted in order to reduce the changeovers and thereby improving the production.

Sales department

Industrial planning department

Industrial manufacturing

Warehouse

Customer

SANA planning department

Shipping

Customer

SANA manufacturing

Figure 3.2. Order flow in Segezha Nørresundby

The order starts at the customers. The order is delivered to the sales department, where the price of the order will be negotiated. If it is a returning customer and there is no change to the price, the order will be sent directly from the customer to the planning department. The planning department is divided in two. The industrial planning department and the SANA planning department, which handles different products as mentioned in chapter 1. The planning departments configure a production plan, which contains the sequence which the product are scheduled to be manufactured in. The production plan is sent to the production line connected to the planning department. Hereafter the order is produced and packed on pallets and stored in the warehouse, waiting to be shipped by lorries to the customers. The changeovers are occurring in the manufacturing department and the changeovers are related to the sequence the products are being produced in. If two orders containing identical products are being produced in a row, there will be almost no time used by the manufacturing department to change from producing the first order to the second. The new 12

3.2. Order flow

Aalborg University

order has to be scanned into the computer at the production line, before the production line can start producing the products. If the order contains a different product, than the currently produced product, the machines in the production line have to be adjusted. The time needed to perform the changeover depends on the amount of adjustments which are required. For instance an adjustment in the length of the sacks from 1000 mm to 1050 mm does not need the same amount of time as an adjustment from 1000 mm to 300 mm. The more similar the products of two orders are, the less time the changeover will take. It is the planning department which is determining the amount and time of the changeovers by deciding the sequence of the production orders.

3.2.1

SANA planning department

In the following section an analysis of the order flow within the SANA planning department will be conducted in order to understand where in the planning process the sequencing is being conducted. The order from the customer is defined as a sales order. When the order is registered by the planning department it becomes a plan order. It becomes a production order, when the order is scheduled to production [Figure 3.3].

Sales order

Plan order

Production order

Figure 3.3. Sales, plan and production order

The detailed order flow in the SANA planning department starts with a customer request, received from the sales department or the customer. The customer request becomes a sales order, which is sent to the warehouse, or configured into a plan order that is sent to the manufacturing department [Figure 3.4]. Throughout the process the computer program SAP is used to send and receive information. SAP(Systems Applications and Products) is a software program used for administrative purposes. Many different companies and organisations, from the Danish Army to big manufacturing companies such as Grundfos, use SAP. SAP includes many different program solutions, within planning, financials, human resources, procurement, inventory, manufacturing, logistics, product development and many more [SAP]. Segezha Nørresundby has implemented SAP and uses it to conduct most of the administrative work. When the planning department receives an order, from the sales department or the customers, it is typed into SAP. From here on, the order flow continues in SAP. The order arrivals varies from daily to annual contracts depending

NO

Yes

No NO Mail Telephone yes

Yes

Letter Fax

Figure 3.4. Detailed order flow of the SANA planning department

13

Group B2-14

3. Pre-analysis

on the customers as mention in section 2.3.1. The first, which is done by the planning department, when receiving sales orders, is estimating the priority of the order based on the delivery date and customer. As described in section 2.3.1 some of the customers are requesting a short delivery time, which makes it rush orders, which are forcing Segezha Nørresundby to produce products to stock. Therefore, the planning department checks the availability files in order to know, if supply from stock is durable, this is done by using SAP. If the products are not on stock, the planning department will check the availability file for raw material also by using SAP. If there are no available materials on stock, the planning department will contact the purchasing department, in order to purchase the required material. The obtained information contains details such as sack width, length, thickness, number of paper layers, palletising, type of print, date of delivery, place of delivery and quantity. If the product has been produced before and the data of the sack is already in the system, the planning department only needs to enter the number for the exact sack. The order remains a plan order until it is scheduled. As long as it is a plan order, it can be changed without creating problems in the production. The next step done by the planning department, is to choose which production line in the SANA manufacturing department to produce the product. There are different machines in the department, which are producing different products, see appendix E. Then SAP is used to calculate the time needed to produce the order. The calculations are based on the specifications from the manufacturer of the machine, which have been adjusted through observations and testing by Segezha Nørresundby. The software also calculates a deadline for when the order has to be produced based on the delivery time and the order production time. When the calculation is made, the purchasing department can access the information and acquire the needed material. In this process, the customer will also receive a confirmation by mail. The next step done by the planning department, is to schedule the production, which is an ongoing process done three to four weeks ahead, depending on the order quantities. Changes to the schedule can occur if deliveries of materials are delayed or if an order, which needs prioritising arrives. The first week is more or less closed for changes, due to the production needs to plan the allocation of the workforce. The next weeks in the schedule are more open to changes. When an order is scheduled, it becomes a production order, which means it becomes more difficult to change the specifications of the order. The order is sequenced according to the priority of the order, delivery date and size. When creating the schedule there are different elements to consider such as time for maintenance and holidays, which are coordinated with the production manager. If there are not enough orders in the system, a production line can be shut down in a period and the orders will be allocated to another production line in that period. At the beginning of every week, the planning department delivers a production plan to the manufacturing department.

14

3.3. Changeovers

3.3

Aalborg University

Changeovers

In this section an comparison of the changeover time will be made. The comparison will deal with the average amount of changeovers per week, and a determination of, which production line that has the largest potential for improvements. Five circle diagrams have been made in order to illustrate the percentage distribution of the total plan time in each production line [Figure 3.5]. The circle diagrams have been made, based on data collected from SAP in the period from January 2013 to December 2013. A comparison of the data in the figure shows that production line 7151 has a higher percentage of changeover time, than the other production lines.

7151

ChangeStop time over time 11% 14%

Change7155 Stop time over time

7157

5%

16%

Runtime 75%

7156

ChangeStop time over time 16% 5%

Runtime 79%

Stop time 12%

Changeover time 9%

Runtime 79%

7158

ChangeStop time over time 19% 8%

Runtime 79%

Runtime 73%

Figure 3.5. The percentage distribution of the five production lines set-up time, stop time and runtime [Appendix C]

Figure 3.6 has been made in order to illustrate the amount of changeovers in the different production lines. The data in figure 3.6 shows that 27 percent of the changeovers are being made in production line 7151.

Changeover per week

Comparing numbers of changeovers per week 5,00 4,50 4,00 3,50 3,00 2,50 2,00 1,50 1,00 0,50 0,00

Changeover per week Procent of total

7151 4,52

7155 3,04

7156 2,15

7157 3,83

7158 3,37

27%

18%

13%

23%

20%

Figure 3.6. Changeover per week [Appendix C]

15

Group B2-14

3.4

3. Pre-analysis

Sub conclusion

The analysis of the order process flow within the SANA planning department has shown that, the changeover can be affected in the process of creating the production order. This chapter has described the distribution of runtime, stop time and changeovers in the production lines in the SANA department. The order process flow, and the SANA planning department sections respectively describe the orders way from start to end, and how the planning department in SANA plans and schedules the orders based on the details of the order. Based upon section 3.3 it has been determined that production line 7151 has the highest amount of changeovers, than the other production lines. It is assessed that the possibilities to improve the amount of changeovers are most present at the production line with the highest amount of changeovers, Therefore, it can be derived that production line 7151 is the production line with the highest potential for improvements regarding changeovers. As a result of this, production line 7151 will be the fundamental basis of the problem statement and the case analysis.

16

Problem statement

4

As outlined in the problem analysis, a way to improve the current production of paper sacks in the SANA department can be to reduce the amount of changeovers. The project will be delimited to focus on how the planning of orders is being sequenced. With the aim of delimiting the project, the focus will be on how the planning of orders is being sequenced. In order to improve the production of paper sacks by reducing the amount of changeovers, it is a requisite that there are a significant amount of changeovers to be reduced. Therefore, it is assessed that the production line with the highest amount of hours allocated to changeovers, and the highest amount of changeovers per week has the highest potential for improvement. As the calculations and figures in section 3.3 concluded, production line 7151 was the production line with the highest amount of hours allocated to changeovers and the highest amount of changeovers per week. In continuation hereof, it can be determined that production line 7151 has the highest potential for improvements regarding the amount of changeovers. The following problem statement is hereby formulated. Can the amount of changeovers on production line 7151 be reduced, by using sequencing methods in the SANA planning department? The sub questions will be answered in order to give a comprehensive answer to the problem statement. • How is the AS-IS planning being conducted? • Which sequencing methods can be determined to reduce the amount of changeovers the most, by using data from the year 2013? • Which sequencing method can be determined to be the best method by using simulation?

17

Methodology

5

The purpose of this section is to elaborate, which methods will be used to describe and analyse the production system at Segezha Nørresundby. The collected data will be acquired by applying different tools to the project, in order to assure the reliability and validity of the data.

5.1

Data Selection

The problem statement highly influenced the way the data were chosen. To be able to gather data for the problem statement, a set of various data have to be collected. The problem statement indicated that some data surrounding the amount of changeovers have to be acquired, along with information about the priority of the different orders. Data about the actual amount of changeovers has to be acquired in order to be compared to the theoretical results. The data concerning the amount of changeovers have to be acquired through data sheets from SAP which made it measurable. The priority of orders is more difficult since it cannot be achieved by quantitative data, but it will have to be achieved through a verbal interview with the account coordinator from the firm. The actual data concerning the amount of changeovers will be acquired through SAP files.

5.2

Data Collection

Based upon the selected data, three methods for collecting the data will be applied • Qualitative interviews • Observations • Quantitative data

5.2.1

Qualitative interviews

The qualitative interviews, which were performed expanded over three visits at the Segezha plant in Nørresundby. The first visit was with the process manager and the production manager. This interview was mostly based upon information concerning the history of the firm, the organisations structure, the products Segezha Nørresundby produce, and a tour around the factory. The purpose of this meeting was to present ourselves to the firm, and then acquire some general information about Segezha Nørresundby. At the second visit, the group was divided into three subgroups which focused on planning, history in details 19

Group B2-14

5. Methodology

and the production line procedure. The planning group cooperated with the account coordinator, this meeting served the purpose of understanding the whole planning process, figuring out how the firms planning department chooses to prioritise their different orders [Appendix A.1.1]. The second group cooperated with the production manager to get a deeper understanding about the history of both Segezha Nørresundby, and the entire Segezha group. The last group focused on the production in the SANA department together with a machine operator. This interview was performed down in the SANA department, where a tour through every step of the production in production line 7151 was given. The third visit involved the account coordinator and the process manager where four of the group members attended. The purpose of this meeting was to get an understanding of Segezha Nørresunby’s planning department. The meeting was mainly based on a further explanation of the received data. Another interview was performed with Brian Kristensen a previous worker of the firm. Brian Kristensen worked for the firm for 20 years in the industrial department, and he worked closely together with the planning department, when the LEAN concept was implemented in Segezha Nørresundby.

5.2.2

Observations

Observations have been used to gather information of how production line 7151 operates, and thereby being able to locate areas where improvements are possible.

5.2.3

Quantitative data

Quantitative data was received through SAP files which dealt with the amount of changeovers. These files were supplied by the account coordinator. The group has also gathered quantitative data through Segezha Nørresundby and the Segezha group’s webpages.

5.3

Data Analysis

The data collected for this project has been analysed using different methods. In the following section it will be described how the data will be applied to this project, and the reasons regarding the chosen analysis methods will be explained. The method of analysis, which is overall used on this project is the problem based learning method. Through this method a possibly reliable solution is proposed with the help of relevant theories, methods, materials etc.

5.3.1

Theory based analysis

Theoretical knowledge is used for analysing the gathered data. Applying theoretical knowledge to an operation is an approach which helps create the ideal image of the design characteristics an operation must possess to be as effective as possible. Theory based analysis includes simulation methods and mapping theory, both of which will be applied to this project.

20

5.3. Data Analysis

5.3.2

Aalborg University

Simulation based analysis

Simulation methods are different from other methods, since it is only based on quantitative data. Simulation will be able to give an overview of where in a process an improvement can be made, or test if an improvement through other means is possible by simulating given parameters. In this project it will be used as a tool for estimating the potential of a possible solution since simulation makes it possible to simulate an actual process, and it generates information with which it is possible to evaluate a process. In this project Monte Carlo simulation will be applied. Monte Carlo simulation Reality is a massively complex system, which is not always possible to analyse. For simulation, it is possible to test some circumstances, where concise analytical tools are not available. Simulation shows a model of reality, which is governed by rules and delimitations. Monte Carlo simulation is based on a RNG (Random Number Generator) which ranges from 0-1, and a probability distribution. The RNG creates x numbers which is then translated to a value from a probability distribution. The values are used to simulate an event, which can then be used in the case study [Bernard W. Taylor III, 2010]. To explain with an example, a set of 10 random generated numbers will be shown: 0.57, 0.11, 0.76, 0.87, 0.31, 0.09, 0.64, 0.85, 0.97 and 0.15 In this example it is assumed that different orders needs to be created to show which are rush and which are not, and is done so by reading values of a cumulative distribution. The distribution which will be applied in this example has the values, 0-0.90 correspond to the value not rush orders, and the 0.91-1 to rush. By analysing the data, which the RNG has provided, it can be shown that 1/10 of the orders are rush orders and 9/10 are not rush orders. The reason for using a RNG, is to be provided with an unbiased set of data to evaluate the stability of a model. The Monte Carlo system is the foundation of almost all simulation models, and is what will determine the validity of the findings in chapter 8.

Figure 5.1. Monte Carlo example.

The simulation, which was made in the project, has not been established as a viable simulation model by anyone other than the group members. Because of the lack of experience of the group members, regarding simulation, the reliability of the simulation model can be questioned. Since the simulation was done manually, it is also disposed to human errors and thus might have influenced the results of the simulation.

21

Group B2-14

5.3.3

5. Methodology

Mapping based analysis

The intention of applying mapping theory is because it describes a process in a simplistic form, and thereby makes it easier to analyse the whole process. In this project mapping has been used to describe and visualize the order process flow.

5.4

Reliability and validity

This section will elaborate how high the reliability and validity will be for possible solutions in this project. The reliability and validity of the data used in this project is of high importance, since it determines whether or not a possible solution could be implemented in the firm. The reliability and validity will be evaluated through an analysis which focuses on the reliability and validity concerning quantitative data and credibility, transferability, dependability and conformability concerning qualitative data.

5.4.1

Quantitative data

Reliability Reliability focuses on the quality or consistency of a measurement. A measure which is considered reliable is supposed to generate the same result every time [Trochim and Donnelly, 2008]. The reliability of the quantitative data in this case the SAP files is evaluated to be high. The data collected is fully representable of the amount of changeovers in the SANA department. The account coordinator at Segezha Nørresundby has supplied the group with data spanning over a large time period which is a factor that makes the reliability even higher. Even though the data comes from SAP, employees at some point keyed it into SAP. Thus, human errors could have influenced the data. It has not been possible to triangulate the data with observations, due to the limited time frame of the project. The changeovers, which should have been observed, occur very randomly. Moreover, the combination of changeovers that should have been observed would have been too comprehensive. Since the changeovers occur as randomly as they do, a time frame of approximately a year would have been necessary in order to gather the required data. Validity Validity is the extent to which an analysis measures what it claims to measure [Trochim and Donnelly, 2008]. In proportion to the quantitative data the validity is seen as high, since the acquired data is an important factor in the analysis. The quantitative data which in this aspect is the SAP files contains information concerning the amount of changeover. This is information that is needed for being able to generate an analysis which answers the problem statement correctly.

5.4.2

Qualitative data

The reliability and validity of the qualitative data which in this case are the conducted interviews will be evaluated on four elements Credibility, transferability, dependability and conformability. 22

5.4. Reliability and validity

Aalborg University

Credibility The credibility of the qualitative data is based on a subjective evaluation based on the group’s impression of each interview and the participants of the interview. The credibility of the participants in the interviews can very easily be affected by several factors such as the participants can misunderstand the questions. The human factor also plays a role in this matter, the participants subjective feelings, opinions and level of interest can easily affect the answer which is given. To minimise the credibility issues the first meeting, which was held with the firm involved both the production manager and process manager. When there are two participants at the same time, it helps increase the credibility, since the two way communications enables both persons to clarify if a question is unanswered or misunderstood. Another factor that increases the credibility is that several members from the group took notes, which made it possible to compare these notes and the overall understanding of the meeting. Transferability In this case transferability refers to the degree, which the results can be generalised and transferred to other contexts [Trochim and Donnelly, 2008]. The information gathered from the interview concerning planning is likely to be transferable to other packaging companies for instance the other companies owned by InvestLesProm since Segezha Nørresundby has a lot of experience, and have been a firm with a positive profit for several years. The information gathered from the interview concerning the production and history in details has a low transferability, since these two matters are subjective to this specific firm.

Dependability Dependability in this context is concerned with, if the same results would be achieved if the interviews will be performed more than once [Trochim and Donnelly, 2008]. The data gathered from the interview concerning the firms history in details is expected to generate the same result if the interview were to be performed again. Of course there is a risk of a specific answer changing, but the main information from this interview revolved around the history concerning general facts about the firm, and these facts were also received through a written document which the production manager supplied the group with, see appendix A. Concerning the interview with the machine operator the dependability seemed to be high, this is based on the group’s subjective opinion. A way of increasing the dependability would be to conduct a new interview with another machine operator, and thereafter compare the answers. The interview with the account coordinator was mostly concerned with the entire planning process of the firm. The account coordinator controls the firms planning in the SANA department by herself, and therefore she is seen as a dependable source. Also the content of the interview was concerned with her explaining the quantitative data, which increases the dependability.

23

Group B2-14

5. Methodology

Conformability Conformability is referred to as the degree to, which the results can be confirmed by an external source [Trochim and Donnelly, 2008]. Brian Kristensen a previous worker of the firm, has functioned as an external person to help confirm whether or not the results from the interviews were correct, to the extent he could, with the knowledge that he possess. Due to this external source the conformability is seen to be high.

24

Theoretical approach

6.1

6

The theory of waiting line management

To improve the order sequence several options are possible. One of these is the methods which exist in the theory of waiting lines. The theory of waiting lines makes it possible to conclude which type of queue is the best fit for an organisation. Waiting lines exists everywhere, and is encountered all the time. This occurs because some customers (e.g. machines, products or humans are waiting to be processed) demands are satisfied instantly while others have to wait, this is due to arrival rate not matching the rate of demand. Waiting lines or queues are an important aspect to understand in operation management. Waiting lines are the foundation in creating job design, inventory management, and schedules. In this section, the basics of waiting line theories and the tools to solve them will be discussed. A queuing system in stores as well as in factories, consists of three major components, arrival, queuing system and exit. In order to describe a queuing system, it is important to elaborate the three components, which are illustrated in figure 6.1. !

Queuing%system%

!

Waiting%line%%

%Servers%% Server%1%

Order% %arrivals%

Server%2%

Exit%

Server%3%

Server%4%

Server%5%

Figure 6.1. Queuning system

6.1.1

Arrival

The first area of focus is arrival. For an organisation to achieve the capability of sequencing and scheduling the flow of the orders correctly, the organisation need to possess knowledge of how and in which order customers or waiting units are arranged, and at what rate they arrive. The more information an organisation has concerning the arrival, the better opportunity it has to make the most effective scheduling and sequencing. The number of 25

Group B2-14

6. Theoretical approach

arrivals can be limited - few/many, unlimited, infinite, and constant/variable. A constant arrival distribution implies having exactly the same time between arrivals. This makes the arrival rate predictable, and therefore simplifies the planning process. It is however much more common to see a variable (random) arrival distribution. The queuing system has to adapt to the arrivals in the system. If the arrival is unlimited the servers in the queuing system need to be able to process the products/customers faster than they arrive. If the process is equal or slower than the arrival rate, a queue will appear. The servers will not be able to catch up with the production if the servers should break down, but if the process is faster it would be possible for the queuing system to catch up.

6.1.2

Queuing system

The queuing system consists of waiting lines and a variable number of servers. Within this section the issues pertaining the queuing system are discussed. Factors to consider include the line length and the number of lines. Waiting lines consist of customers (humans, machines, products) waiting to be processed. Servers are the facilities that process the customers in a queue. In any queuing system there may be any number of servers configured in different ways. Length can be divided into infinite potential length and limited capacity. The length of an infinite queue depends on the arrival rate compared to the production rate. If the arrival rate is higher than the production rate there will always be an infinite queue. If the arrival rate decreases to the point of which it is under the production rate the queue diminishes. Then it would be complimentary to reduce the number of server to reduce the production rate. The goal of an infinite queue is to have 100 percent utilisation of the servers. Limited capacity depends on the capacity of the organisation. If the capacity is full, the potential customers cannot order any more before the servers are free. Expressed differently, they cannot produce more than their max capacity. So if the organisation cannot deliver to the customers’ needs, they probably lose a customer and thereby revenue.

6.1.3

Sequencing

There are many ways to handle waiting lines, and each method has a different focus. It could be in what order units arrive in, or how they are prioritised by category. Categories like length of service time. Some of the most common methods used are: • LOT (longest operation time) • SOT (shortest operation time) • FIFO (first in first out) • LIFO (last in first out) The choice of method depends on the strategy of the organisation and the circumstances of the arrival, the servers and the product/customer. Further description of the method is elaborated in section 6.2.

26

6.2. Planning and control activities

6.1.4

Aalborg University

Exit

When a unit is completed or a customer is served, two exit fates are possible: First, the customer may return to the source population and immediately become a competing candidate for service again. Second, there may be a probability of re-service. The first case can be illustrated by a machine that has been routinely repaired and returned to duty but may break down again; the second can be illustrated by a machine that has been overhauled or modified and has a probability of re-service over the near future.

6.1.5

Simulation in queuing theory

When a series of services is performed in a sequence where the output rate of one becomes the input rate of another, simple formulas can no longer be used to solve waiting line issues. The best technique to solving this kind of problem is computer simulation. The practical purpose of queuing theory is to provide an analytical tool, in order to help the system become more efficient. Queuing theory and analytical tools can improve the understanding of model building and simulation. Simulation is a method used to test the models and evaluate their efficiency by simulating different elements in a queuing system such as service rate and sequencing. It is a method for testing changes in a production system through mathematical models. The simulation gives a simplified representation of reality with the purpose of giving a clear view of the process [Chase et al., 2004,p. 243-249].

6.2

Planning and control activities

As the exposition of the planning department shows, the planning and scheduling of the orders are being made of primarily one person, which uses few elements from planning theory. It is assessed that, by using theoretical methods to a greater extend, possible improvements in the planning process can be made. In order to improve planning there has to be a good fit between the four activities; sequencing, scheduling, loading and monitoring & control. To achieve a good fit, it is important to understand the relation between these four activities, and the way they interact with each other.

6.2.1

Loading

Loading is the work, which is assigned to a workstation. In theory, a work station in a manufacturing firm is available 168 hours a week, but in reality this simply is not possible. Several factors, such as maintenance, failure, breakdowns and changeovers influence the manufacturing. Therefore, these factors needs to be taken into consideration, when the sequencing and the scheduling of a production plan is being made [Slack et al., 2013,p. 299].

6.2.2

Sequencing

Sequencing is a method to decide how and in what way an order will be handled. A firm can have different operations rules and priorities, which influence the way the sequencing is being conducted and some of them can be rather complex. The most relevant will be summarized below. 27

Group B2-14

6. Theoretical approach

Physical constraints In some cases, the physical nature of the inputs being processed, can determine the priority of the work. An example could be the printing of the sacks. Again, it might be advantageous to begin printing paper sacks with the lightest colours and then darken the colours slightly, since it is not possible to remove darkness from the colour mix [Slack et al., 2013,p. 301]. Customer priority Occasionally the importance of the customer, influence the sequencing of the orders. It is not always the time of the order, which decides when the order is produced. Some customers have a higher priority than others do. For instance, some customers are often prioritised on the behalf of other customers. It is simply the amount of profit that can be made of a customer, which decides the priority. Sometimes urgent needs arise among customers which can influence the priority and the way the sequencing is being made. An example could be a hospital. Even though operations have been sequenced and scheduled, the conditions of some patients can be more urgent than the ones already scheduled and as a result of this they are given a higher priority. In proportion to Segezha Nørresundby concerning customer priority, one of the customers Segezha Nørresundby generates the most profit by is Aalborg Portland, which is why they have a higher priority than others do. Even though priority customer sequencing can result in a better service, it can also influence the overall performance of an operation. If the priority of the customer decides how the sequencing is being conducted, it can slow down the production since the work flows can be interrupted [Slack et al., 2013,p. 301]. Due date Due date prioritising, means to sequence work in proportion to when it is due for delivery. When this method is being used, the size of the orders and the importance of customers are irrelevant. Due date sequencing can improve delivery dependability and speed, but it is not always the most optimal way of sequencing. Last in first out (LIFO) When using the last in first out method, the orders received last will be produced, first. First in first out (FIFO) First in first out is the opposite. When using this method, the orders received first will be produced first. Longest operation time (LOT) When using longest operation time sequencing, the orders with the longest operation time is produced first. An advantage of this method is that work stations will be occupied for a longer period and it can be a way to utilise the capacity. Shortest operation time (SOT) Shortest operation time is a sequencing method, which can be used in order to get payments quicker. Companies who have cash-flow problems often use this method. Furthermore, it 28

6.2. Planning and control activities

Aalborg University

is a way to improve delivery performance. A downside of this method is that it can affect the productivity and result in a worse service to customers who places large orders, which demands long operations time [Slack et al., 2013,p. 304-305].

6.2.3

Scheduling

When it has been determined in what way the sequence of the orders is being produced, some companies, require a scheduling of the work. Often scheduling consists of a detailed timetable, which shows when and what orders to produce, and at what time these orders should be finished. Scheduling is often used in companies where planning is essential in order to make certain that customer demand is met. The two most common ways of scheduling are forward- and backward scheduling. Forward scheduling is to start producing orders as soon as they arrive. Backward scheduling is the opposite. It is about producing as late as possible in order to save resources. There are pros and cons to both methods. When using forward scheduling labour utilisation and flexibility is high since there is available time for unforeseen jobs. However, it can result in unnecessary capital tied up in finished orders stored in warehouses waiting to be delivered to customers. Backwards scheduling tries to minimise the material costs until the last moment. This method is used by most companies who use MRP (Materials Requirements Planning), LEAN, or JIT (Just In Time) planning. It tends to schedule and sequence the orders by customers “due dates” and it is less exposed to customer changes. The primary task of scheduling is to ensure that the capacity of a production meets the level of demand at any time [Slack et al., 2013,p. 307-310].

6.2.4

Monitoring and control

Once the planning of a production line has been made by using loading, sequencing and scheduling the different transformation processes needs to be monitored and controlled, in order to ensure that the planned activities happens according to the plan. When or if deviations from the plans occur, some kind of intervention is necessary, which most likely will result in replanning.

6.2.5

Relationship between the four element

Figure 6.2 has been made, in order to illustrate how each activity relates and interacts with the other activities. The four activities interacts with each other, but the activities does not necessarily influence each other to the same extend. For instance, sequencing influence how the scheduling is being done more than it influences how the loading is being done. The sequencing circle has a larger overlap with the scheduling circle, than it has with the loading circle. Figure 6.2. [Slack et al., 2013,p. 278].

29

Case studies and data presentation

7

In order to understand how the AS-IS planning is being conducted, it is necessary to identify what is affecting the planning. This will be done by applying the theoretical approach described in chapter 6 on production line 7151 in SANA department. The components of a queuing system will be mapped and described in order to gain an understanding of the AS-IS situation at production line 7151 in the SANA department.

7.1

Components of the queuing system

A queuing system, in a manufacturing company, consists of the three components; order arrivals, production system and exit as described in chapter 6. Figure 7.1 illustrates the components of the queuing system at production line 7151 in the SANA department. The three components will be described in the following text.

Queuing system Production system

Order arrivals

Production planning (Waiting line)

Production line (Server)

Production schedule

Production schedule

Production schedule

Production schedule

4. week

3. week

2. week

1. week

7151

Exit

Figure 7.1. Components of the queuing system at production line 7151

7.1.1

Order arrivals

The order arrivals component is the part of a queuing system, which determines the way orders Produce are arriving to the production system. It has Make to to stock order 40% not been possible to separate the data associated 60% to production line 7151, therefore the data and the diagrams illustrated in this section have been made based upon the data from whole SANA department, which is presented in appendix D. Figure 7.2. Distribution of numbers of orders in 2013 In the SANA department 60 percent of the products, 31

Group B2-14

7. Case studies and data presentation

as illustrated in the circle diagram[Figure 7.2], are MTO(Make To Order), which implies that the order has been placed by the customers. 40 percent of the products are PTO(Produced To Stock) [Slack et al., 2013,p.275], in order to respond to rush orders and utilise the capacity of the production lines. 30

Number of arrivals per day

The distribution of numbers of arriving orders from customers in 2013 is illustrated in figure 7.3. The orders are being received randomly, which means it is a variable arrivals distribution.

25 20 15 10

5 In order to determine if the time 0 between arrivals and the number of arrivals per day follows a statistical distribution, an analysis Date has been made and the results are illustrated in figure 7.4 and figure Figure 7.3. Distribution of numbers of arrivals over a year 7.5. The arrival is exponentially distributed, which support that the orders occurs in a random fashion.

The probability of new arrivals occurring within a day is 83.8 percent. There is a probability between 0.3 percent and 1.5 percent that there will be more than one day between arrivals [Figure 7.4]. The probability of zero arrivals per day is 18.1 percent, which is illustrated in figure 7.5. The probability of one or more arrivals per day is 81.9 percent (100 percent - 18.1 percent). This fits with the calculations of days between arrivals, which has a 83.8 percent possibility of zero days between arrivals. The difference between zero arrivals per day on 83.8 and the 81.9 percent probability of one or more arrivals per day, is due to round-off of the calculations. 90,0%

20,0% 18,1% 18,0%

83,8%

80,0%

16,0% 14,0%

60,0%

Probability

Probability

70,0%

50,0% 40,0%

12,0% 10,0% 8,0%

30,0%

6,0%

20,0%

4,0%

9,4%

10,0%

1,5%

1,8%

2,0%

0,8%

0,5%

0,3%

2

3

4

5

6

7

0,0% 0

1

Days between arrivals

Figure 7.4. Distribution of arrivals

7.1.2

2,0%

9,5% 8,8% 8,7% 5,8% 4,4% 4,7%

6,2% 5,8% 1,7% 1,8% 1,9%

2,4%

2,8%

0,0% 0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627 Arrivales per day

Figure 7.5. Frequency of arrivals per day

Production system

In this section the analysis is based on data from production line 7151. The production system, as figure 7.1 on page 31 illustrates, contains the two components; waiting line and server, which will be described in this section.

32

7.1. Components of the queuing system

Aalborg University

Production planning The production planning forms a waiting line of weekly production schedules. Figure 7.6 has been made based on an analysis of the distribution of time from an order is created, to the time it is scheduled to be delivered. The data the analysis is based upon, does not include production to stock and collection from stock, which is when a customer is supplied from the inventory of Segezha Nørresundby.

Probability

The results of the analysis which 18,0% are illustrated in figure 7.6, shows 16,0% 14,0% that the greater part of the wait12,0% ing time is between 5 and 20 10,0% days. The waiting time over 95 8,0% days is most likely to be the 6,0% local authority described in sec4,0% tion 2.3.1. 2,0% 0,0% Each production schedule consist 5 15 25 35 45 55 65 75 85 95 105 115 125 135 of a number of production orders, Days from order created to delivered which have been sequenced in orFigure 7.6. The probability of waiting time der to utilise the capacity of the production line to its fullest potential. The sequencing of the production orders are performed by the planning department, which is using a set of unwritten rules. The sequencing is done manually without use of tools, as the orders enter the production system. The production schedule is being sequenced according to the length, width, colour of the sacks and the number of layers of paper in the sacks, which the red marked columns in figure 7.7 is an example of. If two orders have the same length, width, colour and layers, the order with the lowest quantity will usually be produced first. The reason for this, is that in case of lack of materials, it is more convenient to have an absence of finished goods in the largest order, rather than the smallest order. A lack of finished goods in the smallest order will have a greater impact on the customer. The planning department is thereby using elements of the sequencing method; shortest operation time.

Figure 7.7. Example of sequencing

33

Group B2-14

7. Case studies and data presentation

The planning department is at a weekly basis delivering a production plan to the manufacturing department. The production plan is consisting of the sequence which the orders are planned to produce. An example of a sequenced production plan over a period of four weeks is illustrated in figure 7.8. The first week is frozen, which means that the production plan is difficult to change. This is due to that the manufacturing department has to be able to plan the amount of the labour force, which is depending on the load rate of the production lines. In the example the letters represent different types of orders. For instance "AA, AA" indicates the same product needed to be produced in two different orders. For instance if the last product in the weekly production plan scheduled to be made, is "FG", then the first product in the following week will also be "FG" or a product as similar to this product as possible. Therefore in the example every other week the letters are in reverse order compared to the week before, which will reduce the amount of changeover when starting a new week. Alle materialer har tilknyttet en routing, som angiver hvilken linie, de skal produceres på.

AA, AA, AA, AB, BB, FG, FF, FF, FF, EF, EE, AA, AA, AA, AB, BB, FG, FF, FF, FF, EF, EE, Routingen erDD, oprettet standard Set”DD, (referenceplan), derBB, er oprettet til deCC, BB, BB, BC,”Reference CC, CC, Operation DE, DD, DD, CD, BB, BC, CC, DE, DD, DD, CD, ud fra et respektive Til enkelte erDD, der mere daCC, kapaciteten kan være forskellig deDD, respektive CC, CC, BC,produktionslinier. BB, BB, CD, DD, DD, DE, end én,CC, BC, BB, BB, CD, DD,for DD, DE, materialer. EE, EF, FF, FF, FF, FG BB, AB, AA, AA, AA EE, EF, FF, FF, FF, FG BB, AB, AA, AA, AA

4. week

3. week

2. week

1. week (Frozen)

Figure 7.8. An example of sequenced production plans over a period of 4 weeks EE, EE, EE, EE, EF, EF,

CD, CD, CD, CD, DD, DD, DD, DD, DD, DD, FF, FF, FF, FF, FF, FF, DD, DD, DD, DD, DD, The server, which in Segezha FF, FF, FG, FG, FG, FG DD, DE, Nørresundby’s DE, DE, DE EF, EF, FF, FF, FF, FF, Production line

BB, BB, BB, BB, BB, AA, AA, AA, AA, AA, BB, BB, BB, BC, BC, AA, AA, AA, AA, AA, BC, BC, CC, CC, CC, AA, AA AB, AB, AB, case is the production is BB, generating CC, CC, CC, CC, CC AB,line, BB, BB, BB

the output of the factory. This is done according to a production plan. The planning department use a standard time in order to calculate the production plan. The standard 4. weekaccording to the 3. week 2. week 1. week time variates production line and each production line has its own standard (Frozen) changeover- and production time, which can be read in figure 7.9. Production line 7151 has a standard production time of 1000 sacks in 45 minutes, including breakdowns and a Referenceplanerne angiver: up pr. ordre og TM=Machine time pr. 1.000in stk.the red marked row in standard changeover time TS=Set of 200 minutes, which is illustrated figureDisse 7.9.er grundlag for beregning af samlet konfektioneringstid pr. ordre. 6WOO51IP TM

6WOO55IP TM

6WOO56IP TM

6WOO57IF TM

6WOO57IH TM

6WOO57IP TM

6WOO58IP TM

The på, standard time and production time Til f.eks. linie Figure 7151, som7.9. I har fokus anvendeschangeover 6WOO51IP som skabelon til oprettelse af routing på alle nye materialer på den linie, - på trods af de store udsving i dens formatområde.

34

Her kunne man have oprettet en referenceplan for hvert formatområde – med de respektive produktionstider. I stedet ændres maskintiden i forbindelse med oprettelse af routingen på materialet.

7.1. Components of the queuing system

Aalborg University

Calculation based on data of a year showing the distribution of changeovers and production time in production line 7151 are illustrated in figure 7.10 and figure 7.11. The production time has a 46 percent probability that it will be between 40 and 50 minutes. This probability fits well to the standard production time, which is used by the planning department to calculate the total production time. The probability of the changeover time is almost exponential negative with a few deviations. There is a probability of 39.1 percent that the changeover will take between 0- and 30 minutes, which is the highest probability. There is a probability on 60.4 percent (39.1+14.8+6.5) that the changeovers will take under 90 minutes. The standard time on 200 minutes is placed near the average changeover time, which can be read off figure 7.11. A machine operator has in an interview, stated that the normal changeover time is 90 minutes. Based on the interview and the diagram in figure 7.11 it can be estimated that the standard changeover time on 192 minutes used by the planning department is incorrect and an average changeover time on 90 minutes is more likely to be accurate. 50%

46%

45,0%

45%

Average Max Min

40% 30% 25% 18%

20%

39,1%

Average 192,0237 Min 3 Max 1645

35,0% 30,0% 25,0% 20,0% 14,8% 15,0%

15%

10,0%

4%

0%

2% 2% 2% 2% 2% 1% 1% 1% 1% 2%

0%

6,5% 5,3% 2,4%

5,0%

0,6%1,2%

0,6%

Figure 7.10. Distribution of production time