• Author / Uploaded
• Dolana Ruiz Granados

Citation preview

Design for Logistics Ebbe Gubi* Center for Industrial Production, Aalborg University, Denmark

Abstract The nature of a given product has a major influence on the performance of the supply chain associated with it. Conversely, logistics costs are a significant share of total product costs. So it stands to reason that a company may be able to reduce costs, and thereby gain an edge on its competitors, by improving the fit between products and logistics. The research is related to an on-going process to improve the logistic organization’s influence on product creation at Bang & Olufsen, a Danish Manufacturer of audio and video equipment. Based on a literature review, a company-supportive approach to Design For Logistics is proposed, including certain critical elements and guidelines for implementation.

Keywords: DFM, DFL, Logistics, Product creation, Concurrent Engineering

1. Introduction and background This paper presents initial thoughts in a PhD. research project entitled ‘Concurrent product and supply chain creation’. The purpose of the research project is to investigate the potential interactions between products and logistics. The incentive is that the product itself has a major influence on the performance of the supply chain. At the same time logistic costs are often estimated to amount 15-20% of product costs (e.g. Lee, 1992) which makes an improvement of the supply chain performance a potential way to gain competitive advantage. The research assumes that mutual connections exist between a company’s product structure and the associated supply chain the company operates. Though many scholars differentiate between logistics and supply chain management (e.g. Lambert et al, 1998), the supply chain is the system where in logistic activities takes place: ”Logistics is that part of the supply chain process that plans, implements, and controls the efficient, effective forward and reverse flow and storage of goods, services, and related information between the point of origin and the point of

consumption in order to meet customers' requirements” (www.clm1.org). Throughout this paper the terms logistics and supply chain are used interchangeably.

1.1. Research approach To illustrate various approaches when creating products and logistic systems the model in Figure 1 was developed. The aim for the research project is to increase understanding of the different approaches, especially the situation where product and supply chain creation is carried out simultaneously and leveraging each other. Numbering the different approaches in Figure 1 is not an indication of a desirability order; their usefulness will depend on the specific situation (market, product, etc.) a company find itself in. However, the author considers a good ‘fit’ between products and supply chains beneficial to most companies. This particular paper focuses primarily on the approach in quadrant 1, Design For Logistics (DFL), in which the product creation process needs to consider logistics. One example from the literature is Hewlett-Packard’s (re-)design of a printer. Instead of an integral power supply with

* Center for Industrial Production, Aalborg University, Fibigerstræde 16, DK-9220 Aalborg, Denmark, [email protected]

47

E. Gubi: Design for Logistics

e.g. product families. For example, Nokia Networks have different delivery processes for different product types and different customer needs (Hoover et al, 2001; Tissan & Heikkilä, 2001). Finally, the two approaches will be positioned against one another in order to extract elements to the desired approach in quadrant 2. Hence, the research is divided into three steps, with this paper being part of the first.

1.2. Business incentive for the research

Figure 1: Various approaches to designing products and supply chains different fixed power plugs fitting each country standard, the power supply was designed out of the product and could be attached at the latest possible point. This allowed for lower inventory of printers; before, they needed to stock both printers for the US and for Europe with risk of stock-out of one version and excess stock of the other (Lee, 1992). As a next step in the research the approach in quadrant 4, in which the logistic system needs to be flexible in order to match different product, will be considered. A preliminary idea is to tailor focused supply chains to distinct product types,

According to Paashuis & Boer (1997) Concurrent Engineering is not widely implemented in industry, mainly due to a lack of normative methods on how to configure CE to the specific company situation - that is what situational parameters to address. This research focuses distinctively on increasing the understanding of the interrelationships between products and supply chains, through implementing DFL in a company (an inductive approach). The empirical part of the research is based on an in-depth case study in Bang & Olufsen, a Danish manufacturer of audio and video equipment. This remainder of this section gives a short introduction to the case company. Case: Bang & Olufsen Bang & Olufsen is traditionally a very product oriented company, in which the logistic system has previously been considered a ‘necessary

Figure 2. Examples of products

48

E. Gubi: Design for Logistics

Figure 3. The Bang & Olufsen supply chain evil’, tolerated in order to accomplish manufacturing and product delivery. Recently, the organization has begun to view the logistic system as a value adding process instead. The product range consists of TV’s, stereo equipment, loudspeakers, etc., and the products are placed in the upper price segment in the market, focusing on a unique design and value-added service. Product examples are shown in Figure 2. The company follows a service leader strategy rather than a cost leader strategy (cf. Christopher, 1998). This is counteractive to most competitors (e.g. Phillips and Sony) but they are also much larger than Bang & Olufsen. Philips, for example, had 219.429 1 employees compared to Bang & Olufsen’s 2.783 employees in the year 2000. Philips turnover in the same year was 37.862 million Euro, while the turnover in Bang & Olufsen was app. 465 million Euro. The difference between Bang & Olufsen and Sony would be even bigger. Because of the service leader strategy the company constantly seeks new ways of adding to customer satisfaction, which is reflected in the company vision: ”Courage to constantly question the ordinary in search of surprising, long-lasting experiences”. The supply chain is shown in Figure 3. The decoupling point is situated at the Assembly plant, where only a small inventory of semi-finished goods is held. The first-tier suppliers, serv-

ing the assembly plant, are Bang & Olufsen’s own Electronic and Mechanical plants and some selected suppliers. Second-tier suppliers serves the Electronic and Mechanical plants. Product creation teams consist of designers from various groups: Electronics, Mechanics and Software. Furthermore representatives from Operations are invited. Traditionally, those attending have included representatives from the production plants and from purchasing, each ”looking out” for their own domains, and with no appointed spokesman (i.e. no one has responsibility for the overall perspective). The logistics organization is divided into two parts. One part is a Logistic Network, i.e. logistic coordinators in each facility making the daily operation run. The other part is a Supply Chain Development department that continuously works on improving and developing the company’s supply chain concept. A Director of Logistics, directly referring to the Managing Director, is heading the logistics organization. The company experiences a growing complexity in their supply chain management. Also, despite their service leader strategy, they need to cut operational costs to stay competitive. Though being niche leaders, competitors are moving closer and also customers demand to logistic service increase, for instance home delivery. Therefore the supply chain development depart-

1 Phillips’ homepage (2001)

49

E. Gubi: Design for Logistics

ment initiated a process of improving the logistic organization’s influence on new product creation. One of the steps was to initiate a research project together with the Center for Industrial Production, hereby providing the research opportunity. The author is placed in the Supply Chain Development department for a three year period with the logistic director as sponsor. The research project is supervised by a research steering committee, consisting of two senior researchers from the Center for Industrial Production, plus top managers from both R&D and Operations in Bang & Olufsen. After illustrating the objective and approach of the research the existing body of knowledge is reviewed.

2. Literature review The existing literature contains some contributions on how supply chains should be created based on the nature of the product (e.g. Fisher, 1997; Christopher, 1998; Fine, 1998). The opposite approach, Design for Logistics (DFL), is not so thoroughly covered in the literature. Therefore, this paper is based on the idea to investigate the wider known Design for Manufacturing (DFM) in order to extract elements to sup-plement the existing DFL-knowledge. The section starts with a presentation of the DFM concept and a discussion of selected contributions. Also Design For Assembly is briefly discussed. Hereafter some elements in DFM are addressed in order to adapt them to a DFL context. Lastly, a simple framework for deploying DFL in the case company is proposed.

2.1. The early days – Design for Manufacturing Companies creating and manufacturing complex products has been familiar with Design for Manufactur-ing for many years. Even though the discipline got its breakthrough in the 1960’s, the underlying thoughts can be dated back to the

50

very birth of industrial production: ”In one case we found that by using two cents more worth of material in a certain small part we were able to reduce the total cost of it by 40%. That is, the amount of material under the new method cost about two cent per part more than the old, but the labour was so much faster that, under the new method, the cost which was formerly $.2852 was now only $.1663.” (Henry Ford - quoted in Herbertsson, 1999). What Henry Ford is referring to is the very essence of Design for Manufacturing; that some design choices from one (e.g. the purchasing) point of view may seem improper, but from another (e.g. the manufacturing) it may seem well disposed. Later the term evolved into an actual discipline consisting of several methods for lowering manufacturing costs. In 1960, the General Electric Company developed a Manufacturing Producibility Handbook for internal use in the company (Bralla, 1996). The term manu-facturability further evolved, and in the 1980’s the term Design For Manufacturing and its abbreviation, DFM, came widely into use. In the 1960’s Design For Assembly (DFA) gained serious interest as a result of increasing wages. In par-ticular Design for automatic Assembly became widespread as a means of lowering assembly costs (Herbertsson, 1999). Today, it seems from the literature that DFA covers two perspectives: 1) the actual activity of assembling (by hand or automated) and due to vertical disintegration of companies 2) the as-sembly sequence throughout the supply chain. The first perspective focus on ”assembliability”, joining principles, etc. (e.g. Boothroyd & Dewhurst, 1994), and will often focus on a specific assembly system. This perspective can be considered a special instance of DFM, where the manufacturing task is that of assembling parts. The second perspective is more towards the ideas of DFL, since the assembly sequence heavily impacts the supply chain costs of offering product variety (e.g. Martin & Ishii, 1997). This second perspective is interesting from a DFL viewpoint.

E. Gubi: Design for Logistics

Many contributions have been made in the area of DFM (e.g. Bralla, 1986; Fabricius et al., 1994; Ulrich & Eppinger, 1995). They state that DFM needs to be performed throughout the entire product creation process, from the very beginning of concept development to final production. Ulrich & Eppinger (1995) mention DFM as an integrative methodology involved in product creation and suggest the need for a crossfunctional team to perform DFM. According to Fabricius et al. (1994), DFM fits into the wider framework of Concurrent Engineering (CE), by supplying both a basic mindset and concrete guidelines to ”synchronize” the development and the manufacturing. DFM is thus a bridging tool between two types of engineering, New Product creation and Operations Management, and concurrence refers to designers and manufacturing engineers working to-gether. Other parts of the literature regard concurrent engineering a tool in DFM (cf. Herbertsson, 1989). Finally some refers to CE and DFM as being synonyms (e.g. Fine, 1998). It therefore seems that there is some confusion in the exiting body of knowledge of the interrelationships between CE and DFM/DFA. Also, an unambiguous definition of DFM (or CE for that matter) does not exist. Nevertheless, there seems to be an agreement on the axiom that a great deal of the manufacturing costs is disposed during product creation. Therefore, the term is related academically to the theory of dispositions (Olesen, 1992). This theory treats relationships between parameters of a product and the parameters of the system that realizes the product. Further the theory of dispositions states that a large portion – often estimated to 70-80% - of a product’s lifecycle cost is ”locked” in the early design phases. Figure 4 shows this suggested relationship between allocated (disposed) and used product cost in a product creation project. Since logistics is a part of the product lifecycle, the logistic cost is likewise allocated in the early phases of product creation. With a large share of a product’s value stemming from out-

Figure 4. Relationship between allocated and used product cost side the company, from the suppliers, considering the entire supply chain when designing products seems uttermost rational. This assumption is a basic proposition in the research project. In the following the above-discussed concept DFM will be extended to a supply chain context.

2.2. Broadening the scope When Henry Ford built his Highland Park assembly plant in 1913, he reached the maximum level of vertical integration. The Ford Company produced every raw material needed for producing a car: they even grew rubber trees in order to extract rubber for the tires. While this vertical integration was the overall paradigm for industrial production for many years it is far from practice in today’s business where a considerable part of the company’s product cost stem from purchased materials and parts (Dowlatshahi, 1996). This change in business context calls for an extension of the term manufacturing to include the entire supply chain, which consist of both manufacturing processes and sourcing, storing, transportation etc. Charles Fine claims that traditional DFM no longer provides a significant differential advantage and claims the need for three-dimensional concurrent engineering, adding design and development of

51

E. Gubi: Design for Logistics

the supply chain to the traditional productprocess unity (Fine, 1998; 2000). A supply chain is a chain, or rather a network (e.g. Christopher, 1998, Lamming et al., 2000), of produc-tion and assembling units linked together by logistic activities, e.g. transportation and storing, with the purpose of making products that satisfies the needs of the ultimate customer. The idea here is to broaden the scope from a process/company point of view to a more holistic approach embracing the entire supply chain. The pioneers within DFM, however, cannot be accused of excluding this perspective - it simply did not exist in the early days of industrial production, i.e. in the vertical integrated companies. Therefore it was not considered.

2.3. Different DFM perspectives Design for Manufacturing and the theory of dispositions is comprehensively investigated from a product approach. Fabricius et al. (1994) suggested links between the manufacturing system and the product designs considered at several levels: • Corporate level: The interaction between the product and other types of company products • Family level: The relationship between different variants in the same product family • Structural level: The relationship between the different sub-systems/components • Component level: The design/specification of each individual component. Beginning from the bottom, the component level aims at optimizing the individual components against individual manufacturing processes. At the next level, the aim is to optimize the composition of these individual components for the most cost effective production system. The family level considers the flexibility of the production system in order to produce more variants of the same product, including considerations of shared components and/or modularization. The final level requires considerations

52

about shared components and modularity across product families and reduction of need for flexibility in the production system in order to maintain cost effective productivity on a company level. Instead of investigating dispositional mechanisms from a product point of view it is possible to take a system-oriented perspective. A supply chain consists of several companies/facilities, each featuring one or more processes. Therefore DFM/DFL can be considered at several system levels: • Supply chain level: The overall perspective including inter-company activities/processes • Company/facility level: The individual production and assembly systems • Process level: The individual manufacturing and logistic processes At the process level traditional DFM is situated to deal with an optimization of the product components for the manufacturing processes. At company level, DFA is useful for optimizing the product (or part of a product) - which the company delivers to its customers - towards the company’s production and assembly system. Boothroyd & Dewhurst formulates in this way: ”To manufacture refers to the manufacturing of the individual component parts of a product or assembly and that to assemble refers to the addition or joining of parts to form the completed product” (Boothroyd & Dewhurst, 1994). At the supply chain level Design for Logistics should be practiced, aiming at optimizing the product structure against the entire supply chain. Though the supply chain is constituted of companies and processes, DFL is not solely the sum of the individual (sub-optimized) DFM and DFA solutions/suggestions. Because of the system perspective, it is possible that DFL solutions counteract one or more individual processes/companies in order to get a better overall performance of the system. By having proposed DFL a methodology that embraces the entire supply chain, we now turn to discuss the contents of the term, as proposed in the literature.

E. Gubi: Design for Logistics

2.4. Design for Logistics In 1992, the practitioner Mather called attention to that products should not only be cheap to produce and assemble; they should also ”be designed so the customer can be delighted with availability, responsive-ness, and flexibility to the marketplace dynamics”. The wit is that a company, unable to provide availabil-ity when the customers need the products, need not at all to consider how easy it is to produce (Mather, 1992). His answers are standardization of components and modules, and adding variety at the latest pos-sible moment. However, he does not give any suggestions to how this should be applied, either in general or in a specific company. This seems to characterize later contributions as well. Lee (1992) who suggest that the following elements are important in DFL (DFSCM): • Delayed product differentiation • Localization • Part commonality, concurrent processing and decoupling of tasks • Product line restructuring Delayed product differentiation is also known as Postponement, (Pagh & Cooper, 1998). The benefits are flexibility; lower stock levels (not finished products but subassemblies) and more precise forecasts, due to aggregation of forecasts for each product variant into one forecast for the common parts. According to Lee (1992) product designs that allow for delayed product differentiation usually involves a modular structure of the product. Hence modularity is an important design strategy in supporting logistic leverage (also Martin & Ishii, 1997). The logistic cost of storing and moving materials and products through the supply chain and finally to distribute it to the customer can be a significant part of the product costs. Therefore localization of the entities in the supply chain should be considered when designing the product. It can be argued that selecting suppliers close to the focal company can solve this, and indeed in the automotive industry clustering of

suppliers around the car manufacturer has been predominant. In smaller companies, like the case company, it is much more difficult to force relatively larger foreign suppliers to co-locate around the company. Similarly (technology) suppliers are chosen because they posses a unique technology and they can too be spread out in the world. Again, for smaller companies the bargaining power of the case company is not very high towards these suppliers. Part commonality refers to using the same entities for different products and even different product families. This issue has also been pointed out by Martin & Ishii (1997) and is further related to both modularity and product platforms (Meyer & Lehnerd, 1997). The benefits are cost savings in part number administration, inventory reduction, and supplier management. Further it can leverage delayed product differentiation. Product line structuring relates to the structuring of the products so that the many models and versions of end products are assembled from relative independent subassemblies and auxiliary systems. The benefit is that these subassemblies can then be manufactured concurrently instead of serially, reducing manufacturing lead-time and increasing flexibility to meet customer demands. Other scholars in DFL are Dowlatshahi (1996) and Simchi-Levi et al (2000). Dowlatshahi argued that DFL was only little discussed in the literature: ”a review of literature reveals that little or no work has been done on the interface between product design and logistics.” He proposed that a framework for involving logistics in a concurrent engineering environment consists of four subsystems: Logistics Engineering, Manufacturing Logistics, Design for Packaging and Design for Transportability. Simchi-Levi et al extensively draws on Lee’s work, discussed above, in their design for logistics approach. The existing body of knowledge does not say exactly what should be done in DFL, but state a need for considering logistics in product creation. Issues pointed out as important to take into account are product variety (Gubta &

53

E. Gubi: Design for Logistics

Krishnan, 1998) and postponement (Pagh & Cooper, 1998), and part commonality and modularity (Martin & Ishii, 1997). This section has discussed the concepts of DFM, DFA and DFL. In the following some selected DFM methods is considered, with the purpose of transferring them to a DFL context. First though, before exploring selected issues in the DFM toolbox, we take a look at the interaction between logistics and manufacturing.

2.5. Manufacturing vs. Logistics The interrelationships between logistics and manufacturing has been discussed by several authors (e.g. Chikán, 2001; Pagh, 2000; Bowersox & Closs, 1996). One thing mentioned is that manufacturing focuses on capacity exploitation while logistics focuses on flow. Narrowly viewed it can be said that the production system favors big lot sizes and few changeovers while the nature of logistics is to move the goods to the next stage in the supply as soon as possible in order to meet customer demands. Chikán (2001) lists some common features and conflicting focuses. The common features are e.g. that both functions concentrate on physical activities such as moving, storing and form transformation of goods. Though both functions have a strategic context their activities have to be operational since they provide a direct impact on the daily profitability of the company. Supplementary to the common features some conflicting focuses exist. These conflicts arise even though the two functions strive for the same goal because they pursue them with different means. Both functions add value to the product, but in different ways: manufacturing ads form value where as logistics ads time and place value. As a consequence of the value type manufacturing will give emphasis to quality, while logistics will emphasize costs. Organizationally the two functions also are different since the manufacturing sub-functions are relatively concentrated in the organization as opposed to the

54

logistic sub-functions, which are spread out through the organization. Chikán stresses the importance of solving these conflicts in order to minimize the frictions (costs) in the company. Pagh (2000) proposes that the task of the entire logistic system – that is the task of the overall supply chain - should provide the background for determining the tasks of the manufacturing and logistic functions. Since the supply chain comprises manufacturing systems as well as supply and distribution systems, the objectives of these two other functions/systems also need to be included in the logistic objective. In this way, when practicing Design for Logistics purchasing, storing, moving etc. is considered. Finally, the objective must be aligned with other systems in the company – e.g. the quality system.

2.6. Exploring the DFM toolbox The existing DFM methods make up a wide spectrum of contributions. Boothroyd & Dewhurst (1994) are very hands-on, where as the methods suggested by Fabricius et al. (1994) are more conceptual. Also the aforementioned product levels are reflected in the treated contributions. Bralla (1986) focuses mainly on the Component level (e.g. castings, molded and machined components). Other authors seems to focus more on the Structural and Family levels and addresses the product architecture: ”… the scheme by which the function of a product is allocated to physical components” (Ulrich, 1995). The product architecture is an important concept when discussing DFL. This is where the product is divided in to parts, that need sourcing considerations (make/buy), and where the interfaces, later to be joined, are defined. Finally, some discuss DFM in an organizational context (e.g. Herbertsson, 1999). Table 1 summarizes selected elements in DFM. While most of the contributions in Table 1 provide tools and methods, the contribution from Herbertsson (1999) considers the actual implementation in a company. Herbertsson divides

E. Gubi: Design for Logistics

Author

Elements

Boothroyd & Dewhurst (1994)

Specific manufacturability process considerations Supplier involvement in product creation Cross-functional teams with clear and common objectives Cross-functional teams Metrics to compare alternative designs Three laws for DFM Implementation of DFM in an Enterprise

Fabricius et al. (1994) Ulrich & Eppinger (1995) Herbertsson (1999)

Table 1. Selected elements in DFM DFM into three main ac-tivities, as shown in Figure 5. As an example, if we assume that the perception of DFM in Figure 5 can be extended to a supply chain perspective (DFL) the application in the case company would be: Preparatory DFM (DFL) is a task for a supply chain development department, in cooperation with the logistic organization in the compa-

ny, and includes the continuous development of a competitive logistic concept that reflects product and customer needs. The abilities and constraints of such a competitive concept must be translated into new DFL-knowledge and applied at the two underlying levels of DFL. Supporting DFM (DFL) refers to the continuous improvements of the activities that deal

Figure 5. Three main activities in DFM (adapted from Herbertsson, 1999)

55

E. Gubi: Design for Logistics

with the material flow. This task is mainly to investigate the frictions between the (new) logistic concept and the existing products. Further the supporting task is to bring this knowledge into existing and new DFL-tools and design methodologies for the product creation projects. Operational DFM (DFL) is carried out in distinct product creation projects. It consists of various DFL-tools, the design methodology and cross-functional knowledge. This DFL activity is a task for each of the individual product creation teams with an overall responsibility placed with the project managers. Other companies will have a different division of the tasks depending on the structure of their logistic organization. Since the division of tasks in figure 5 is not specifically related to DFM, but could be about any continuous improvement effort, it is considered applicable for DFL as well. Figure 5 lacks feedback from the operational level to the two overlying levels, which will be needed in order to continuously adjust and improve the DFL tools and methods. Table 1 lists some factors being important elements in DFM. In order to use them in constituting a proposal for a framework for Design for Logistics these are discussed in the following. 2.6.1. The three laws of DFM Generally, all approaches to DFM shares the same goal: to give the product satisfactory manufacturing properties. In doing this, manufacturing consequences of the product design is evaluated during product creation. In order to do this the manufacturing processes must be known. Also there must be an objective for performing DFM, and some measures must be present. Herbertsson (1999) have proposed three laws for DFM: 1. DFM cannot be performed without knowledge of the manufacturing process 2. DFM must be performed with a specific objective in mind 3. A measure of manufacturability must be defined

56

It is assumed that that these laws apply to all Design for X activities in the product creation process, since any design property needs knowledge about its application area as well as objectives and measures in order to decide whether the activity was successful or not. Therefore these laws are suggested to apply to DFL as well. 2.6.2. Cross-functional teams Concurrent Engineering means that employees from various functions - traditionally design and manu-facturing engineers - are involved in a company’s product creation to ensure that vital knowledge of manufacturability is available. If we look to the supply chain (e.g. Figure 3) DFL cannot be performed merely by involving delegates from the individual links in the chain since, according to the chosen per-spective, a holistic knowledge is needed. This among other things calls for supplier involvement in prod-uct creation, but also for knowledge about supplier relations, performance etc. Finally knowledge of the overall performance of the supply chain is needed, which can be regarded as being the message in the three laws of DFM (knowledge, objectives and measures). Cross-functional teams are mainly assumed to be a prerequisite for DFL activities at the operational level (cf. figure 5). However cross-functional coor-dination is also needed at the preparatory and supporting level where it is accomplished through meetings. At the operational level some members of cross-functional teams could be representatives from selected suppliers. In the SCM literature supplier involvement in product creation is viewed as a groundbreaking step to de-sign forward-looking products better and be able to achieve order fulfillment in a more effective way (e.g. Monczka & Morgan, 1996). Taking into account that up to 70% of product costs stem from supplied ma-terials and parts, and keeping the theory of dispositions in mind, the benefits of early supplier involve-ment seems evident. On the other hand this should only apply for selected suppliers in order to

E. Gubi: Design for Logistics

counteract unfocused supplier integration. For this purpose supplier segmentation (e.g. Kraljic, 1983; Bensaou, 1999) should be considered. 2.6.3. Timing DFL Extension of DFM methods into DFL methods does not exclude the need for DFM – or DFA. As aforementioned the three DFX’s constitute a hierarchy with a supply chain level (DFL), a company level (DFA) and a process level (DFM). It has been proposed that once the product architecture is revealed the extent to which the product impacts the supply chain can largely be determined (Erens & Verhulst, 1997). Thus DFL should be applied in this phase. At this time, however, DFM and DFA are not that applicable since the detailed design yet remains. In the System-level Design phase the physical chunks are fixed, and their interfaces are specified, and thereby it is defined how the product will be assembled. In the Detail Design phase the individual parts that constitute the chunks are designed and thus the manufacturing processes are selected. Thus, the three DFX should be applied sequentially during the product creation process. Figure 6 shows the timing of DFL, DFA and DFM in a generic NPD process model. 2.6.4. Objectives and Metrics The need for objectives and metrics are stated as DFM law no. 2 and 3 (Herbertsson, 1999). This is not a trivial point since logistic performance is

measured in the operating system and not in the development system. When a product is released for sale it is possible to measure delivery performance and logistic costs. But how is it done before, and during, new product creation? Martin & Ishii (1998) proposes three indices for evaluating a given product architecture: • Commonality Index (CI): Measure of how well the design utilizes standardized parts • Differentiation Index (DI): Measure where differentiation occurs within the process flow • Setup Index (SI): Indirect measure of switchover costs contribution to the overall product costs The DFL approach presented in the next section also will address this issue.

3. A DFL approach Having discussed selected elements in DFL and when to deploy DFL, a simple framework is now proposed. It is based on the previous discussion and is recently proposed to the case company, where it will be tried out in a forthcoming new product creation project. In order to evaluate the logistic performance (cost, flexibility, lead-time, etc.) of a given product architecture, that is before the product is detailed designed, it is proposed to sketch supply chain scenarios for three alternative product architectures. The scenarios are based on the most important parts, those that will carry to product functionality, and therefore are likely to be known very early in the NPD project.

Figure 6. Sequential DFX (product creation process model adapted from Ulrich & Eppinger, 1995)

57

E. Gubi: Design for Logistics

These scenarios will reveal the number of links (tiers) in the supply chain as well as the ramification. The number of links is nevertheless just one of the factors used to calculate the possibility of meeting the de-sired performance (lead-time, flexibility, etc.). Another factor is where the product is made customer-specific. Based on a supplier segmentation model the necessary resources required from the case company, to run the supply chain in each alternative scenario, are estimated. Questions such as how reliable each of the suppliers are and how competently they manage their suppliers will need to be answered. At the same time each scenario is evaluated in a quality and a cost perspective. Afterwards each scenario is described and together with the sketches these descriptions will make up the basis for top management’s choice of scenario. The case company follows a stage-gate model as the one in Figure 6, where the overall product creation process is divided into manageable sub-processes, connected through gates where management can decide whether the project should continue or not (go/kill). In phase 1 just a few members are appointed to the project team, all of them being from the product creation department. These members are called the product architects, since they establish the product

architecture. Subsequently additional members are invited to join the team, some of them, as stated earlier, industrial engineers with the objective to promote product designs that will fit their functional area well. Now, it is proposed that a realization group will be established parallel to the product architects (see Figure 7). The task for this group is to draw up the before mentioned supply chain scenarios, based on a conceptual description of the project, and evaluate these before phase 1 is completed. The group will con-sist of engineers from each of the three plants (Mechanics, Electronics and Assembly) and personnel from the Central Purchasing Department, who are responsible for supplier relations and hence supplier in-volvement in product creation. Lastly the group will be assisted by a supply chain architect, an employee with profound supply chain understanding, in order to impart overall supply chain knowledge. In this manner cross-functional knowledge is present. To both groups a project manager will be appointed with responsibility for coordination the work of the two groups. The reason for not forming one big cross-functional team is that it is the experience of the company, that too large teams are inefficient. Since the product architecture is not yet fixed in phase 1, the supply chain scenarios will

Figure 7. Future organization of architectural phase at Bang & Olufsen

58

E. Gubi: Design for Logistics

be based on a few universal parameters (links, distance, supplier capability, maturity of technologies etc.) supplemented with the experience of the members of the realization group. Therefore the selection of team members is based on their level of experience in the company as well as their knowledge about the company supply chain according to Herbertssons law no. 1. After getting some first experiences with the framework it will be analyzed and revised for further use in the company. This could in time diminish the need for a high level of experience in the team members. The scenarios are visualized by placing small plastic bricks on a plastic board, where every link in the network will be indicated with a brick having a color: green, yellow or red. The colors indicate the rating of the link, where the rating shows the overall judgment of the parameters. A green brick indicates ‘no danger’, a yellow indicates ‘warning’ and red indicates ‘danger’. This ‘traffic light’ rating is an ordinal scale, i.e. the rating is relative and directly not comparable across companies, industries etc. Likewise the whole scenario is rated green, yellow or red based on an overall judgment of the composition of the color of the applied bricks. If a scenario contains one or more yellow bricks the ‘yellow’ parameters must be described accompanied with an indication of where the responsibility for handling the specific parameter is based in the company. If a scenario contains one or more red bricks a plan of action must be presented. Here it should be noticed that red is not necessarily bad – it just requires action. The objective (law no. 2) of this work is to visualize potential pitfalls in the scenarios and thereby be able to act proactively. Evaluation of the scenarios will be presented for the top management of the company as a part of the documentation for phase 1 in the product creation process. This does not necessarily mean that management will choose the scenario that suits logistics best - parameters like cost and quality are evaluated as well. However, it ensures that the selection is made on a quali-

fied basis and that potential problems become visible and can be dealt with proactively. For the logistic organization this will be an improvement compared to the present situation. At the same time the evaluations can be used for sharing experience across development projects. The company has already pilot tested this method, albeit when the product architecture was fixed and furthermore only a few parameters were contained in the evaluation. Further the objective of this approach was not made clear to the product creation team. Therefore the result was inadequate for decision-making. With this lesson learned the objective will be made clear to the product creation team and the responsibility for the evaluation of logistic, quality and cost parameters of the scenarios will be set in place prior to implementation. Though this approach applies to the operational level, metrics and objec-tives are needed in all three main activities of DFL. Consequently, metrics and objectives for the pre-paratory and supporting levels need to be developed too. The next step in the research project will be to test the approach in the case company. Hereafter a complete DFL framework will be developed.

4. Discussion The DFL-approach proposed in this section relies on existing body of knowledge within DFL combined with existing theoretical contributions from the closely related field DFM, transformed into DFL-application. The latter should be subject to careful consideration since DFM often relates to the product level (cf. Fabricius, 1994) whereas DFL should address the product family or product portfolio levels. Reuse of a module across products will not necessarily benefit component manufacturing to the same extent as the positive impact on logistic performance. Also DFM/DFA means optimizing products to a certain manufacturing or assembly system, where as DFL targets the performance of the entire supply chain.

59

E. Gubi: Design for Logistics

Most existing DFX literature considers optimizing single products towards the X. When discussing DFL we should focus on the product portfolio, encouraging the use of product platforms and modularity. The critique raised by Paashuis & Boer (1997), that CE is not widely implemented in industry, mainly due to a lack of normative methods on how to configure CE to the specific company situation has been addressed. By testing the idea proposed in this paper and continuously revise it to fit the context of the company it is the anticipation that some guidelines for making CE company-supportive can be found.

Bowersox, D. J. & Closs, D. J. 1996, Logistical management, 3. edn, McGraw-Hill.

4.1. Summary

Fabricius, F., Ahm, T., Christensen, B., Olesen, J., Hein, L., & Mørup, M. 1994, Design for Manufacture, DTU

This paper provides a literature review on the existing knowledge of Design for Logistics (DFL), and the related concepts of Design for Manufacturing (DFM) and Concurrent Engineering (CE). The literature review shows some confusion on the connection between DFM, DFL, etc. and CE. Furthermore, there is a lack of normative methods; especially how to make CE specific to companies (Paashuis & Boer, 1997). The existing literature base within Design for Logistics is found scarce and rather conceptual. Thus, the wider area of DFM has been explored in order to find some elements that can be used in a DFL context as well. In addition, this paper has presented a company-supportive approach to Design for Logistics, yet initial and untested. It needs to be tried out and continuously adjusted and improved. This will be initiated in the case company and a future paper will have to tell this story.

References Bensaou, M. 1999, Portfolios of Buyer-Supplier Relationships, Sloan Management Review, Volume 40, Number 4, pp. 35-44 Boothroyd, G. & Dewhurst, P. 1994, Product Design for Manufacture and Assembly, Marcel Dekker.

60

Bralla, J. G. 1996, Design for excellence, McGrawHill, New York. Chikán, A. 2001, Integration of Production and Logistics - in Principle, in Practice and in Education, International Journal of Production Economics, vol. 69, no. 2, pp. 129-140. Christopher, M. 1998, Logistics and Supply Chain Management, 2. edn, Pitman Publishers. Dowlatshahi, S. (1996), The role of logistics in concurrent engineering, International Journal of Production Economics, vol. 44, no. 3, pp. 189-199

Fine, C.H. (1998): Clockspeed: Winning Industry Control in the Age of Temporary Advantage, Perseus Books Fine, C.H. (2000): Clockspeed-Based Strategies for Supply Chain Design, Journal of Production and Operations Management, vol. 9, no. 3 Fisher, M. L. 1997, What is the Right Supply Chain for Your Product?, Harvard Business Review, vol. 97205, no. March-April, pp. 105-116. Gubta, S.& Krishnan,V.(1998), Product family-based assembly sequence design methodology, IIE Transactions, vol 30, pp. 933-945 Herbertsson, J. 1999, Enterprise Oriented Design for Manufacture, PhD disertation, Linköbing University, dept. of Mechanical Engineering. Hoover,W.E.; Eloranta,E.; Holmström,J.; Huttonen,K. (2001), Managing the Demand-Supply Chain Value Innovations for Customer Satisfaction, John Wiley & Sons inc. Kraljic, P. 1983, Purchasing Must Become Supply Management, Harvard Business Review, vol.9, no.1. Lambert, D., Cooper, M. & Pagh, J. 1998, Supply Chain Management: Implementation Issues and Research Opportunities, International Journal of Logistics Management, vol. 9, no. 2

E. Gubi: Design for Logistics

Lee, H. L. 1992, Design for Supply Chain management: Concepts and Examples, MIT working paper Martin, M. & Ishii, K. 1998, Design for Variety, Product Variety Management, Research Advances, Kluwer Academic Publishers, USA Meyer, M. & Lehnerd, A. 1997, The Power of Product Platforms - Building Value and Cost leadership, Free Press, N.Y. Monczka, R. & Morgan, J. 1996, Supplier Integration: A new level of supply chain management. Purchasing 120 (1), 110-113.. Boston, Cahners Magazines. Ref Type: Magazine Article Olesen, J. 1992, Concurrent Development in Manufacturing - based on dispositional mechanisms, Ph.D. dissertation, Technical University of Denmark (DTU) Pagh, J. D. 2000, Udvikling af virksomhedens transformationssystem, Danish Ph.d. dissertation, DTU

Pagh, J. D. & Cooper, M. C. 1998, Supply Chain Postponement And Speculation Strategies: How To Choose The Right Strategy, Journal of Business Logistics, vol. 19, no. 2, pp. 13-33. Paashuis, V. and Boer, H. 1997, ”Organizing for concurrent engineering: an integration mechanism framework”, Integrated Manufacturing Systems, Vol. 8, No. 2, pp. 79-89. Simchi-Levi, D., Kaminsky, P. & Simchi-Levi, E. 2000, Designing and Managing the Supply Chain, McGraw-Hill Tissan, T. & Heikkilä, J. (2001), Successful Re-engineering – Learning by doing, Int. Journal of Logistics: Research and Applications, Vol.4, No. 3, pp. 329-344 Ulrich, K. T. 1995, The Role of Product Architecture in the Manufacturing Firm, Research Policy, vol. 24, pp. 419-440. Ulrich, K. T. & Eppinger, S. D. 1995, Product Design and Development, McGraw-Hill.

61