IN THE FIELD OF TECHNOLOGY DEGREE PROJECT
MECHANICAL ENGINEERING AND THE MAIN FIELD OF STUDY INDUSTRIAL MANAGEMENT, SECOND CYCLE, 30 CREDITS STOCKHOLM SWEDEN 2018,
Introducing a Modularzation Strategy for High Performance Products
NINO BOROJEVIC JOEL LAVERGREN
KTH ROYAL INSTITUTE OF TECHNOLOGY
TRITA ITM-EX 2018:466
Introducing a Modularization Strategy for High Performance Products
Nino Borojevic Joel Lavergren
Master of Science Thesis TRITA ITM-EX 2018:466 KTH Industrial Engineering and Management
Machine Design SE-100 44 STOCKHOLM
Examensarbete TRITA ITM-EX 2018:466
Introduktion av en modulariseringsstrategi för högpresterande produkter
Nino Borojevic Joel Lavergren
Godkänt
2018-06-07
Examinator
Sofia Ritzén
Handledare
Mats Magnusson
Uppdragsgivare
Anonym
Kontaktperson
Anonym
SAMMANFATTNING
I industrin idag så är många företag tvungna att hantera många olika kundkrav som varierar från dag till dag. Detta ställer krav på flexibiliteten på produktion och produktutveckling. Det här behovet för flexibilitet har gjort att Företag Xs komponentflora och antalet artikelnummer har ökat. Att skapa och lagra ett artikelnummer är dyrt, vilket påvisar betydelsen av att hålla ner komponentfloran, samtidigt som företag kan anpassa sig för de varierande kundbehoven. Detta kan åstadkommas med en modulariseringsstrategi.
Modularisering kan ses som en nedbrytning av en produkt med standardiserade gränssnitt mellan komponenter. Företag X är ett företag som utvecklar och producerar stötdämpare för fordonsindustrin. Det här examensarbetet behandlar en modulariseringsprocess för en utvald stötdämparfamilj i applikationer för motorcyklar.
Modular Function Deployment (MFD) metoden valdes för att utveckla en modulär arkitektur för den utvalda stötdämparfamiljen. MFD-metoden integrerar kundvärden i den resulterande lösningen för att försäkra om att produkten inte kompromissar med kundens behov.
Semistrukturella intervjuer med konsumenter och “workshops” med Företag Xs anställda hölls för att kunna formulera kundbehov samt för att utföra modulariseringsarbetet. En “bottom-up” analys metod användas för att bryta ner produkten i mindre komponenter. Från MFD-processen så kunde 14 moduler skapas från nuvarande arkitektur. Resulterande arkitekturen indikerade att ytterligare utveckling krävs. Det rekommenderas att Företag X utför en “top-down”-analys för att undersöka om arkitekturen kan förändras för att kunna främja en modulariserad struktur. Det noterades att några av anledningarna till att komponentfloran har växt skulle kunna vara på grund av bristen av kommunikation mellan avdelningar där produkterna har utvecklats separat, bristen på verktyg och databaser för att välja rätt komponenter, tillsammans med att Företag X är för inriktade på prestanda i vissa applikationer. Det ska tilläggas att där “tournament goods” teori kan appliceras och prestanda är viktigt, ska inte en modularisering göras, dessa dämpare ska förbli unika.
Master Thesis TRITA ITM-EX 2018:466
Introducing a Modularization Strategy for High Performance Products
Nino Borojevic Joel Lavergren
Approved
2018-06-07
Examiner
Sofia Ritzén
Supervisor
Mats Magnusson
Commissioner
Anonymous
Contact person
Anonymous
ABSTRACT
Companies are today struggling with rapid changes of customer needs. This puts a demand on flexibility in production and product development. This need for flexibility has made Company X’s component variance grow. Creating and storing an article number is expensive, which means that keeping the component variance down while being able to adjust to the rapid changes of the customer is important. One way to accomplish this is with a modularization strategy.
Modularization can be seen as a breakdown of a product with standardized interfaces between components. Company X is a firm that develops and produces dampers for the automotive industry. This thesis covers a modularization process for a selected damper family in motorcycle applications.
The Modular Function Deployment (MFD) method was chosen to develop a modular architecture for the selected damper family. The MFD method incorporates the customer values in the solution, to make sure that the modularization does not compromise their needs. Semi-structured interviews with consumers and workshops with Company X’s employees were conducted to formulate customer needs and execute the modularization work. A bottom-up method was used for a breakdown analysis of the product. From the MFD process, it was found that 14 modules should be generated while using the current product architecture. The resulting product architecture indicated that there is a need for further refinement. For further analysis it is suggested that Company X deploys a top-down method in order to investigate if the architecture could be changed to further support a modularized structure. It was noted that some of the reasons why the component variance has grown might be the lack of information flow between departments where products have been developed separately, the lack of tools and databases for selecting components, together with that Company X might be too focused on performance for some applications.
However, where performance is key and the tournament theory can be applied, it was concluded that modularization should not be adapted; these dampers should be allowed to be unique.
ACKNOWLEDGEMENTS
Acknowledgements are due to everyone that has directly or indirectly supported this work in any way. A special thanks to our supervisor at Company X, Martin, for the opportunity to conduct this study and withouts whose support this would not have been possible.
We would like to direct a very special thanks, and show our deepest gratitude to our supervisor Prof. Mats Magnusson at The Royal Institute of Technology, who has provided invaluable insights and guidance throughout the whole project with his expertise and experience in the field.
Nino Borojevic & Joel Lavergren 28/5-2018 Stockholm
NOMENCLATURE
Table A and Table B present all nomenclatures used in this report. Table A contains a list of the abbreviations used, and Table B explains the used variables.
Table A. Abbreviations.
ABBREVIATION MEANING
A Attachment
A1 Product A
B1 Product B
C Command and control
CI Customer intimacy
CR Customer research
CVR Customer Value Ranking DPM Design Property Matrix
E Evaluation
IM Interface Matrix MDM Module Driver Matrix
MFD Modular Function Deployment
MG Module Generator
MIM Module Indicator Matrix OE Operational Excellence
OEM Original Equipment Eanufacturer PL Product Leadership
QFD Quality Function Deployment R&D Research and Development
S Spatial
T Transfer
Table B. Variables.
VARIABLE MEANING
C Cost of investing
E Utility
P Probability of winning W1 Payoff from winning W2 Payoff from losing
Μ Equipment investment
TABLE OF CONTENT
1 INTRODUCTION ... 1
1.1 Background ...1
1.2 Purpose ...2
1.3 Delimitations ...2
2 PROBLEM ANALYSIS ... 3
2.1 Research Question 1 ...3
2.2 Research Question 2 ...3
2.3 Research Question 3 ...4
2.4 Research Question 4 ...4
3 THEORETICAL FRAMEWORK ... 5
3.1 Modularization ...5
3.2 Modularization as Business Strategy ...7
3.3 Product Platforms ...8
3.4 Customer Value Driven Modularization ...9
4 METHODOLOGY ... 15
4.1 Data Collection ... 15
4.2 Data Analysis ... 17
4.3 Design Study ... 18
4.3.1 Customer Value Mapping ... 18
4.3.2 Modular Function Deployment (MFD) ... 19
4.3.3 Evaluation... 22
4.4 Limitations ... 22
5 RESULTS AND ANALYSIS ... 25
5.1 Results of Customer Value Mapping ... 25
5.1.1 Analysis of Customer Value Mapping ... 27
5.2 Modular Function Deployment ... 27
5.2.1 CVR ... 27
5.2.2 QFD ... 28
5.2.3 DPM ... 28
5.2.4 Module Generator... 28
5.2.5 MDM ... 29
5.2.6 IM ... 29
5.2.7 Modular Function Deployment Analysis ... 29
5.3 Evaluation Results ... 31
5.3.1 Analysis of Evaluation ... 34
5.4 Additional Empirical Observations ... 35
5.4.1 Analysis of Additional Empirical Observations ... 35
6 DISCUSSION ... 37
6.1 Standardization ... 37
6.2 Knowledge Management ... 39
6.3 Variance ... 40
7 REFERENCES ... 43
APPENDIX A – CUSTOMER NEEDS ... i
APPENDIX B – INTERVIEW GUIDES ... i
APPENDIX C – MFD ... i
1 INTRODUCTION
This report is the result of a master thesis conducted during the spring of 2018 at an industrial company which develops and produces dampers. The company mainly produces dampers for motorcycles, automobiles, and mountain bikes, and will further on be denoted as Company X due to confidentiality agreement. The report describes the background, purpose, and the delimitations for the study along with the selected methodology for conducting the work. Later, the results are presented along with an analysis and recommendations discussion regarding the outcomes of the study.
Keywords: Modularization, high performance products, modular function deployment, voice of the customer, customer research.
1.1 Background
In conjunction with the global competition and market complexity in the industry, manufacturers of physical products have been forced to handle a growing product variety and customization of products in order to maintain the firm’s competitive status (Antonio, Yam & Tang, 2007; Ericsson
& Erixon, 1999). As a consequence, high product variety drives the increase of development costs, creating pressure on the competitive capabilities, e.g; product quality, flexibility, delivery, and low price. Furthermore, a high variety leads to greater complexity in operations, both in management and manufacturing (Magnusson & Pasche, 2014). The high variety has grown from the fact that firms are forced to compete in even more specific market segments, requiring shorter development lead times (Ericsson & Erixon, 1999).
To prevent these negative effects of product variety from emerging, companies have started to adopt modularization strategies (Magnusson & Pasche, 2014). Modular components are components with standardized interfaces that vary within a certain range. The notion “module”
stands for standardized unit or an interchangeable part (Erixon, 1998). The modular architecture of a product is flexible as product variations depend on the interchangeability of different modular components. Furthermore, the modular architecture and components allow for development upgrades for individual modules without the need to re-design the whole architecture (Sanchez &
Mahoney, 1996).
The intention when modularizing is not to limit the customers’ choice when selecting products to purchase, but rather to increase the variants available while at the same time minimizing the number of parts (Ericsson & Erixon, 1999). These variants need to meet the customer needs, which mean that the designers are pressured to relate customer needs to one or several modules through the use of tools and processes. However, as product complexity increases, the modularization process becomes more complex as functions get shared by more parts in a product (Persson & Åhlström, 2006). The challenge then is to relate customer values to modules while at the same time retain product performance (Antonio et al, 2007; Persson & Åhlström, 2006).
During Company X’s history, they have been highly successful in meeting tough customer demands and needs, resulting in a high external efficiency, “doing the right things”. However, this
affected the internal efficiency, “doing things right”, requiring longer lead times and increased production-, logistics- and development costs for meeting the customers’ needs. One way for Company X to be even more successful is to become more efficient and effective (Sköld, 2016).
Efficient and effective in terms of meeting their customers’ needs and simultaneously utilize the available resources in the most efficient ways. According to Sköld (2016), in order for companies to excel in efficiency and effectiveness, adapting a modular approach in product development is key. Only utilizing one type of efficiency can lead to difficulties.
Company X has always been working close to the customers, ensuring safety, quality, delivery, and economy for their applications. Due to this, the number of parts produced has increased, leading to high complexity in managing the product assortment. A lot of development time is spent on making sure that the products fit the customers’ needs perfectly. With this complexity, it is assumed that previous development work gets repeated, in other words they “reinvent the wheel” in new projects. Using modularization, the company could decrease the variety in the products by having off-the-shelf solutions, thus decreasing the risk of repetitive development work and shortening lead times.
1.2 Purpose
The purpose of this thesis work is to investigate how modularization can be used to decrease production- and development costs of dampers, without compromising on customer needs and performance. Furthermore, the report will provide Company X with a proposed modular solution for a damper, and a methodology for modularizing further products in the future.
1.3 Delimitations
Only the motorcycle segment was included in this study. This choice was made to reduce the complexity of the project as it would require a more extensive customer research which in turn would have required more time. Furthermore, because of issues with time limitations for Original equipment manufacturers (OEM) firms and a season involving a lot of development work, there was no possibility to include their input in this thesis.
Modularization is associated with standardization (Erixon, 1998). With this in mind, the thesis will exclude products that are used in racing applications where the demands for performance prioritized. Every damper in the racing segment is considered to be unique, in order to give maximum performance for each specific customer. It is assumed that modularization would affect performance in these segments negatively for the customer.
2 PROBLEM ANALYSIS
In this chapter, the practical implications for Company X are presented in order to formulate research questions.
2.1 Research Question 1
Company X is today developing products in close relation with the customer. Their products are often associated with racing and top-model products, which put a high demand on performance.
They are typically involved in the later stages in projects, when the other products and components of the motorcycle already have been integrated in the design, and the dimension specifications are set. These conditions, together with a desire to take on projects, have led to many custom-made products and a large amount of different article numbers. Creating and keeping an article number running creates costs in production by, for example, increasing downtime in production as switching between the assembling of different products takes time. As an effect, costs associated to development increase. With this in mind, a first research question can be formulated.
RQ1: What effects will a modular architecture have on production- and development costs?
2.2 Research Question 2
The customers are keen on that Company X’s products work better than the competitors for their application area. This means that extensive testing is needed for every model that will have a Company X damper, to make sure the damper is best suited for the exact application. During Company X’s history, performance has been the key to success and it should remain so in the future.
The demand for high performance has led to a high price for Company X’s products, and because of this it is important that they are, and are seen as, the best on the market. Since a modularization to some degree always will induce standardization (Ericsson & Erixon, 1999), it is important that the perception of Company X’s brand, from the users view, is taken in consideration as it is one of their greatest strengths. Their distinguishable colour and design today signals quality and high performance in the eyes of all motoring enthusiasts. This leads to the formulation of the second research question.
RQ2: What potential brand and performance issues will modularization induce?
2.3 Research Question 3
Due to the fact that many of Company X’s customers are OEMs, with their own test drivers who often are very keen on the bikes performing the fastest lap times on track, there might be a gap in what they ask for and how the end user actually drives the bike. What is it that the end customer expects from the bike? Is there a gap between the needs from the OEM and the end customer?
The question then arises, can instead the customer values that need to be fulfilled, be identified? A third research question was formulated.
RQ3: How could customer values be linked, in a beneficial way, to modules?
2.4 Research Question 4
To utilize a modular strategy to the fullest, there might be certain changes in processes and ways of working which potentially have to be made at Company X. This leads to the formulation of a fourth research question.
RQ4: What actions could be taken to facilitate a modularization strategy?
3 THEORETICAL FRAMEWORK
In the following chapter, the theoretical framework that builds the foundation for this study is presented. The framework is designed to give an overview of different types of modularization strategies and then funneling down to how modularization could be used as a business strategy to better fit the needs for both the company, and at the same time, the end customer.
3.1 Modularization
As technology is developing, so is the complexity of the products available on the market. Higher complexity and product variety is often correlated to higher operational costs (Magnusson &
Pasche, 2014). A complex system is, according to Simon (1962), a system that consists of a large number of parts that are interacting in a “non-simple way”. He also argues that hierarchic systems will develop far more quickly than non-hierarchic systems, analyzing systems at comparable size.
Here, a hierarchical system, or hierarchy, is described as a system that is divided into small, or moderate, number of sub-systems where the components are interrelated.
The increasing complexity of products is one of the reasons to why many companies have chosen to adopt modularization or product platform development (Magnusson & Pasche, 2014). Sanchez
& Mahoney (1996) describe modularity in forms of loosely- vs. tightly coupled components, where the degree of which components are coupled is decided by what extent to “which a change in the design of one component requires compensating design changes in other components”. The goal of modularization is to create a high degree of interdependence, or “loose couples”, between components by standardizing interfaces in order to be able to develop subsystems as independently as possible (Dasu & Eastman 1994). Ericsson and Erixon (1999) argue that by breaking down complex structures into subsystems, companies can regain control of all the product-related activities, and achieve an improved product architecture. They also argue that the decomposition of the systems is driven by company-specific strategies, which means that if two different companies were to modularize the same product, they could end up with different results.
Modularization is associated with standardization; however, it should not reduce the customer’s choice. It should increase the number of variants and at the same time decrease the number of components in a product (Ericsson & Erixon, 1999). It is important to note that not all components can be viewed in the same way.
Roberson & Ulrich (1998) make a distinction between two characteristic components, namely;
differentiating attributes and chunks. Differentiating attributes are described as characteristics that customers find important in distinguishing between products. Chunks are the major physical elements of a product, a sharing of these parts will result in high levels of commonality. They claim that if the commonality is too high, customers will not be able to distinguish between products, and if the commonality is too low, manufacturing cost will be high due to the extensive amount of components produced. When utilizing customer driven development, companies run the risk of having low levels of commonality.
Magnusson & Pasche (2014) argue that an extensive customer customization to some degree hampers a thorough product modularization. In line with this, Antonio et al. (2007) claim that
modular products gain economies of scale and component standardization, which in itself will lead to less customer intimacy. Hölttä & Salonen (2003) also state that one of the purposes of modularity is to obtain scale advantages, which also points in the direction that modularization may affect customer customization in a negative manner. Despite these arguments, there is a clear lack in literature when it comes to modularity and how it affects companies with a high degree of customer driven development. Antonio et al. (2007) found that product modularity is significantly correlated with delivery, flexibility and customer service, but not with low price and product quality. Product quality is here described as a competitive capability, and refers to a company’s ability to provide products with better quality performance than competitors (Flynn & Flynn, 2004). This suggests that modularity may not be related to the product quality, which would point towards a benefit from modularization, for companies with a high degree of customer driven development.
Foss (2001) claims that, generally, product development can be divided into a set of activities. The sum of the time accumulated for a product to undergo all these steps, from the start of the process to the launch, can be denominated as lead time. Due to the versatility modularization induces, it can be a useful tool when it comes to minimizing lead times (Simpson, Siddique & Jiao, 2007). Due to the loose couples between modularized components, concurrent engineering, or more specifically the “parallelism principle”, can be implemented (Dasu & Eastman, 1994). The parallelism principle exploits the possibility of executing multiple development stages at the same time, thus reducing development lead time (Yassine and Braha. 2003).
Ericsson & Erixon (1999) present a number of evaluation criteria, where lead time is included, that are affected by modularization. Some of the evaluation criteria that are included in this study are presented below in Table 1. These criteria will all be affected, and can be optimized, depending on the type of product architecture.
Table 1. Evaluation criteria (Ericsson & Erixon, 1999).
CRITERIA EFFECT
Lead time in development
Lead time in development can decrease when the possibility to work in parallel exist. In recorded cases for certain part-by-part products, the lead time was reduced by 30-60%.
Development costs
Parts with the possibility to be transferred between product generations without any design changes have a positive influence on development costs.
Development capacity
The decision whether to develop modules in-house or to outsource is a factor in development costs. With modularization there is a possibility to outsource certain modules to keep costs down. A company in the automotive industry was able to reduce expenses with 40% when outsourcing nine modules.
Product costs Direct material and labor are significant in product cost. In general, material costs increase, or decrease between 3% and 10%.
Quality
Modules can allow for separate functional testing in development, therefore ensuring quality on a module level. Case studies have revealed that the amount of rework can be reduced by 37-75%.
Persson and Åhlström (2006) present four different types of product architectures, mapping out how the functional elements are connected to the physical components.
Modular design: The simplest form. For each module, there is a specific function that addresses the customer needs.
Integrated design: The most complex form. Several functions are integrated into several different modules.
Function sharing and function distribution: Between modular design and integrated design, Ulrich (1995) describes a hybrid modular-integrated architecture which combines function sharing and distribution.
The described types of product architectures have different degrees of modularity, these describe the degree of coupling between component in modules, and the degree of coupling between modules themselves. When a modular design is achieved, the product has a 100% modularity degree, and when the product architecture is integrated the modularity degree is 0%. Persson &
Åhlström (2006) presented a case study where the managers found that there were no suitable modularization methods to deal with complex products. The functional interdependencies between modules makes it nearly impossible to remove them, and simplify the product architecture while maintaining performance. As a recommendation for managers, Persson & Åhlström (2006) put emphasis on deciding the degree of modularity which should be sought to achieve. An important step is then to map out all the physical and functional interdependencies that exist to allow for re- designing functional interdependencies which could be eliminated. Important to highlight is that removing functional interdependencies has a negative impact on overall product performance but a positive impact on local performance. Apart from the strictly structural part, the business strategy also plays a big role in the work of modularization.
3.2 Modularization as Business Strategy
The link between the product development and business strategy is not always clear. Ericsson &
Erixon (1999) use a framework called Modular Function Deployment (MFD) in order to illustrate this relationship. MFD creates a link between the product development and the business strategy with the use of module drivers. These module drivers are then incorporated in the product architecture by relating them to technical solutions. “The chart”, as Nilsson & Erixon (1998) name the tool of MFD, creates an unbroken link of information flow, starting from customer values and ending with the proposed modules (Nilsson & Erixon, 1998). This relation tends to be overlooked in literature today due to the market demand for variety often being considered as given. The majority of the modularization work focus on optimizing the cost structure in R&D and manufacturing, and literature taking an engineering focused perspective (Magnusson & Pasche, 2014). This implies that there might be a gap between the customer values and the product development, something that MFD tries to address.
When developing a business strategy, there is a strong relation to tactics. Here, three value disciplines can be identified, which steers the business in different directions, namely; Operational Excellence (OE), Product Leadership (PL), and Customer Intimacy (CI), see Figure 1. These disciplines can be related to the business whether it seeks to increase variety, improve customer satisfaction, shorten lead time, or reduce costs (Lange & Imsdahl 2014). Typically, businesses tend to seek one of these as their main focus, while minimizing the other. Furthermore, Lange &
Imsdahl (2014) describe that modules often are described by referencing at least one of the following three conceptual attributes:
What the module consists of; correlated to “Voice of the Customer” and what needs are spoken.
Physical limitations; responding to a “Voice of the Engineer” and what technical and physical constraints are present, for them to be able to produce modules.
Why there is a model; representing the “Voice of the Business”, tasked with configuring a product variant using the modules.
MFD, however, promotes taking all three voices into account, by defining a module as a functional building block, driven by company-specific reasons (Lange & Imsdahl 2014).
Figure 1. The three value disciplines (Modified from Lange & Imsdahl, 2014).
3.3 Product Platforms
Like modules, product platforms are used to reduce the complexity for a product assortment by decreasing the number of parts, and therefore also decreasing the number of operations and processes. Product platforms make use of sharing components across a platform of products.
Generally, platform products share many development and production assets. A platform can be defined as; “the collection of assets that are shared by a set of products” (Robertson & Ulrich, 1998). The platforms can be used for a range of products, utilizing economies of scale means that components and subsystems can be produced in higher numbers. Furthermore, product platforms can enable development of new products based on old platforms (Magnusson & Pasche, 2014).
One of the effects of this is significantly shorter development lead time, as technology transfer
from old to new products require less engineering hours (Cusumano & Nobeoka, 1998). As development lead time decrease with the use of product platforms, so does development costs.
The use of product platforms is seen as a method of achieving successful mass customization as products can be produced in large volumes. At the same time, the product assortment can meet the needs of many individuals, as highly differentiated products with minimized consumption of resources can be delivered (Robertson & Ulrich, 1998). Mass customization offers satisfaction of diverse expectations and economies of scales concurrently (Simpson, Jiao, Siddique & Hölttä, 2014). Giovani, Denis & Flavio (2001) note that mass customization can be defined as “the ability to provide individually designed products and services to every customer through high process agility, flexibility and integration”. This could be delivered through modularization.
Pasche & Magnusson (2011) argue that if the changes in customer demands are rapid and the demand for a high degree of customization is present, a modular product architecture is the preferred course of action. While if the rate of change is low and customers prefer a cost-efficient functionality, the company should focus on implementing product platforms. If however there is a high demand for customization together with a willingness to pay a premium price for it, both modularization and product platforms may be considered as obstacles to achieve the desired product flexibility. This is why, when looking at a customer driven modularization, product platforms might not be the optimal choice.
3.4 Customer Value Driven Modularization
Lindstedt & Burenius (2003) estimate that the number of inventions that do not succeed is larger than those that do. Any invention is required to offer a certain degree of value in order for an organization to be successful. For a company to innovate, the products and services created need to let the customer perform a job faster, better, more conveniently, and less expensively than it was done before (Ulwick, 2005). Furthermore, for a customer to purchase a product, the value of it must exceed the expenditure. Equation 3.1 illustrates how Lindstedt & Burenius (2003) define customer value as:
𝐶𝑢𝑠𝑡𝑜𝑚𝑒𝑟 𝑉𝑎𝑙𝑢𝑒 = 𝑃𝑒𝑟𝑐𝑖𝑒𝑣𝑒𝑑 𝐵𝑒𝑛𝑒𝑓𝑖𝑡𝑠
𝑇𝑜𝑡𝑎𝑙 𝐶𝑜𝑠𝑡 . (3.1)
It is however to be noted that the customer values, or competitive factors, differ. Not only in their relative importance, but also in their nature (Slack & Lewis, 2002). Basic needs, or order-qualifiers, are needs that are taken for granted by the customer (Lindstedt & Burenius, 2003; Matzler &
Hinterhuber, 1998). Meeting these criterias is not going to give a competitive advantage, but if they are not fulfilled, it will lead to customer dissatisfaction. What is sought to be achieved with these needs is neutrality. Exciting needs, or order-winners, are needs that are not addressed in the existing product but are capable of creating negative or positive comparative benefit. However, there is some criticism against dividing competitive factors into order-winners and order-qualifiers. Slack
& Lewis (2002) point out that the idea is based on how customers behave when considering only a single transaction, disregarding any types of long term relations. They also argue that the concept is based on considering past sales data, missing out on large groups of customers.
With modularization, firms can address a mass market by creating modules tailored to individual market segments (Hölttä & Salonen, 2003). The essence of mass customization, an effect that is achieved through modularization, is to provide each customer with exactly what he or she wants at the right time (Du, Jiao & Tseng, 2003). If a company can create a flexible product, which facilitates adding or removing different functions, the customers can assess the value of different functions and features (Lindstedt & Burenius, 2003). It is therefore beneficial, both for the customers and the company, to involve the customer in the design stages in the product development, since no one else understands the customer needs better than the customer him- or herself (Du et al., 2003). Ulwick (2005) argues that not all companies consider the end user directly, especially not OEMs and firms that sell their products through channels. This is often the case when the product is not the primary purchase made by the end users, i.e. consumers. In those cases, the company may be tempted to only communicate with buyers or purchasers, missing out on important customer needs (Ulwick, 2005). In these situations, manufacturing companies mistakenly think that they can receive the right customer input through their channel partner, as the channel partner has a relationship with the customer. What has been revealed is that channel partners are only qualified to provide their own cost-related outcomes, which does not reveal what the customers value. Only the end user is qualified of providing the necessary inputs for improving existing products or creating new ones (Ulwick, 2005).
The first stage in the modularization process, as Du et al. (2003) describe it, is to conduct the product definition, which is the process of translating the voice of the customer into product specifications. Collected data in form of customer needs from customers is subjective, therefore it is necessary to translate the needs into objective specifications. Griffin and Hauser (1993) argue that in order for a product to be profitable, all parts of an organization, such as engineering, manufacturing, and R&D, need to understand how customer needs are linked to functions and features. Quality Function Deployment (QFD), a step in MFD (Ericsson & Erixon, 1999; Erixon, 1998), originating and developed in Japan at Mitsubishi and used extensively by Toyota, has proven its usability in connecting the voice of the customer, i.e. customer needs, to how they might be achieved (Slack & Lewis, 2002). The customer values are put into relation to how they relate to product properties. Product properties are defined as measurable and controllable properties of the product (Ulrich & Eppinger, 1995). They also specify that these product properties should be dependent, in the sense that they describe the properties of the entire system, and not on a component level. The product properties should also be practical, meaning that the metric should be directly observable or analyzable and easily evaluated. Some needs, however, cannot be directly translated into metrics, as they are subjective. These needs are related to a subjective product property.
A commonly used method for designing product concepts is the use of function-means tree. From a top-down approach, means and functions, e.g. for a mechanic architecture, are arranged in a hierarchical structure (Chakrabarti, 2002; Simpson et al., 2014). Chakrabarti (2002) defines the function-means tree as a hierarchy of effects, i.e. functions, contributing to a realization of the overall purpose for a mechanical architecture, determined by the means to realize the effects. The result is often expressed in a graphic illustration, called function-means map, in the shape of a tree structure, presenting the functions and the means for providing them (O’Sullivan, 2002). Functions should, preferably, be provided simultaneously in one mean. This is known as entity sharing, and
it is the backbone for enabling the designer to remove redundant parts from the architecture (O’Sullivan, 2002). On the other side of the spectrum there is the bottom-up approach, which is a reactive design. It takes an already existing product and consolidates them into a single set of components, and after this assigns functions to the components, or technical solutions (Simpson, Jiao, Siddique & Hölttä, 2014).
After the functional decomposition, a set of functional requirements with technical solutions can be identified. The technical solutions are the embodiment of the product properties, therefore the relations to the product properties need to be set (Lange & Imsdahl, 2014). In Nilsson & Erixon’s (1998) study, a gap between the QFD and the later described module drivers was identified in the MFD process. The process needed a tool for connecting the two halves. Nilsson and Erixon (1998) were the first to present the Design Property Matrix (DPM) which documented this relationship between product properties and technical solutions (Lange & Imsdahl, 2014).
According to Erixon (1998) there are a number of different driving forces for modularization that have been identified in case studies. These are referred to as “module drivers” and are driven from company strategy. The module drivers are used to identify modules by studying the interrelationship between technical solutions and module drivers (Erixon, 1998). Table 2 presents the module drivers.
Table 2. Module drivers (Erixon, 1998).
STRATEGY MODULE DRIVER
DESCRIPTION
Product Leadership
Technology Push
Refers to parts that are likely to undergo changes due to changing customer needs and technology shifts.
Planned Development
Concerns parts that the company intends to develop in change in order to launch new products, better fulfill customer requirement, or decrease costs.
Operational excellence
Carry Over
Concerns parts or subsystems that most likely will not undergo design changes during the life cycle of a product platform. Therefore, the part/subsystem may be transferred across product generations.
Common Unit
Refers to a part that can be used in the entire product assortment or a large part of it. For example, parts carrying a function that is required by all customers are candidates.
Process/
Organization
To make production more efficient, parts requiring the same production process may be clustered to form a module.
Recycling
Growing interest and emphasis on environmental issues and sustainable design has led to attention for choice of materials.
Modules should preferably not contain mixes of different materials.
Separate Testing
To increase quality significantly, components should be clustered so that they could be put to separate testing.
Strategic Supplier
Modules that should be purchased from vendors as standard modules.
The vendor takes all responsibility for manufacturing, development and quality.
Customer Intimacy
Styling Parts that are visible and used to underline product identity.
Technical Specification
Variation should be allocated in as few parts as possible do decrease product complexity.
Upgrading
Modules that enable upgrading in order to offer the customer the possibility of changing the product in the future to add functions or increase performance.
Serviceability Parts exposed to service and maintenance are clustered to create a service module.
As a final evaluation of the module concept, all interfaces should be examined as they affect the final product and its flexibility (Erixon, 1998). The interface evaluation should illustrate if the interfaces are fixed, moving, or transmitting media between the modules. This step is suitable to take from a production perspective as it indicates what type of assembly principle should be applied. Erixon (1998) distinguishes two different principles, hamburger assembly, and base part assembly. These assembly principles are illustrated in Figure 2.
Figure 2. Assembly principles (Modified from Erixon, 1998).
These two assembly principles are ideal from many standpoints as they allow for simultaneous development and simplified process planning (Erixon, 1998). Thus, a well performed modularization should allow for either a base part assembly, or a hamburger assembly.
4 METHODOLOGY
In the following chapter the data collection and analysis methods are presented along with the designed methodology for modularizing a selected product. This thesis work will be a project in modularization in order for Company X to obtain a methodology which could be repeated for other products in the assortment. The customer driven modularization will be conducted on a selected product family, denoted as A1, which is a common damper for several motorcycles and automobiles in different segments.
4.1 Data Collection
The identification process of customer needs is often a qualitative research task (Griffin & Hauser, 1991). Griffin and Hauser (1996) argue that for a qualitative study, ten up to 30 interviews are needed in order to obtain all customer needs, see Figure 3. However, the same study shows that at 20 respondents interviewed there is a significant drop in new discovered customer needs. Griffin and Hauser (1996) estimate that over 90% of all customer needs are discovered after 20 interviews.
In this study, eleven interviews with twelve respondents were conducted with the goal of extracting customer needs.
To better understand the current situation and future possibilities, and to build a better understanding for the background of this thesis, two interviews were conducted with respondents from the production. They were asked to discuss what aspects of the production that was well functioning, and where opportunities for significant improvements exists. The study ended with five short rounds of interviews in order to evaluate the proposed modular architecture, compared to the current situation
.
Figure 3. Percentage of identified customer needs from The Voice of the Customer (Modified from Griffin & Hauser, 1993).
Semi-structured interviews were conducted to allow the respondents to talk freely within certain boundaries (Westlander, 2000). See Table 3 for more information about semi-structured interviews.
Table 3. Semi-structured interviews (Westlander, 2000).
INTERVIEW METHOD INTERVIEWER’S ROLE RESPONDENT’S ROLE
TOOLS
Semi-structured one-on-one interview
Discover the respondents’
opinion on the subject and understand. Interviewer decides themes and sub- themes discussed.
Has the freedom to discuss anything within the specific themes and sub-themes.
Interview guide, notes, and audio recording with consent.
The studied sample for the customer research (CR) consisted in a mix of test drivers, managers and a sales executive from Company X, retailers, and representatives from service and distribution centers. Important to point out was that the respondents were motoring enthusiasts, which was an important factor for this study in order to define customer needs and not technical solutions. For the short round interviews for the evaluation (E), some of the interviewees were asked to represent again. The interviewed respondents are represented in Table 4.
Table 4. Summary of interviewed respondents.
RESPONDENT INTERNAL/
EXTERNAL
POSITION NOTES DATE STAGE
Respondent 1 Internal R&D 6/2-2018 CR
Respondent 2 Internal R&D 7/2-2018 CR
Respondent 3 Internal R&D 8/2-2018 CR
Respondent 4 Internal Sales and market 8/2-2018 CR
Respondent 5 Internal Test driver 19/2-2018 CR
Respondent 6 Internal Test driver 20/2-2018 CR
Respondent 7 External Retailer 26/2-2018 CR
Respondent 8 External Retailer 26/2-2018 CR
Respondents 9 External Retailer/service centre Focus group 28/2-2018 CR
Respondent 10 Internal Test driver 6/3-2018 CR
Respondent 11 External Consultant 8/3-2018 E
Respondent 12 External MC journalist 8/3-2018 CR
Respondent 13 Internal R&D 13/3-2018 CR
Respondent 14 Internal Production 10/4-2018 E
Respondent 15 Internal Production 11/4-2018 E
Respondent 4 Internal Sales and market Short
interview 2/5-2018 E
Respondent 14 Internal Production Short
interview 2/5-2018 E
Respondent 16 Internal R&D Short
interview 2/5-2018 E
Respondent 15 Internal Production Short
interview 2/5-2018 E
Respondent 17 Internal Sales and market Short
interview 2/5-2018 E
Respondent 18 Internal R&D 2/5 -2018 E
4.2 Data Analysis
Data recovered from interviews was transcribed to identify customer needs. The transcripts were analyzed multiple times in order to define customer statements from all respondents. The transcripts were analyzed by two analysts to identify all expressed customer needs in the transcripts.
However, Griffin and Hauser (1993) point out that if the raw data is analyzed by two analysts, only approximately 70% of the expressed needs can be identified. This is illustrated in Figure 4.
Figure 4. Percent needs identified depending on number of analysts (Modified from Griffin & Hauser, 1993).
The statements were translated to customer needs, see Appendix A. Ulrich and Eppinger (1995) present a well-functioning method for analyzing raw data and translating it to customer needs. It translates the statements to needs on the same detailed level while maintaining full traceability.
4.3 Design Study
The study was conducted in three major steps. The first step, Customer Value Mapping, for identifying customer values. Following the Customer Value Mapping, the MFD process was deployed to create the modules. The final step was the evaluation of the proposed modularized concept. The steps are presented in detail in the following chapter.
4.3.1 Customer Value Mapping
To hear the voice of the customer, a systematic and reliable process is needed, which guarantees a credible and comprehensive final result (Lindstedt & Burenius, 2003). Some of the steps for ensuring this are described as follow:
Exploration: Visiting the customer and conducting in depth interviews.
Verification and Validation: Receiving feedback from the finalized concept.
This stage in the process was designed to extract both basic and exciting customer needs. The customer needs were identified by allowing the respondent to speak about how a motorcycle should feel, rather than describing the suspension itself. In total, five interview guides were used, these can be located in Appendix B. Ulrich and Eppinger (1995) propose that the questions focus on how the customer uses the product. An example of the asked questions are presented below:
How do you use the product?
What should it feel like?
What do you like with existing products?
What do you dislike?
The interviews regarding Customer Value Mapping were also used to identify market segments for Company X, which facilitated the module generation later in the study.
4.3.2 Modular Function Deployment (MFD)
The modularization was conducted with the use of the method MFD in order to construct modules from the gathered data. MFD is a strategic method for creating a modular product design while keeping the customers’ needs in consideration. As a mean for conducting the MFD process, a software for facilitating the creation of modules with technical solutions and business strategies called Palma was used.
An interview with a modularization expert was conducted with the aim to consult and validate the method that would be used in this study. The modularization expert suggested that it would be favourable to let Company X participate as much as possible during the whole MFD process. Both in the sense that change is often easier to achieve and more widely accepted if it is driven from the inside, and the technical knowledge possessed in house is far superior than the one which can be obtained during this the period of this thesis. QFD and DPM were the stages which put the highest demand on product expertise, therefore it was deemed necessary to include experienced personnel in arranged workshops. A total of three workshops were conducted to gather input for the MFD process.
The MFD process consists of five steps. These steps are illustrated in Figure 5. The tool in Palma is referred to as The Chart by Nilsson and Erixon (1998), and is used as assistance when creating a modularized breakdown of a product. These matrices and steps are described in detail later.
.
Figure 5. Modular Function Deployment process. The Chart.
Customer Value Ranking (CVR): Here the identified customer values were ranked on a one to five scale, based on the importance for each segment. For each segment, a persona was created, describing a typical end user for that market segment. The basis for the rankings were the conducted interviews and consultation with experts at Company X. The ranking then weighs the importance of the customer value for the segments, based on the score, for the further steps of the MFD process. The decision was made that for all customer values relating to safety would to be give the highest score, as safety is a priority in all applications for Company X.
Quality Function Deployment (QFD): Here the ranked customer values were put in relation to how much they affect the product properties. To do this, the product properties had to be defined.
In order to do this a specification process was used.
Otto and Wood (2001) argue that the specification process describes the measurable criteria which the product should satisfy. They state that this process can be divided into two aspects; engineering requirements, which are on a dimension that support units, and holds a target value which establishes required performance. The second is a translation of qualitative customer needs. When developing the engineering requirements, they also argue that it is important to separate between functional requirements and constraints. Where functional requirements are statements of specific performance of design, or what the device should do, while constraints are external factors that limits the selection of system characteristics. Otto and Wood (2001) claim that, the customer needs should be understood, and also how the current product satisfies these needs.
The relation between the customer values and the product properties is translated to points in the matrix based on the correlation between the customer need and the product property, see Table 5.
The relations were determined during a workshop with two experts at Company X.
Table 5. symbols for linking customer values to product properties.
RELATION SYMBOL WEIGHT DESCRIPTION
Strong relation 9 p Clear and positive effect on the customer value, confirmed by all customers.
Medium relation 3 p Positive effect on the customer value, confirmed by most customers.
Weak relation 1 p Positive effect on the customer value in some cases.
Negative relation - 0 p Negative effect on the customer value.
The QFD relates to the voice of the customer, this is where the customers’ needs are incorporated in the modularization. The accumulated score for each product property is then carried over to the next step in the MFD process, the DPM.
Design Property Matrix (DPM): Product properties were here related to technical specifications.
A similar point based ranking system as in QFD was used for this matrix. The technical specifications can be described as the structure broken down in smaller pieces (Erixon & Ericsson, 1999). Each technical specification should answer to a specific function for the entire system. These technical solutions can be identified by using two types of methods, top-down or bottom-up. Due to the complexity of the products and the limited knowledge about them, a bottom-up approach was used. Design drawings were studied and designers were consulted with in order to first list all components and then assign functions to them. The technical solutions were related to the product properties during workshops with a total of three experts from marketing, R&D support and engineering design, in order to make sure that the interest of all parts of the company are incorporated in the solution. The relation is based on if the technical solution drives variance for the product property. If the technical solution drives one variant for the product property, also called specification point, an empty circle is appointed. If it drives two to three variants a half circle is appointed, and if it drives three or more, a full circle. With this, a score for the complexity of the technical solution is calculated in Palma. Palma also helps with clustering the technical solutions which drives variance to similar product properties, as these often can be put into the same module (Lange & Imsdahl, 2014). With this step the voice of the engineer is taken into account in the modularization of a product.
Module Driver Matrix (MDM): The technical solutions were here put into relation to module drivers. The module drivers were set after a discussion with design engineers with several years of experience and history at Company X. Technical solutions that significantly increase product performance were given product leadership strategies, the more “structural” components were given operational excellence strategy and the technical solutions which satisfy specific customers were given the strategy customer intimacy.
Module Indication Matrix (MIM): Incorporated in the same step as MDM, the MIM relations gives the module, containing the technical solutions, a strategy. The module strategies were set to match the strategies for the components. This is where the MFD process incorporates the voice of the business into the solution.
Interface Matrix (IM): In order to receive an overview of the interface relations, an interface matrix was created, see Figure 6. All the modules were put into relation against each other and given a category for which type of relation they had. The categories were:
Attachment (A) - If there is a physical connection between the modules.
Transfer (T) - If material or energy passes through the interface.
Command and control (C) - If operational signals pass through the interface.
Spatial (S) - If there is a volume or area constraint to the interface.
Figure 6. The concept of IM (Modified from Erixon, 1998).
The marked area in Figure 6 should be avoided as much as possible. Relations there deviate from the assembly principles and thus, increase complexity. If there is a need for unwanted relations, further work is required to minimize them.
4.3.3 Evaluation
The proposed concept was evaluated in order to be compared to a non-modularized solution. In order to do this, representatives from production, development, and sales/market were interviewed and presented with the new concept which they compared to the current. The interviews main focus were three parameters, namely performance, brand and variance.
4.4 Limitations
The selected methodology lacked tools for identifying customer needs which were order-winners and order-qualifiers. Categorizing these needs would not affect the end result of this study significantly, other than giving the customer values in the CVR more, or less, importance. However, the data could have been valuable for Company X to receive as it could serve as data for product development. Therefore, it is proposed that if a customer research were to be conducted again, the data received from the qualitative study is complemented with a quantitative study in form of questionnaires, using scales to decide the importance for each customer value.
Due to the limitation of time and competence regarding the products, a top-down method has not been used. This choice of method would have given greater product knowledge as the functional interdependencies would have been revealed.
One of the limitations to this project is that no OEMs were interviewed. It is however to be noted, that since OEM’s have their own parameters upon which they evaluate the dampers, some information might have been overlooked when they were not included in the sample.
Most of the conducted interviews were internal, with respondents employed at Company X. There is a danger in this, as technically knowledgeable persons are more likely to talk about technical solutions rather than the actual customer values, as they are familiar with the product. In order to avoid this, the interview guide was reformed and separate interview guides were used based on the technical competence of the respondent.
5 RESULTS AND ANALYSIS
The results from the Customer value mapping, MFD process and the Evaluation are presented in this chapter. As an exploratory method was used, additional key input was received during the interviews, these are presented in section 5.4, Additional Empirical Observations. After each result, an analysis of the result is presented. The results and analysis are further discussed in chapter 6.
The interview guides used to gather information and evaluate the projects can be located in Appendix B.
5.1 Results of Customer Value Mapping
In the Customer Value Mapping, the transcripts from the interviews were analyzed and customer values identified. Quotes were put into a matrix and then interpreted into a customer need. The interpreted customer values from the customer research stage are presented in Table 6 along with a short description where needed. The Customer Value Mapping resulted in 38 identified customer values. The origins from the customer values with the statements from the interviews are located in Appendix A.
Table 6. Identified customer values.
CUSTOMER VALUE DESCRIPTION
Jumps with high loads are supported
Stable on track Stiff, good performance on track.
Stable at high velocities
Predictable behavior Predictable, feedback delivered, does not wobble, pot holes and bumps absorbed, vibrations reduced.
Feedback is delivered continuously through the seat.
Does not wobble during full throttle Stable at full throttle.
Feels sporty without removing softness
Rear damper provides support during acceleration
Ride height preserved during cornering Does not scrape against the tarmac during hard cornering.
Retained ride height during loads
Application-unique settings The setting is customized for a specific motorcycle.
Absorbs pot holes and sharp edges
Provides confidence Helps the driver push the limits. Should feel like you are a better driver.
Tires in contact with road During cornering and acceleration.
Better performance than standard dampers
Harsh ride Stiff ride.
Reduces vibrations Small vibrations from surface roughness that exhaust the driver are reduced.
Chassis feels numb
No spikes in damping At sharp edges.
Installation Fits well for every application.
Height adjustable
Pre-tension is automatically controlled So that the ride height and stiffness can be adjusted quickly and easily.
Dynamic functionality Harder and sportier ride.
Not adjustable outside performance interval For safety reasons.
Settings generated through digital service Adjustable characteristics
Actively adapting while driving
Easy adjustability by hand Setting and pre-tension are easy to adjust.
Click changes, visible and noticeable Customer is able to return to previous settings, and help to find new are available.
Preserved handling characteristic During high workloads, such as racing and motocross, the damper should not fade during the later stages of the race.
Instills pride Driver feels proud to own a Company X damper. Racing connection, looks good.
Damper is visible In application.
Good looks
Looks preserved over time Surface coating, withstands water, dirt, etc.
No oil leakage between service intervals
Withstands humidity Does not corrode.
Preserved functionality over time Should not leak, keeps its characteristics.
Long service intervals
