A METHOD FOR OPTIMISING THE PRODUCT ARCHITECTURE OF EXISTING TRAINS

(1)

A METHOD FOR OPTIMISING THE

PRODUCT ARCHITECTURE OF

EXISTING TRAINS

Merijn van der Roest

Student number: 1288180

University of Groningen

Faculty of Management and Organization

MSc. Technology Management

NedTrain Refurbishment and Overhaul

Haarlem

Supervision Faculty of Management and Organization: Dr. Ir. J.A.W.M. Vos

Dr. Ir. J. Slomp

Supervision NedTrain Refurbishment and Overhaul: Ir. J. van Dongen

(2)

(3)

Preface

The research presented in this thesis is the final assignment of my study Technology Management at the University of Groningen. The research is executed at the Engineering department NedTrain Refurbishment and Overhaul (R&O) which overhauls and refurbishes trains.

My acknowledgements goes out to all the people of NedTrain R&O who contributed to this research. Not only I have learned a lot about overhauling and refurbishing of trains, but also how it is to work in a professional organisation. Special gratitude goes out to Jan van Dongen for his supervision on my research. He gave excellent feedback on my research which helped me for improving this research. Thanks, Jan!

This research mainly took place at the department of Engineering of NedTrain R&O. I am grateful to the department of Engineering to give me the opportunity for this research, especially Matthijs Keur, the manager of the department of Engineering. Thankfully to the good conversations and discussions about product architecture with Jos Dittmar and Mo Ben Salah, a better understanding of the organisation, issues, problems and so on, was achieved. Special thanks goes to the consultants of ADSE especially Niek Boersema and Marcel Dijkman. They gave me excellent feedback on the results of this research.

Special thanks goes to my supervisor of the faculty of Technology Management especially Jeroen Vos. I would like to thank Jeroen Vos for his excellent and useful feedback, for his patience, our critical conversations, and instructive questions. Due to Jeroen I have learned a lot about doing good research. Thanks Jeroen! Next, I would like to thank Jannes Slomp for his feedback in the last phase of my research.

Thanks to my friends, especially Alexander Raven who reviewed my thesis. Next, I would like to thank my family and Marieke for their love and support.

Merijn van der Roest

(4)

4

Management Summary

This research is executed at a division of NedTrain BV, namely NedTrain Refurbishment and Overhaul (R&O). NedTrain provides the maintenance, cleaning, overhauling and

refurbishment of trains. She overhauls components such as bogies and wheel sets, engines or complete trains. NedTrain R&O takes care about overhauling and refurbishing trains.

NedTrain R&O has many competitors in Europe who are willing to take over NedTrain R&O biggest customer, namely the Nederlandse Spoorwegen (NS). In order to stay ahead of the competition, NedTrain R&O has to be able to deliver low costs, high quality with a short delivery time.

The department of Engineering of NedTrain R&O introduced product modularity to the ICM project, an important characteristic of product architecture. Product architecture is about the arrangement of functions, physical components and their interfaces. The

introduction of product modularity seems to be very successful in helping NedTrain R&O being more competitive. For introducing product modularity, the department of Engineering has used a method to define modules for the ICM project.

Because product modularity is an aspect of product architecture, the management of the Engineering department wants to enlarge it’s scope and wonders if there is a better method than they have been used so far. These improved or new method should help for future

projects to be more competitive. This leads to the following research question:

Which method will help to define the product architecture of a train in an optimal way so that the labour costs of the department of Material Overhaul will be reduced and the throughput time will be shortened?

When looking at the situation at NedTrain R&O, the following aspects determines the most suitable method for optimising NedTrain’s product architecture, namely: high influence of physical aspects on the products architecture, existing products with the preliminary decisions of design and desire for a matrix representation, and manipulation method or step-by-step method. The method of Gu and Sosale (1999) meets the selection criteria. Due to the

flexibility of the this method, the method of Gu and Sosale (1999) is preferable than others. The new method exists of three steps, namely: Problem definition, Interaction analysis and Module formation. In step one, the problem for optimising the product architecture is analysed in detail. This leads to the right objective for optimising the product architecture, which are used in step two. In step two the objective for optimising the product architecture is divided in a functional-objective and sub-objective, which are divided further in factors which have influence on the objective for optimising the product architecture. Step two is the most important step, because in this step all information is gathered, quantified and used for formation of modules. The most difficult part of this step is choosing the right factors and quantifying these factors for optimising the product architecture. All information is represented in a matrix which allows easy formation of the modules in step three. An algorithm which could used for this step is the simulated annealing algorithm. This method is able to reduce the labour costs and shorten the throughput time of the

department Material and Overhaul. But to ground this more, a better practical case should be used than used in this research.

(5)

1 Introduction

This chapter is dedicated to the introduction of this research. To gain a better understanding of the objective of this research, the following subjects will be described. Paragraph 1.1 gives a short description of NedTrain and the division NedTrain Refurbishment & Overhaul. The second paragraph describes the research problem. NedTrain R&O is interested in reducing the labour costs of the department Material Overhaul and shortening the throughput time for the overhauling process of a train. Therefore, paragraph 1.3 gives more insight in the labour costs and throughput time of trains during an overhaul process. Product architecture is described shortly in paragraph 1.4. Than, the motive for research is described in paragraph 1.5.

Furthermore, paragraph 1.6 explains the research question, followed by the objective and sub-questions of this research.

1.1 Short introduction to NedTrain

This paragraph will give a short introduction to NedTrain and NedTrain R&O, with the purpose to make the motive of research more comprehensible. Firstly, information is given when NedTrain was established. After that, the main activities of NedTrain are described. Furthermore, the company structure is shown. Because this research takes place at the division Refurbishment and Overhaul, the structure of this division is shown as well.

The Nederlandse Spoorwegen (NS), the Dutch railway company, was established a 150 years ago. Up to 1999, a division of the NS maintained, overhauled and refurbished trains. After that, this division became NedTrain. NedTrain is a private and a fully subsidiary company of the Nederlandse Spoorwegen (NS). Ninety percent of NedTrain’s business demand is fulfilled by the NS.

NedTrain provides the maintenance, cleaning, overhauling and refurbishment of rolling stock. She overhauls components such as bogies and wheel sets, engines or complete trains. NedTrain refurbishes train interiors and installs safety systems to modernize the trains to today’s standards. Examples of such systems are air-conditioning and safety systems. Furthermore, NedTrain repairs damaged trains and system errors, and replaces derailed trains back on track. Furthermore, NedTrain assists her customers in purchasing and operating trains. The objective of NedTrain is to offer her customers the lowest possible life cycle costs, and a desirable level of safety and quality. In short, NedTrain assists her customers in

(8)

8

Figure 1.1 Organization chart of NedTrain and NedTrain R&O.

Figure 1.1 shows the divisions of NedTrain and the structure of the Refurbishment and Overhaul department. One of the NedTrain’s division is Operations. This division is

responsible for refurbishment, overhaul, components, locomotives, maintenance, services and the quality of trains. Refurbishment and Overhaul (NedTrain R&O, Haarlem) provides the refurbishment and overhaul of trains. The motive for this research came from that particular department, and , the research was also conducted on that location.

1.2 Problem description

This paragraph is devoted to a full description of the problem NedTrain has to deal with. Firstly, the origination of the problem is described, secondly, the reason to solve the problem. And finally, the objectives that resulted from the need solve the problem are described.

NedTrain has many competitors in Europe. In the year 1999 the NS became a privately owned company. Before that, the NS had her own maintenance, overhauling and

refurbishment department for their trains. In 1999, this department became NedTrain

company. The same fate applied to many national railway companies in Europe. Resulting in many privately owned companies that provide maintenance, refurbishment and overhaul for trains. Nowadays, NedTrain is the fifth largest company in Europe that provided maintenance for rolling stock. In total there are twenty serious competitors for NedTrain in Europe, but companies like NedTrain in East Europe has 50 percent lower labour costs and is therefore a real threat for NedTrain.

NedTrain

Fleet Management NedTrain Consulting Business & Development

Finance HRM

ICT Communications

Operations

Components & Locomotives Refurbishment & Overhaul

Maintenance Services NedTrain R&O Finance HRM Overhaul Boogies Product group safety

Damage repair Project management

Internal supply

Purchase

Quality

Logistics & Planning

(9)

Since NedTrain is a privately owned company, the NS lets NedTrain perform maintenance, refurbishment and overhaul to the trains of the NS. In order to maintain the relationship between NedTrain and the NS, NedTrain has to be competitive with other companies, which means to be able to deliver low costs, high quality with a short delivery time. Because there are twenty more companies in Europe that provide maintenance, refurbishment and overhaul of trains, the NS has the possibility to compare the prices, quality and delivery time of other companies with the prices, quality and delivery time of NedTrain. The NS is interested in the company that offers the lowest costs, with the right quality and the right throughput time. So, the international competition has made it clear to NedTrain that if it is intent on surviving competition, lowering of costs is required. If NedTrain is not able achieve this, they might end up loosing their biggest customer.

In order to stay ahead of the competition and maintain a healthy relationship with the NS, NedTrain launched three objectives for their next overhauling and refurbishment project for the NS:

1. Reduce material costs with thirty percent.

2. Reduce labour costs of the department of Material Overhauling with ten to twenty percent.

3. Shorten the throughput time of an overhaul of a train from seventy days to fifty-seven days.

The problem description given above, makes it quiet clear that NedTrain must invest in her business in order to survive. To maintain the relationship with the NS, NedTrain has to meet these three objectives.

1.3 Costs and throughput time of ICM trains

This paragraph supplies more details about the costs and throughput time of the ICM project. This will give a better understanding of why this research is performed. Firstly, an

explanation is given of where these costs and time schedules are based on. Secondly, an overview of the costs and throughput time of the ICM project will be given.

In 1998, NedTrain R&O started the ICR project, which covered the overhauling of InterCityRijtuigen trains. During this project different problems arose concerning

overhauling. For example, the product design was not optimal, which led to many

adjustments, which were performed by employees of the department of Material Overhaul (Materieel Revisie, MR). These adjustments took a long time to execute; leading to high costs and a long throughput time. After the ICR project, NedTrain R&O started the overhauling of the InterCity Material trains (ICM), which consists of 87 train sets. The experiences of the ICR project was taken into account in order to be able to reduce labour and material costs, and shorten the throughput time. The pre-try-out phase of the ICM project was initiated during this research, therefore, this research is based on estimated costs and time schedules. Because costs and time schedules are based on many years of experience, these variables are highly valuable.

The total costs of the ICM project is 165 million euro. The main costs pool is for material expenses, namely: 42 percent; the second largest is for the labour costs the department MR, namely: 16 percent. Other costs pools consist of small percentages of the total costs. More details will be given in chapter 3.

(10)

10 body of the train takes 18 percent and painting and brushing takes also 18 percent of the total throughput time. Due to the drying process it is not possible to shorten the painting and brushing process. Therefore, the total throughput time can be reduced by shortening the following two steps: assembling and preparing body of the train. This will result in fewer trains that are out of running.

The main cost pool consists of the material costs. In order to reduce these costs, NedTrain R&O has invested in better purchasing and communication methods with suppliers, also, sought out more reliable suppliers and, furthermore, outsourcing to suppliers. Since labour costs are the second largest cost pool for the ICM project, NedTrain R&O has invested with the intent of reducing the these costs. This is achieved by investing in product

modularity, which is an important aspect of the discipline product architecture. For example, this is done by designing a modular toilet, which could be assembled in ten hours, instead of hundred hours. Before the motive for research is described, product architecture and product modularity will be explained.

To sum up, to meet the required costs objectives, NedTrain R&O has to reduce her largest costs pools, namely: material costs and labour costs of MR. In order to meet the objective of throughput time, NedTrain R&O has to shorten the assembling and construction adjustment steps. The costs of material are reduced by increased investments in suppliers. The costs for labour of the department of MR and the throughput time are reduced by investing in product modularity.

1.4 Product architecture and product modularity

To gain better understanding of the motive for research, product architecture and product modularity are described. Than, the manner in which product modularity helps NedTrain R&O to meet the objectives is described.

Ulrich and Eppinger (2003) define the architecture of a product as the scheme by which the functional elements of a product are arranged in physical chunks and by which the chunks interact. Chunks are the building blocks of a product. In short, product architecture has to deal with:

1 The arrangement of functional elements.

2 The mapping from functional elements to physical components.

3 The specification of the interfaces among interacting physical components. An important characteristic of product architecture is product modularity. For the ICM project, NedTrain R&O invested in product modularity. According to Sanchez and Mahoney, product modularity is a special form of design which intentionally creates a high degree of independence or ‘loose coupling’ between components designs by standardizing component interface specifications (in Hsuan, 1998).” Baldwin and Clark define product modularity as follows: “Product modularity is the building of a complex product process from smaller sub-systems that can be designed independently, yet function together as a whole. It is a strategy for organizing complex products and processes efficiently (in Fredrikson, 2006).”

Sako and Murray define modularity as the ability to pre-combine a large number of

(11)

combined into a module. Because the time to place the components is reduced, labour costs will be reduced also. Secondly, due to the fewer tasks and adjustments that are required, the time needed to place components, combined in a module, is less. This results in a shorter throughput time, the third objective of the ICM project.

Combining the components off-line in a module will costs time, which will reduce the benefits of placing the module faster; which will costs more labour time and results in more labour costs. But, by combining components to a module, functions can be integrated. For example, each individual component placed on the body of the train needs a construction whereon the component can be placed. Instead of placing individual components and making for each component a construction on the train, components can be combined in a module which needs only one construction for holding the module.

To sum up, product architecture is about interfaces, functions and building blocks. An important characteristic is product modularity. Combining components into building blocks, it facilitates NedTrain in meeting her objectives.

1.5 Motive for research

This paragraph is dedicated to the description of the research motive. NedTrain R&O is of the opinion that it will meet the objectives easier by investing more in optimising the product architecture.

During this research, the ICM project is in the pre-try-out-phase. However, the

introduction of product modularity seems to pay off. Estimated labour hours of MR are fewer, which will results in lower costs. The same holds for the estimated throughput time, which will be shorter. Labour costs and the throughput time of the ICM project are compared with an earlier project, namely the before mentioned ICR project. This reduction is mainly the result of product modularity.

The department of Engineering introduced product modularity to the ICM project. The introduction of product modularity seems to be very successful in helping NedTrain R&O meeting her objectives. With regard to future projects, NedTrain R&O expects to also make use of product modularity in order to achieve her objectives. Because product modularity is an aspect of product architecture, the department of engineering is rather curious to discover how product architecture can be optimised more to meet the objectives. So, NedTrain R&O is widening her scope with the intent of optimising their product architecture in manner that will meet her set of objectives.

The department of Engineering has used a method to define modules for the ICM project. A method is defined as a series of steps taken to achieve an objective. The used method consist of the following steps:

1. Calculate cost and throughput time of current situation.

2. Gain conscious by employees for improvement by product architecture. 3. Make an inventory of suggestions for improvement.

4. Investigate suggestions for profitability on labour cost and throughput time. 5. Select profitable suggestion for execution.

6. Formulate specifications.

7. Decide to execute task by NedTrain R&O themselves or by suppliers. 8. Execute suggestions for optimising product architecture.

(12)

12 establish whether they were defining product modularity in the most optimal way. Because product modularity is an aspect of product architecture, the management of the Engineering department wants to enlarge it’s scope and wonders if there is a better method than they have been used so far. These improved or new method should help for future projects.

The new methods, with regard to optimising the product architecture, must optimise in such a way that the labour costs will be reduced and throughput time will be shortened. Because NedTrain R&O has already invested to reduce the costs for material, this research is aimed at reducing the costs of the second largest cost pool. Therefore, the objective of this research is to deliver a method that is able to optimise the product architecture, consequently helping to reduce labour costs of MR and shorten the throughput time of trains. The reduction of overall costs always be taken into account. For example, the labour costs could be reduced by outsourcing tasks of the department of Material Overhaul (MR), but this will increase the material costs. So, the sum of labour costs of the department of MR and the material costs should be reduced.

This research is about how the engineers could design the product in such a way that it will reduce the labour costs of the department MR and shorten the throughput time of

overhauling a train. The throughput time can be shortened by optimising the process architecture, which has a relation with product architecture, but for this research only the product architecture is considered.

To sum up, NedTrain R&O is interested in a method for optimising the product architecture in a way that the labour cost of MR will be reduced and the throughput time will be shortened. The method should be useful for future projects.

1.6 Research question

This paragraph is dedicated to the research question, which is:

Which method will help to define the product architecture of a train in an optimal way so that the labour costs of the department of Material Overhaul will be reduced and the throughput time will be shortened?

The process of answering the research question will lead to improved or new methods, which will help the department of Engineering to optimise the product architecture. Because the ICM project has already begun, the method will help for future projects. By optimising the product architecture in a way that labour costs will be reduced and throughput time will be shortened, the objectives of NedTrain are easier to meet.

1.7 Objective and sub-questions of research

In this paragraph, the objective of this research will be further explained. Furthermore, the sub-questions are described.

The objective of this research is to deliver a method that help define the product architecture in an optimal way, so that the objectives of reducing labour costs and shortening throughput time will be met.

To be able to meet the objective of this research and to answer the research question, the following sub-questions should be answered. These are categorised in ICM project, product architecture and methods.

ICM project

• What were the objectives of the ICM project?

• Which modules are introduced in the ICM project?

(13)

Product architecture

• What are the characteristics of refurbishing and overhauling trains?

• What is product architecture?

Methods

• Which method is used to define product modularity for the ICM project?

• Which method, used by the ICM project for defining product modularity, could be used for future projects?

• Which method is better and could be used for future projects?

(14)

14

2 Methodology of research project

Chapter 1 has given a problem description and explained the research question. Research is needed in order to answer the research questions and reduce the problem of NedTrain R&O. A methodology is needed to do valid research, and this chapter describes the research methodology. Paragraph 2.1 reveals the research model, and following the steps taken in the research model are explained in paragraph 2.2. In paragraph 2.3, the chapter arrangement and their relations are shown. This paragraph explains where the information, gathered by the research steps, is located. And finally, paragraph 2.4 gives a summary.

2.1 Research model

This paragraph is dedicated to describe the research model.

The objective of this research is to recommend a method that can be used for optimising the product architecture for future projects. Figure 2.1 describes the research model. This research model shows how the objective of this research is met, and how the research question is answered. Roughly, this research model shows three areas of

investigation, namely: methods for optimising product architecture, product architecture and the consequences of product architecture on total costs, labour costs of MR and throughput time. The numbers in figure 2.1 represent the steps taken in this research.

-Total costs -Labour costs MR -Throughput time ICM Without PA -Used method Method -Total costs -Labour costs MR -Throughput time ICM With PA -Functions -Interfaces -Components/Chunks Product Architecture -New or improved method Method -Total costs -Labour costs MR -Throughput time Future project -Functions -Interfaces -Components/Chunks Product Architecture 1 6 5 4 3 2 Legend: PA = Product Architecture Step 1: Analyse ICM without PA Step 2: Analyse used methods and tools Step 3: Analyse how used method resulted in PA Step 4: Compare the PA of ICM project with and without PA Step 5: Gather new method and compare used and new method Step 6: Analyse how new or improved method resulted in new PA Step 7: Compare PA ICM project with PA future project

7

Figure 2.1: Research model.

2.2 Steps in research model

This paragraph explains the steps taken in this research model. Each step taken, is represented by a number in Figure 2.1. When step 1 till 5 are taken, the research question has been

(15)

this research should be interpreted. Firstly, the overall research process will be described, followed by the individual steps that are to be taken.

The department of Engineering is interested in a method that can help optimise the product architecture in such a way that the objectives are easier to meet. Therefore, three subjects have to be analysed: costs and throughput time, methods and the product architecture. Before beginning to investigate the methods, the characteristics of a train and of the process of overhauling a train should be known. This will make any consequences of these

characteristics on product architecture more comprehensible. So, these characteristics should be taken into account when searching for new or improved methods. To investigate the consequences of new or improved methods on the objectives, costs and throughput time should be known. With that information in mind, the consequences of these new methods can be analysed. How to analyse the three subjects is described in more detail in the following steps.

1. ICM project without product modularity

In 2004 the ICM project was started. A cost proposal for the project was made, divided in important cost objects like materials and labour costs. In 2004, product modularity was limited introduced. So, on the basis of ICR trains, an estimation was made concerning costs, labour time and throughput time in order to overhaul the ICM trains. Due to the many years of experience, this information is quite accurate and the cost proposals are still available. This makes it possible to evaluate the total costs, the labour costs of the

department MR and the throughput time. This can be achieved by doing desk research and results in quantified variables. By having these variables quantified, it becomes possible to compare them with the ICM project, where they did invest in product architecture.

2. Used method

The department of Engineering has used a method in order to introduce product

modularity in the ICM project. The used method are investigated and explained in step 2. Before meeting the objective of this research, the used method will have to be analysed, making it possible to detect any weak points, and improve or recommend a new method. The method is investigated by doing qualitative research. This will be accomplished via open interviews with employees of the department of Engineering, of the department MR, and project controllers.

3. Product architecture of ICM project

By gaining a more profound insight in the product architecture, the consequences on total costs, labour costs and throughput time can be evaluated. The optimised product

architecture was realised mainly by the use of modularity. For the ICM project, the department of Engineering has invested solely in product modularity, which is the reason that in this particular phase of the research project, only the consequences of product modularity could be evaluated.

The modules that are introduced for the ICM project consist of components. By looking at these components, before introducing product modularity, and looking at the particular modules that are introduced, it is possible to see which components are

(16)

16

4. Consequences of product modularity on ICM project

The modules that are developed for the ICM project have consequences for many

variables, for instance on total costs, labour costs and labour hours of MR and throughput time of a train. The first part of this step investigates these variables.

Because NedTrain R&O is interested in the consequences on total costs, labour costs of MR, labour hours of MR and the throughput time of a train, these variables will be compared with the variables before product modularity was introduced, which is investigated in step 1. This is the second part of this step. The consequences on these variables investigated in step 1 will be quantified through comparison with the variables investigated in step 4. This comparison makes it possible to form a judgement regarding product modularity’s effectiveness on the variables total costs, labour costs, labour time and throughput time.

5. Selecting new or improved method for future projects

Step 5 involves a search for literature aimed at finding methods for optimising product architecture. The appropriate scientific literature is sought out by using the following criteria:

• product architecture,

• product modularity,

• methods, product architecture,

• methods, product modularity and,

• Design For Assembly.

Characteristics of the product and process that influence the product architecture should be taken into account when selecting useful method. Therefore, the information gathered in step 3 is put to use. When investigating a new method, the newly discovered ones can be used to be compared to the old one. By doing so, any favourable aspects of the used methods could be used for future projects. These analyses are based on a qualitative comparison. Step 5 will result in method that will help optimise the product architecture in a way that the objectives of NedTrain R&O will be met.

By taking these steps, the research question will be answered, resulting in recommendations and conclusions for NedTrain R&O.

The following steps are out of scope of this research, but mentioned for further research. These steps are explained in more detail in chapter 7.

6. Product architecture of future projects

A result of using a new method is that the product architecture will change. Step 6 investigates these consequences of using new or improved methods for product architecture. Therefore, a comparison is made between the product architecture of the ICM project and a future project.

7. Consequences of product architecture on future projects

(17)

2.3 Chapter relations and arrangement with regard to research

steps

This paragraph is dedicated, firstly, to the question of how the different chapters are related, and secondly, to the explanation of the respective chapters’ contents. For a better

understanding of the exact content of this research, the steps in the research model are settled differently compared with the chapter arrangement.

Figure 2.2: Chapter relations.

Chapter One: Introduction to NedTrain and this research. Chapter one has given an

introduction to NedTrain, and a description of the research problem, followed by the research question, the research objective, and further related sub-questions. This chapter describes why this research is executed and forms the basis of this research.

Chapter Two: Methodology of research project. Chapter two revealed the research model and

explained the research methodology. This chapter is important for determining how further information is gathered for the following chapters. This is made clear in figure 2.2, by the respective position it holds with relationship to the other chapters in the figure. Chapter two describes how the information of the different steps of the research model is investigated and describes the chapter relation.

Chapter Three: Current situation of process flow, labour costs and throughput time of NedTrain R&O. Chapter three gives more insight into the total costs, labour costs of the MR

department and the throughput time of the ICM trains. Furthermore, train characteristics that will have consequences on the product architecture are given. This information is investigated according step 1 of the research model. Based on this information the consequences of the introduced product modularity in the ICM project is calculated, which is explained in chapter

(18)

18 5. This chapter describes the influences on product architecture and therefore the influences on a method for optimising product architecture. This information will be used in the following chapters.

Chapter Four: literature about product architecture and methods. This chapter is basic

information for further chapters. Firstly, chapter four will explain product architecture from a scientific point of view. Also, product modularity is further explained. This chapter describes motives for optimising the product architecture and which motive is most applicable for NedTrain R&O. This will give the reader of this research more understanding of product architecture and the influences of the characteristics of NedTrain R&O on the method for optimising the product architecture.

Secondly, this chapter gives a definition of a method and a tool. This will avoid misunderstanding of these terms. These definitions are used in chapter five and six.

Finally, this chapter gives an overview of different methods for optimising product architecture. This gives the opportunity to select the most applicable method for NedTrain R&O.

Chapter Five: Used method for optimising product architecture of ICM trains. This chapter

describes the used method by explaining each step of the used method. This is done according step two of the research model.

The last paragraph of chapter five will describe the results of the used method on product architecture of the ICM trains. This will be done by calculating the costs of the department MR and the throughput time of the trains after modules where introduced. This is done according step three and four. For this calculation the information of chapter three will be used.

Chapter Six: New Method for optimising product architecture for future projects. Chapter six

will describe a new method for NedTrain R&O. Firstly, a general description is given, and thereafter an example will illustrate the new method. Useful aspects of the used method are taken in the new method and therefore described. This chapter is investigated according step 5 of the research model. Information of chapter three, four and fife is used.

Chapter Seven: Recommendation. This chapter will give recommendations for using the new

method. Thereafter, more general recommendations will be given. Therefore, chapter three, four, five and six are used.

(19)

3 Current situation of process flow, labour costs

and throughput time of NedTrain Refurbishment &

Overhaul

Chapter 1 described the objective of this research, namely recommendation of a method that can help optimise the product architecture of future projects. This method should help in a way that the labour costs and throughput time of overhauling trains will be reduced. Chapter 2 described the methodology that will be used to meet the objective of this research. In order to be able to do this adequately, the current situation of NedTrain R&O has to be clarified, especially concerning the variables labour costs and throughput time. Also, the influences of the trains, overhaul process and requirements on the product architecture have to be

understood.

This chapter describes the variables of the current situation and their influence on product architecture. The variables and influences are based on the ICM project, because this project can be regarded as a model for future projects, with the variables and influences connected to it. To gain a better understanding of the process flow, labour costs, throughput time and influences on the product architecture, appendix 1 gives an general description of the ICM project, the production flow, the motive for an overhaul and the path of a project.

Paragraph 3.1 describes the influences of the overhaul process, and the product that has to be overhauled on product architecture. Further, paragraph 3.2 describes the throughput time; followed in paragraph 3.3, by a description of the labour costs.

3.1 The overhaul of an ICM train

Since the ICM trains have operated for twenty-four years, they need an overhaul that is up to current standards and long-term maintenance, a task NS has given to NedTrain BV. The aim of overhauling an ICM train is to enlarge and ensure it’s life time with another fifteen years. The ICM project involves the overhauling process of the ICM types 1 and 2, resulting in a total quantity of 87 trains. The ICM-1 and ICM-2 each consist of three coaches, creating a total of 261 coaches for the ICM project.

An overhaul consists of four types of adjustments, and each will be described concisely: 1. Long Term Maintenance – is aimed at ensuring that the train is able to operate for fifteen

more years, which is achieved through the replacement of parts that are worn out. 2. Overhaul of interior and exterior – Interior overhauling of interior is aimed at couches,

tables, dust-bins, lockage racks, lighting, floor and separation walls. Exterior overhauling is aimed at conserving body of train, paintwork and stickers. Every aspect of overhauling is described in concept designs, which are transformed to basic designs, and these are finally transformed to detail designs. Table 3.1 gives the components that are overhauled. 3. Improvement for conductors and engine drivers - is based on the desires of the train

operators, and are focused on the engine drive cabin with the intent of improving it’s comfort. This also applies to the cabin of the conductor.

4. Modifications: The input for modifications is based on the desires of NS and the advise

(20)

20 Table 3.1: overhauled components

Static transformer Air-conditioning Side windows

Interior: couches, tables, dust-bins Information systems

Toilet

Destination indicator Balcony

Conserving of train

Summing up, the ICM project is aimed at the overhauling process of trains that were built in 1982, amounting to a total of 261 coaches. The overhaul process consists of four elements, namely: long term maintenance, overhaul of interior and exterior, improvement for

conductors and engine driver and modifications.

3.2 Influencing characteristics of trains on product architecture

This paragraph describes the consequences of overhauling on the nature of the product

architecture of a train and processes of NedTrain R&O. For example, the state in which a train enters the overhaul process, determines how the product architecture will be. The possible consequences are derived from evaluation of the production flow and by conducting open interviews with the employees of the department of Engineering and Material Overhaul. These consequences on product architecture also have a profound effect on the selection of a method for optimising the product architecture, meaning that they should be taken into account when selecting a method. Firstly, the consequences on the product architecture when overhauling existing trains are explained, and secondly, the requirements that have

consequences on the product architecture.

(21)

Figure 3.1: Influences on product architecture of train.

Influences of existing train on product architecture

Decisions for preliminary design. It should be taken into account that, when overhauling a

train, its material has been designed many years earlier. At the actual time of design, many important decisions were made, about for instance: allocations, composition and interfaces of components. Because components that have been used, are from the past, it is rather hard to make any changes concerning the location of that component after the overhaul.

Construction changes. During its running-time, a train has many different construction

changes, which is done of course is to improve the train. For example, these changes are done to a component that has to be replaced too frequently. So, this component will be improved or replaced, to reduce the frequency with which replacement is required.

Deviations, unknown adjustments and reparations. When dealing with an overhaul process,

one should take into account that trains are used for many years. There is always a chance of involvement in collisions, resulting in changes in the dimensions of the body, adjustments, and deviations. After a collision, the train will be repaired, which is a rather labour intensive activity, and it is rarely recorded for administration. The variations of these reparations have rather great influence on the overhauling process.

Low production volume. The ICM project consists of 261 coaches that are being overhauled

each year. This means there is no mass production for NedTrain R&O resulting in low standardisation between different types of trains.

Long technical life of components. Because trains are subject to intensive use by travellers and

are operating for many hours per day, trains have to consist of robust durable components. For example, like is the case for luggage racks. The obvious benefit is that these components have a long technical life and seldom have to be replaced.

Product Architecture Existing train • Decisions preliminary design • Construction changes • Deviations, unknown adjustments and reparations • Low production volume • Low standardisation

• Long technical life

• No technical innovations • Interfaces Requirements of products • Right functionality • Short throughput time of overhaul • Reliability • Safety legislations

• Low life cycle cost

(22)

22

Little need for technical innovations. The necessity for technical improvement in trains is very

low, which results in little need for adjusting the technology on a regular base.

Interfaces. Interfaces of components or modules in a product are of very great importance to

the product architecture, and also have a great influence on it.

Requirements of products

Right functionality. When changing the architecture of a train, the most important requirement

is that the components, modules, or the product have the right functionality.

Short throughput time of overhaul. For the customers of NedTrain it is important to have short

throughput time. This has influences on the product architecture, for example, fast assembling methods for components.

Reliability. Every train has to be reliable in its use. Because many travellers make use of the

train on a daily basis, it is an important requirement for the product architecture.

Safety legislations. The public transport section is held under the demands of safety

legislation, which have to be taken into account when developing a train or modules of a train.

Low life cycle cost. When overhauling a train, the train should life for fifteen more years.

Therefore, life cycle costs should be considered and have great influence on the product architecture.

Need for maintenance and cleaning. Because trains are used intensively, it is necessary to do

daily maintenance activities, like cleaning.

Comfort. Passengers prefer comfort trains, therefore trains should be comfortable.

Appearance. The appearance of the exterior is also of much importance. The visual

appearance of certain components have to be up to standard for travellers, for example the couches.

The above given consequences should be taken into account when selecting a method for optimising a product architecture that is set to help meet the objective. Moreover, they are the result of overhauling an existing train and the requirements of a train.

3.3 Throughput time of ICM train

This paragraph gives more details about the throughput time, which is important because one of the objectives is to create a method that optimises the product architecture in a way that throughput time is shortened. Paragraph 3.3 will help to understand where the bottlenecks are concerning the throughput time, which should be taken into account when selecting a method for optimising the product architecture. It depends on the particular process phase one wants to improve, which method is selected for future projects.

(23)

Throughput time of a train.

NedTrain R&O has changed their production process for the ICM project from a dock production to a flow production. This change was made to create a shorter throughput time compared with earlier projects, to have more control over quality and progression and use more the capacity of the factory. Figure 3.2 shows the throughput time in days of each process phase. The total throughput time is 59 days.

Figure 3.2: Throughput time of each process phase.

Assembling the train takes the longest time to complete. By reducing the duration of this phase, the throughput time can be significantly reduced. The same holds for the process phase train body. But, because paint has to dry before it can go to the next process phase, it is rather hard to reduce the throughput time for this particular phase. NedTrain R&O is researching if there are perhaps alternatives to painting in order to reduce the throughput time. There is a relation between the labour hours needed to process each phase and the throughput time. Therefore, when reducing the throughput time, the labour hours has to be reduced. Appendix 2 gives more details about the relation between labour hours needed, the throughput time and the amount of trains taken out of operations. For the phases Assembling and Train body, most hours are involved. So, if NedTrain R&O desires to reduce the throughput time of a train, these two phases are excellent candidates for improvement.

Because the space wherein the employees need to do their task for assembling and preparing the train body is restricted, it is no option to use more employees for assembling or preparing the train body. More employees will not result in a shorter throughput time, due to more employees will be in the way as a result of restricted space.

3.4 Labour costs of department material overhaul of ICM project

This paragraph will focus on the costs of the ICM project, and how these costs are

determined. Because NedTrain R&O intents to reduce the MR department’slabour costs, they are explained in more detail. Because this research is aimed at reducing the labour costs of the MR department, it is important to explain how these costs are fullydetermined. Firstly, the

Train Body; 11; 19%

Brush and Paint; 9; 15%

Floor assembling; 5; 8% Assembling; 20; 34%

Out; 6; 10%

In; 2; 3% Pre-dismantling; 1; 2%

Grit and Brush; 1; 2%

Dismantling; 4; 7%

In

(24)

24 overall costs are explained, and, secondly, the reason why NedTrain R&O wants to reduce the MR department’s labour costs.

During this research, the estimated costs of the ICM project were available. The costs information is estimationsbecause the real information was not available yet. However, since the estimatedcosts are based on many years of experience, they form valid data for research. The real information will become available if the actual overhaul of the ICM trains willtake place under normal conditions. As mentioned earlier in this paper, during this research the pre-try-out phase commenced. Figure 3.3 shows an overview of costs for the ICM project.

Figure 3.3: Estimated costs of ICM project.

The costs for the ICM project are based on the information given in appendix 3. The largest cost pool is that ofmaterial costs, and it consists of expenses made out to the suppliers of the materials. The MR department’s labour costs are determined by multiplying its required hours to perform the ICM project with the actual costs of an hour’s work by a single employee. So, there is no real overhead involved in the labour costs of MR. Less than one percent of the labour costs of the MR department is made up by P&O.Costs resulting fromineffective hours, such as illness, are taken into account in the total labour costs of the MR department. Appendix 4, 5 and 6 shows more detailed information about the labour costs. The situation is different for the Internal Supply department; they do have overhead costs. However, since these are not that significant compared to those of Material and Labour costs, they are considered as integral costs. This is also the case for the department of Engineering.

The costs information will be comparable to any future project. In order to hold a beneficial competitive position, it is most interesting to reduce the largest cost pools. To reduce the material costs, NedTrain R&O has done the following:invested in better purchasing methods, better communication with suppliers, more reliable suppliers, and more outsourcing to

suppliers. The same efforts will be taken for any future project. But, in order to reduce the overall costs, other cost pools should be reduced as well. This can be achievedby reducing the second largest cost pool for the ICM project: the labour costs of the MR department. A way to reduce these labour costs is to optimise the product architecture. Chapter 4 willexplain this in more detail.

Supply Department (TLB) 6% Material 41% Onetime outgoings 2% Mark-up 4%

Total labour cost MR 16% Calculated depreciation 2% Calculated interest 3% Technical Support -Accommodation 1% Personnel department 0%

Technical Support - Rest 3%

Overhead 3% Purchase & Logistics – Logistics

4%

Engineering 5%

Material Overhaul - Production Support

1%

Material Overhaul - General/Staff

4% Purchase & Logistics - Material

transport - Shunt 1%

Purchase & Logistics - Material transport - Airbed transport

(25)

4 General explanation of product architecture,

methods and most suitable method for NedTrain

The previous chapters described the objective and the methodology of this research. As described in Chapter 1, the objective is to deliver a method to optimise the product architecture in such a way that labour costs and throughput time will be reduced. Before meeting the objective, a definition and a description of product architecture has to be given, if one is to understand the product architecture of trains and the reasonwhy NedTrain R&O is investing to optimise this particular element. Therefore, the first part of this chapter is devoted to the description of product architecture from a scientific point of view, by means of

scientific literature. This knowledge will applied in the following chapters and the second part of this chapter. This first part will conclude with the drivers for NedTrain to optimise the product architecture.

The second part of this chapter is devoted to describing the different methods for optimising product architecture. Before describing the used method and delivering a new one, the definitions of and the relations between the elements: discipline, method, tool, user, and objective will be described. This will help in the understanding of the elements and the difference between a tool and a method. The second part of this chapter will conclude with a suitable method for optimising the product architecture.

Both parts are structured in such a way, that it starts with a general description and ends with conclusions for NedTrain R&O. Part 1 of chapter four starts with paragraph 4.1 and is dedicated to the definition of product architecture. Paragraph 4.2 explains an important form of product architecture: product modularity. Paragraph 4.3 explains the general drivers for changing the product architecture and the drivers for NedTrain R&O for changing the product architecture. Part two of chapter four starts with paragraph 4.4 and explains definitions of a method and a tool. The final paragraph gives an overview of different methods for optimising product architecture and explains a suitable method for NedTrain R&O.

Part One of chapter four: Product architecture

Part one of chapter four is devoted to the general explanation of product architecture. At the end of part the drivers for NedTrain for optimising the product architecture are explained.

4.1 Definition of product architecture

Product architecture will be defined in this paragraph, and similarly: what product deals with. In the 1960s, the awareness of product architecture as a strategy arose and many optimisation models were introduced to investigate this new form of strategy. (Mikkola, 2006). In essence, these models describe how to design, develop, and produce parts that can be combined in a maximum number of ways in order to deal with consumers’ demand for variety and uniqueness. Since then, the subject literature has highlighted various aspects of product architecture.

Ulrich and Eppinger (2003) define the architecture of a product as: the scheme by which the functional elements of a product are arranged in physical chunks and by which the chunks interact. Chunks are the building blocks of a product. In short, product architecture has to deal with:

1. The arrangement of functional elements.

2. The mapping from functional elements to physical components.

(26)

26 A component is defined as: a physically distinct part of the product that embodies a core design concept and performs a well-defined function (Chen and Liu, 2005). Interfaces create the interacting relationship between components. A functional structure describes the

relationship between the components and the interface, which will enable the product to operate effectively. This definition of product architecture leads us to the conclusion: every single design for a product has an architecture.

According to Ulrich and Eppinger (2003) the most important characteristic of a

product’s architecture is its modularity. The definition of product architecture given by Ulrich and Eppinger, is about: building blocks (referred to as chunks), interfaces, and functions. When interpreting this definition more loosely, product architecture is about the construction of a product consisting of building blocks, while taking into account the interfaces and functions it entails. When looking at these building blocks in a broader scope, they can be referred to as modules. So, the literature about product architecture makes it quite clear that product architecture is mainly about modularity, while taking into account the interfaces and function. In conclusion, product modularity is the most important characteristic of product architecture.

4.2 Product modularity: the most important characteristic of

product architecture

According to Ulrich and Eppinger (2003), the most important characteristic of a product’s architecture is its modularity. Because of product modularity’s importance, this paragraph is solely devoted to this particular subject. Another reason for highlighting product modularityis the important role it plays withinthe ICM project. Firstly, product modularity is described in more detail, followed by the benefits of a modular and integral design. Finally, the levels of modularity will be described.

4.2.1 Product modularity

Ulrich (1995) makes a distinction between modular and integral architectures: “A modular architecture includes a one-to-one mapping from functional elements to physical components of the product, and specifies de-coupled interfaces between components. An integral

architecture includes a complex (non one-to-one) mapping from functional elements to physical components and/or coupled interfaces between components”.

(27)

Table 4.1: Source Jose and Tollenaere, 2005: Trade-offs between modular product designs and integral product design.

Benefits of modular design Benefits of integral designs

Faster assembly and less production time Interactive learning

Fast development of products High levels of performance through special technologies

Parallel manufacture of modules Systematic innovations

Outsourcing Superior access to information

Lower life cycle costs through easy maintenance

Protection of innovation from imitation Module task specialization High entry barriers for component and

module suppliers System reliability due to high production

volume and experience curve

Craftsmanship Costs savings in inventory and logistics

Economics of scale in component commonality

Flexibility in component reuse

Postponement of operations of differentiation for fast reaction of the market

Shorter product life cycle through incremental improvements such as upgrade, add-on and adaptations

Increased number of product variants

The first nine benefits of modular design especially apply to Nedtrain R&O’s specific situation. However, the first four benefits of modular design are sufficient to help achieve Nedtrain R&O’s set of objectives, namely: reducing labour costs from the MR department and shortening the throughput time.

4.2.2 Levels of product modularity

There are many different levels on which product modularity can take place: component level, module level, subsystem level, and system level (Hsuan, 1998). Because these terms are used rather frequently, a good understanding of them is required. This particular sub-paragraph is devoted to the explanation of these various levels of product modularity.

1. Component level: This is considered to be the lowest level of modularisation, represented

by standard, off-the-shelf parts, such as: resistors, capacitors, connectors, epoxies, and so on. The specifications of these parts are generally well defined and are accepted as industry standards.

2. Module level: Modules are created through a combination of different parts of the

component level: electrical, mechanical, or chemical. For example, a windshield wipers controller is produced with a set of electrical components (e.g. capacitors), mechanical components (e.g. housing), and chemical components (e.g. solder). Then, these modules are assembled in order to produce a sub-system level component. Most modules are rarely universal in nature, because they can impossibly satisfy the technical requirements and demands of all systems, even when they serve the same applications.

3. Sub-system level: Components in the sub-system level are often highly customized for a

(28)

28 4. System level: Systems are closed structures, where the system is enclosed through

subsystems with clear boundaries, and where the individual subsystems must be linked together via interfaces and linkage technologies. Modularisation, at component level, or at system level, requires companies to understand products on adeep level and be able to predict howmodules will evolve over time.

4.3 Drivers for NedTrain R&O for optimising product architecture

Changing the product architecture is a powerful design strategy and is used for different functional purposes, such as: modularity in design, production and use (Fredrikson, 2006). This paragraph describes the drivers for implementing this strategy, thus explaining the reason why a company invests in optimising of its product architecture. This of course with the intent of revealing the reason why NedTrain is investing in optimising its product architecture. “Drivers” are the reasons, or motivations, for optimising the product architecture. Firstly, the drivers for changing the product architecture will be described.

A list of driversis given by Gu and Sosale (1999). According to Gu and Sosale (1999), there are eight objectives with regard to developing a modular design. The most important drivers for NedTrain R&O are described. The other drivers are described in appendix 7. The following drivers must be considered when making up-front system architecting decisions for NedTrain R&O.

1. Production and assembly improvement. Modules are essentially independent entities in a

product with defined interfaces and with other modules and components. Modules can be manufactured separately on different locations to facilitate production processes and expertise, and to optimise equipment utilization.

Product architecture affects theassembly efficiency of the product. Different

modularity scenarios may result in different assembly procedures withdifferent assembly times and costs. In comparison with integral architectureof products, especially for large, complex products, modular architecture allows for modules to be assembled separately on the most convenient locations and then to be put together to reduce the total assembly time and costs.

Production and assembly improvement are created by the possibility to manufacture separately on different locations, and to assemble the products at a later point, thus reducing the total assembly time and costs. This is an important advantage for NedTrain R&O.

2. Services. Products usually require both preventive maintenance and recovery repairs.

Different components of a product display a variety of different maintenance frequencies and repair requirements. By grouping components into modules that can be easily

disassembled, fault analysis and maintenance of products are much better facilitated. When a failure occurs, the faulty module can be temporarily replaced, the faulty parts within the module are repaired, and then the module is returned to service. The main considerations required for serviceability are: frequency of failure, service requirements, mean duration of repairs, frequency of preventive services, accessibility of components, cost of replacements, and repair complexity in terms of the special skills or tools

required.

(29)

3. Dividing design task for parallel development. Design of complex products may require

design teams consisting of experts from different disciplines. Sequential development of such products may take a long time. By breaking down the overall design and

development task into simpler sub-tasks and by properly defining the interfaces between the sub-tasks, design teams are able to carry out the sub-tasks in parallel to reducing product design and development time.

Design teams are able to carry out the sub-tasks simultaneously in order to reduce the amount of product design and development time.

4. Upgrading. Every product has a specificlife cycle and will eventually break down. Many reasons contribute to the disposal of a product, such as: customer demand for new models and thewear of the product. Today’s highly competitive market and high consumer expectations demand of manufacturers to introduce new models in a relativelyshort period of time. Also, the rapidly changing technology quickly renders products obsolete, even though they might still be usable (e.g. computers). The time and effort required to introduce a new product model are usually substantial. One way to facilitate a more rapid introduction is through there-use of existing design and production processes of old models, with as little changes as possible so that the time and effort for new model development is reduced.

This driver is relevant for components or modules that allow for the possibility of being upgraded. For instance, the NS has a favourable stance towards the possibility of upgrading the couches in a train, and there is no problem of achieving this. So, the motive to upgrade is applicable to the couches. Another motive is the construction changes of a train.

5. Recycling, reuse and disposal. Different components of a product may have adifferent duration of life. When a product expires, there may be some usable components left in it. Modular design can group these components into easily detachable modules, such that they can be easily re-used. In order to recycle components properly, material

compatibility has to be considered, since different materials may require different methods of recycling or disposal. A modular product can facilitate the recycling process considerably.

Different components of a product may have different duration of life. This is an important motive for NedTrain R&O for optimising its product architecture.

There is a wide variety of drivers for optimising the product architecture. The specificmotive for this research is to investigate methods that can help optimise product architecture in a way that labour costs and throughput time will be reduced. By changing the product architecture, the quality requirements obviously still have to be met. Ideally, product architecture is able to achieve all objectives mentioned above. However, conflicts between the various objectives can arise.

The above-mentioned driversare important aspects thatNedTrain R&O has to deal with. For this particular research, production and assembly improvement is the most significant motive. As shown in figure 3.2 the throughput time of overhauling a train is mainly determined by the process phase: assembly. The throughput time is determined by the labour hours needed to assemble the train. When the product architecture is optimised in such a way that the assembling process can be performed more easily, resulting in faster

A METHOD FOR OPTIMISING THE PRODUCT ARCHITECTURE OF EXISTING TRAINS