A DESIGN METHOD FOR DEVELOPING A UNIVERSAL PRODUCT FAMILY IN A DYNAMIC MARKET ENVIRONMENT

Transcription

1 Proceedings of IDETC/CIE 2009 ASME 2009 International Design Engineering Technical Conference & Computers and Information in Engineering Conference August 30-September 2, San Diego, California, USA DETC A DESIGN METHOD FOR DEVELOPING A UNIVERSAL PRODUCT FAMILY IN A DYNAMIC MARKET ENVIRONMENT Seung Ki Moon and Daniel A. McAdams* Department of Mechanical Engineering Texas A&M University College Station, TX USA ABSTRACT Innovative companies that generate a variety of products and services for satisfying customers specific needs are invoking and increasing research on mass-customized products, but the majority of their efforts are still focused on general consumers who are without disabilities. This research is motivated by the need to provide a basis of universal design guidelines and methods, primarily because of a lack of knowledge on disabilities in product design as well as methods for designing and evaluating products for everyone. Product family design is a way to achieve cost-effective mass customization by allowing highly differentiated products to be developed from a common platform while targeting products to distinct market segments. By extending concepts from product family design and mass customization to universal design, we propose a method for developing a universal product family to generate economical feasible design concepts and evaluating design feasibility with respect to disabilities within dynamic market environments. We will model design strategies for a universal product family as a market economy where functional module configurations are generated through market segments based on a product platform. A coalitional game is employed to model module sharing situations regarding dynamic market environments and decides which functional modules provide more benefit when in the platform based on the marginal contribution of each module. To demonstrate implementation of the proposed method, we use a case study involving a family of mobile phones. * 1. INTRODUCTION Innovative companies that generate a variety of products and services for satisfying customers specific needs are invoking and increasing research on mass-customized products, but the majority of their efforts are still focused on general consumers who are without disabilities [1, 2]. Persons with limitations due to age and disabilities are not a static population with static abilities. We will all likely be disabled at some point, be it temporarily through injury or permanently through injury or the effects of age. The number of people with a disability is between 40 and 50 million [3]. This number represents approximately 1 in every 7 Americans. Based on current demographic trends, particularly the aging population, the numbers of people with disabilities is expected to increase for the foreseeable future. Universal design is a recently suggested term for designing for persons with a disability [4]. Universal design specifically suggests the concepts of equity and social justice. Also, in the context of separate is not equal, universal design suggests the design of solutions that simultaneously and equally serve both the fully able and not fully able. Design of new products for everyone requires numerous functions for many individuals and groups often separated by capabilities and limitations due to age and disabilities [5]. These functions can be represented by individual and collective background knowledge and new knowledge synthesized from information gathered during the design process. Modern product design practice consists of gathering customer needs, translating these needs into the required product functionality, synthesizing physical artifacts capable of providing the required functionality, and integrating this artifacts into an overall product: design moves from need, to function, to a form solution. The challenge then is to transform information obtained from an abundant of various sources into useful knowledge for effective universal design. This research is motivated by the need to provide a basis of universal design guidelines and methods, primarily because of a lack of knowledge on disabilities in product design as well as methods for designing and evaluating products for everyone. * Please address all correspondences to this author: dmcadams@tamu.edu 1 Copyright 2009 by ASME

2 For mass customization, companies are increasing their efforts to reduce cost and lead-time for developing new products and services while satisfying individual customer needs. Mass customization depends on a company s ability to provide customized products or services based on economical and flexible development and production systems [6]. By sharing and reusing assets such as components, processes, information, and knowledge across a family of products and services, companies can efficiently develop a set of differentiated economic goods by improving flexibility and responsiveness of product and service development [7]. Product family design is a way to achieve cost-effective mass customization by allowing highly differentiated products to be developed from a common platform while targeting products to distinct market segments [8]. In this paper, we extend methods from mass customization and product family design to create specific methods for universal product family design. The objective of this research is to develop a design method for a universal product family by generating economical feasible design concepts and evaluating design feasibility with respect to disabilities within dynamic market environments. In this case, the universal product family includes products for persons with and without a disability. We use a module-based platform design approach by introducing common modules, variety modules, and typical modules for universal product family design. To determine a preferred platform strategy, we will investigate which modules provide a greater contribution to producer profit if included in the platform. Because the design strategies for universal product family design can be modeled as a market economy where functional module configurations are generated through negotiation with customers, we will use market-based decision-making approaches to consider dynamic design and market environments in determining platform design strategies. A coalitional game is employed to model potential module sharing situations to determine which combination of modules used in the platform provide more benefit. Game theoretic approaches provide a rigorous framework for managing and evaluating strategies to achieve players goals using their information and knowledge [9]. The remainder of this paper is organized as follows. Section 2 reviews related literature and background for universal design and product family design. Section 3 describes the proposed design method for developing a universal product family using a coalitional game. Section 4 gives a case study using a family of mobile products. Closing remarks and future work are presented in Section 5. 2.LITERATURE REVIEW AND BACKGROUND 2.1 Universal Design A team of researchers organized through The Center for Universal Design at North Carolina State University has created seven principles of universal design [10]. The seven principles are: 1) equitable use, 2) flexibility in use, 3) simple and intuitive use, 4) perceptible information, 5) tolerance for error, 6) low physical effort, and 7) size and space for approach and use. For each principle, several guidelines have been created. For example, principle 6 has a guideline of minimize repetitive actions. These principles have been well received by designers in a range of disciplines. Though the seven principles of universal design provide high-level guidance, they provide more of an evaluation aid than a design or synthesis aid for product design. Vanderheiden [11] has developed a set of guidelines for the design of consumer products. Some of these guidelines are useful for evaluation but more challenging to use for product synthesis. Housed in the Center for Inclusive Design and Environmental Access at the University of Buffalo is an active group of researchers with focus on universal design [12, 13]. Though this group is focused on architectural design and comes from an architectural background, they have performed research on appliances and other applications that extend to product design. A team of researchers at the University of Cambridge has produced implementable results for universal design [14, 15]. The focus of this research group has been in modeling user groups, creating product assessment methods, and extending the needs of universal design to modern product design processes. The results of the Cambridge team are the most directly applicable to product design. Their effort has been primarily focused on the user and the design challenges of accommodating that user. Universal design is an active research area. Nevertheless, fundamental work applicable to product design is still a sparsely populated space. Universal design is more of an objective than a systematic design approach. There is little in the way of a prescriptive approach to universal design in more detail than broad design objectives [16]. Additionally, though creating modular products that minimize modification to become universal is a recognized approach to universal design, specific knowledge and methods strategies to do it do not exist [15]. Methods that allow the design of universal products that offer value to the user and profitability to the producer have yet to be thoroughly developed. 2.2 Product family and platform design A product family is a group of related products based on a product platform, facilitating mass customization by providing a variety of products for different market segments costeffectively [17]. A successful product family depends on how well the trade-offs between the economic benefits and performance losses incurred from having a platform are managed. There are two recognized approaches to product family design [18]: (1) a top-down (proactive platform) approach and (2) a bottom-up (reactive redesign) approach. In the top-down approach, a company s strategy provides a guide line for developing a family of products based on a product platform and its derivatives. Meanwhile, the bottom-up approach is focused on redesigning and/or consolidating a group of distinct products to standardize components for sharing and reusing. In platform-based product development, two common types for product families are module-based product family and scale-based product family [18]. Products in 2 Copyright 2009 by ASME

3 a module-based product family are obtained by adding, substituting, and/or removing one or more modules from the platform. In a scale-based product family, products are created by scaling one or more variables related to the platform design to satisfy a variety of market niches. Simpson et al. [18] introduced a method to optimize a platform by minimizing performance loss and maximizing commonality based on a scale-based product family design approach. Gonzalez-Zugasti et al. [19] designed platform modules to minimize design risk and save costs related to develop a product family. Siddique and Rosen [20] described a method to design a platform from an existing group of products by comparing commonalities in assembly processes. Rai and Allada [21] used a two-step approach to determine a modular platform for a product family, which consists of an agent-based optimal technique and post-optimization analysis using the quality loss function. Johannesson and Claesson [22] proposed a configurable product platform design process and model using an operative product structure and a hierarchical function-mean tree to capture parameters describing design information such as rules, variants, requirements, and product configuration possibilities. Thevenot et al. [23] developed the design of commonality and diversity method (DCDM) to provide designers with recommendations for both the functional and component levels by the inherent tradeoff between commonality and diversity during product family and platform development. Moon et al. [24] introduced a marketbased negotiation mechanism to support product family design by determining an appropriate platform level that represents the number of common modules using a dynamic multi-agent system in an electronic market environment. Zacharias and Yassine [25] proposed a mathematical model for developing and evaluating modular product families to provide maximum market coverage by integrating a conceptual design approach, a product development cost model, an economic model. In a highly competitive market, universal design can be considered as appropriate marketing strategies by providing the broadest market segment [5]. For example, safety and convenience are important design factors for everyone, especially the older consumers. In universal product family design for mass customization, a method to produce a variety of products should be considered for dynamic and various market segments to reflect a variety of customer needs and preferences. In addition, dynamic factors, like customer needs and trends, companies strategies, market situations, and technologies, should be considered to increase customer s satisfaction in developing a family of products. Therefore, we need to address how to reflect the dynamic factors in a family of universal products. Market-based product design is one way to reflect various and dynamic market environments in product design. In the next section, the proposed design method for developing a universal product family is discussed in detail 3. UNIVERSAL PRODUCT FAMILY DESIGN Figure 1 shows the proposed process for developing a universal product family based on the top-down and modulebased approaches in product family design. The proposed method consists of four phases: (1) identify market segments, (2) develop platform design strategies, (3) identify design quality, and (4) determine a platform strategy. Customer needs can be collected through surveys and market studies. The market study begins by establishing target markets and customers. In the initial phase, customer needs are analyzed to develop market segments for a universal product family. Customer needs are also used to identify required product functionality for individual products and across a range of products. Then, platform design strategies are developed by module-based design concepts. After evaluating different platform design strategies using universal design principles and a game theoretic approach, a final platform is determined to generate universal product family concepts according market segmentations and design constraints. A description of each phase follows Figure 1: Process of Developing a Universal Product Family 3.1 Phase 1: Identify Market Segments The division of a market into homogenous groups of consumers preference is known as market segments [26]. Because a market segment provides guidelines for determining and directing customer requirements, it can be used to identify the criteria for designing product family more accurately [17]. A successful product family requires balancing the trade-offs between the economic benefits and performance losses incurred from having a shared platform. The basic development strategy within any product family is to leverage the product platform across products that target multiple market segments. In the initial phase, customers are classified into groups based on their characteristics and preferences. Products are also clustered as groups for recommending to customers. For example, Meyer and Lehnerd [26] introduced three platform leveraging strategies based on market segments within the gird during a conceptual design phase. The market segmentation grid is useful for both platform development as well as product family consolidation. 3 Copyright 2009 by ASME

4 In universal product design, customers preference is determined by information related to customers accessibilities or functional limitations. Product reference information can help develop market segmentation for universal product family design by identifying an initial platform based on functional requirements. For example, Figure 2 shows three platform leveraging strategies for universal design based on approaches applied by Mayer and Lehnerd [26]. ;-3*'#( "18#:-6%<1,-+( 76-3*"8()12%'/( "-+9:*61,-+( "-+"#08((./0%"1'(2-3*'#(./0%"1'( 2-3*'#(./0%"1'( )*+",-+( ;-3*'#( 4+%5#61'(2-3*'#( 76-3*"8(=(>( (*+%5#61'(2-3*'#( 76-3*"8(?(>( (*+%5#61'(2-3*'#(!""#%&'#(2-3*'#(!""#%&'#( (2-3*'#(!""#%&'#( ()*+",-+( B"-+-2/( -)("1'#( 76-3*"8(+(>( (*+%5#61'(2-3*'#( 7'1A-62( B"-+-2/(-)("1'#( Figure 3: Module categorization and platform based product configuration concepts Figure 2: Three platform leveraging strategies for universal product family design 3.2 Phase 2: Develop Platform Design Strategies Universal Product Architecture We propose to use a modular product framework to build product families from common and varietal models. The notion of common and varietal modules is generally well understood in product family design. Common modules are those shared across the product family. In general, these common modules are suitable candidates for establishing the product platform. Varietal modules refer to the differing elements used to introduce variety into a range of products in the family. The common elements plus the varietal elements combined create a product family. The framework used to design a product family here is modular, but the notions of common and varietal need not be limited to a modular framework. In the context of designing a universal product family, we will also categorize modules as 1) universal; 2) accessible; or 3) typical modules. Universal modules are those that are the same in function and form for both typical and disabled users. Accessible modules provide specific functionality or form solutions for persons with limitations not desired by a typical user. Typical modules contain functional and form solutions, or both, that are not suitable for user with a disability. Given this modular framework, a product family can build on a common modular platform of universal or accessible elements. In the mobile phone product family example discussed in this article, the (initially) common elements are a universal module that is the circuit board and processing elements. Note that the product platform can also be a common accessible module that makes a family of products appropriate for a disabled user. The ways in which the modules can interact is illustrated in Figure 3. Approaching universal product family design with multiple notions of what constitutes common and varietal allows for the greater potential of leveraging economies of scale. Such thinking may be important when designers attempt to spread costs of an accessible module across as large a product family as possible Platform Design Strategies and Cost Model A designer determines a feasible set of strategies for the platform based on components specifications and his/her design knowledge. The strategies for a platform are represented as alternative designs and can be constructed by combining functions in accessible and typical modules. A well-defined platform reduces production costs by improving economies of scale and reducing the number of different components that are used [17, 24]. The term platform level refers to the amount of common elements in the product family platform. A high platform level indicates more sharing then a low platform level. An appropriate platform level for a product family can be determined by minimizing the production costs associated with commonality levels. For instances, high levels of the platform (i.e., high commonality) decrease interfaces between modules as well as the number of different components. Nevertheless, a high degree of commonality potentially decreases customer satisfaction due to design compromises made to increase commonalty. On the contrary, low platform levels (i.e., low commonality) may increase customers preference while increasing interface between modules and the number of components. As such, the appropriate platform level for the universal product family can be represented as a mathematical programming model in which production costs are minimized [24]. Based on generated product design concepts, product cost can be determined by life time product volume, material cost, direct labor, production resource usage, tooling and capitalization costs, system cost (overhead or indirect costs), and development costs [27]. To develop platform strategies based on current common modules, we introduce the expected strategy cost that represents additional costs for developing a new platform for a product family such as redesigning components. Therefore, a new platform is a combination of common modules and varietal modules. Suppose that a product family consists of I products, PF = (P 1,P 2,...P i,...,p I ). Let A be a set of strategies for additional modules that can be added to current common modules and c(s y ) the expected strategy cost for strategy s y ( y = 1,2,...,S ). Then, the expected strategy cost can be 4 Copyright 2009 by ASME

5 calculated based on additional design cost and the benefit of family design as follows [24]: i I c(s y ) = " # (1) f # r a where C i is the additional design cost of product i for developing a new platform, " is a factor for overhead cost, and f is a strategy weight function as follows: " 1, if a module is unique f = # (2) I, otherwise and r is a volume penalty factor related to products sales quantity. Hence, the expected total product cost, TC, for the product family using platform strategy, s y, can be calculated by: % C i a TC(s y ) = # C i + c(s y ) (3) i" I where C i is the product cost of product i. For a given set of products, the value of c(s y ) varies depending on the strategy for platform design. The expected strategy cost function will be used to determine a platform for a universal product family and can be developed by various cost functions based on products characteristics and/or company s strategy in product family development. The next section introduces a product (strategy) quality model for evaluating accessibility in a universal product. 3.3 Phase 3: Identify Design Quality A usability testing and a participatory design model help designers evaluate design characteristics and functionalities for universal products in an early design process or a conceptual phase [1, 28]. To evaluate and measure accessibility of a product, we propose a strategy quality function that is positively related to functional accessibility level (FL) and usability level (UL) as follows: Q = f (FL,UL) (4) The functional accessibility level represents the interaction of product functionality and product accessibility: it is a measure that indicates what functions are needed to make a product accessible to individuals who have a functional limitation as defined in the ICF [29]. To determine the functional accessibility level, we propose the use of the Function- Universal Principles Matrix (FUPM). This matrix is based on impairment and usability measure developed in the ICF [29] and the seven principles of universal design [10, 30]. Table 1 shows a FUPM template. The first two columns enumerate and list all the potential functions that may be needed by all products in the product family. Across the header row, the 7 principles of universal design are recorded. The last two columns contain the functional accessibility level and the usability level. Table 1: The proposed Functions-Universal Principles Matrix In the FUPM, the values, fl i (i = 1,2,...,n ), of the functional accessibility level for each function can be calculated as follows: n fl i = " a i # p up i, p (5) i where a i is the degree of importance for ith function in terms of accessibilities and the degree is determined based on the accessibility of impairment as follows: " 1 for No impairment 2 for Mild impairment a i = # 3 for impairment (6) 4 for Severe impairment % 5 for Complete impairment And, " p is the importance weight of pth universal principle in terms of functions ( p = 1,2,...,7 ) and can be determined by the Analytical Hierarchy Process (AHP) or group decision-making methods based on product s characteristics and utilization [31]. up i is a binary variable (0, 1) for indicating the dependence between functions and the pth universal principle. For the usability level, ul i, we categorize the usability of a function into five levels based on the difficulty of using the function with respect to impairment and capacity limitation [29]: (1) No, (2) Mild, (3), (4) Severe, and (5) Complete difficulties. The value of the usability level can be determined as follows: " 5 for No difficulty 4 for Mild difficulty ul i = # 3 for difficulty (7) 2 for Severe difficulty % 1 for Complete difficulty The expected strategy quality, Q i, for function i can be estimated by an expected quality function: f i : FL "UL! #. Hence, the real number of f i (FL,UL) represents the quality of strategy i having accessibility level FL for usability level UL. For example, the expected quality for strategy i can be determined as: f i (FL,UL) = fl i " ul i (8) The proposed strategy quality function will be applied to measure accessibility for determining product qualities in terms of platform design strategies. The next section discusses a 5 Copyright 2009 by ASME

6 coalitional game model for determining a platform design strategy Phase 4: Determine a Platform Design Strategy A coalitional game is designed to model situations wherein some of players have cooperation for seeking a goal in a game [9]. A coalitional model focuses on the potential benefits of the groups of players rather than individual players. In the coalitional model, the sets of payoff vectors are used to represent value or worth that each group of individuals can achieve through cooperation. A platform level problem in platform design can be considered as a strategic module selection problem under collaboration situation. The strategic game provides a useful technique for determining a strategy in dynamic market environments. In this paper, we employ a coalitional game to model module sharing situations regarding dynamic market environments and solve the functional module selection problem in given universal product family design. To determine modules for a platform, we decide which functional modules provide more benefit when in the platform based on the marginal contribution of each module. We assume that each module in a product is modeled as a player. Then, consider the following module selection problem for platform design in a dynamic market environment. Each potential coalition can be represented as a platform design strategy and be independent on the remaining players. To determine modules for platform design, we consider the set of all possible coalitions and evaluate the benefits of coalitions based on individuals preferences. In order to formulate the proposed scenario as a coalitional game, we must first identify the set of all players, N, and a function, v, that associate with every nonempty subset S of N (a coalition) [9]. A real number v(s) represents the worth of S and the total payoff that is available for division among the members of S. And, v satisfies the following two conditions: (1) v(") = 0, and (2) (superadditivity) If S,T " N and S "T = #, then v(s "T ) # v(s) + v(t ). Based on the definition of the coalitional game [9], the proposed game can be defined as: N: players who represent (varietal) modules v(s): the benefit of a coalition, S " N where a coalition, S, represents a platform design strategy that consists of several modules as mentioned in Section We use the Shapley value to analyze the benefits of family design and determine modules for platform design [32]. The Shapley value is a solution concept for coalitional games and is interpreted as the expected marginal contribution of each player in the set of coalitions [32]. Shapley value is defined as follows: (T #1)!(n # T )! " i (N,v) = & [v(t ) # v(t \ {i})] (9) n! T N,i%T where " i (N,v) is the payoff of player i, T is the players number in the coalition T, T \{i} is the players coalition excepted player i, n is all players numbers, and v(t) is the payoff of the coalition T. Through the Shapley value, we can determine a platform design strategy based on the payoff of each varietal module by the marginal contribution it makes to the platform. In this paper, we use sales profits of products to evaluate players coalitional benefits for their pay-offs. We assume that the price of a product can increase with increasing quality. Then, product i profit, " i can be formulated based on the design strategy quality and the product cost as follows: " i = Pr i # i # i C i = (Pc i Q i )# i # i C i (10) where, Pr i is the price of product i, C i is the product cost of product i, " i is the sales quantity of product i, Pc i is the coefficient of the price for product i, and Q i is the design strategy quality of product i. Hence, the payoffs of coalitions, v, can be calculated by difference between the expected sales profit and the current sales profit, and formulated by coalitional benefits for a universal product family based on a platform strategy, s y, as follows: v(s y ) = " i (s y ) % " i (s 0 ) i# I i# I (11) = (Pc i Q i (s y )& i %& i (C i + c(s y )) % (Pc i Q i (s 0 )& i %& i C i ) i# I where s 0 is the current design strategy. c(s y ) and Q(s y ) are estimated by the expected strategy cost function and the quality function as mentioned in Sections and 3.3, respectively. To determine the product sales qualities, we use the proportion of market segmentation grids that are covered by a product. The sales quality of product i, " i, is formulated as follows: " i = (# i,d % i,e TD) (12) ' d,e& M i where M i is the set of the market segmentation grids of product i, " i,d is the proportion of the market segmentation grids at x- axis (d=1,2,,d) for product i, " i,e is the proportion of the market segmentation grids at y-axis (e=1,2,,e) for product i, and TD is the amount of total demands for products in the market. For example, if Figure 4 shows the result of market analysis for product A and TD is 10,000. Then, the sales quality of the product A is 3,500. If several products are involved in the same segments, the segment ratio of each product is calculated by the proportion of the number of the products in the same segment. Figure 4: Example of Market Segments Identification for a Product i# I 6 Copyright 2009 by ASME

7 Based on the results of marginal contributions for different modules, we can determine a platform strategy based on consideration of better product sales form accessing broader market segmentations with the tradeoffs of increasing platform cost. The proposed method can provide designers with candidate modules for improvement and opportunities for redesign in dynamic market environments. In the next section, the proposed method is applied to determine a platform design strategy using a case study involving a family of mobile products. 4.1 Phase 1: Identify Market Segments Figure 6 shows current market segmentation grids for the mobile products with respect to vision features and market prices. The products have different vision accessibility features and market prices depending on market segments as shown in Table 2. For example, N73 covers No vision impairment and low price market. In Table 2, we can consider F2, F3, F4, F6, and F9 as common modules for the phone family. And, F1, F7, and F8 are considered as varietal modules. 4. CASE STUDY To demonstrate implementation of the proposed method, a family of mobile products consisting of N73, N76, N78-1, and N79-1 is investigated from the Nokia N70 phone family. These items are shown in Figure 5. The Nokia N70 series family products provide a good example of common and different functions for accessibilities related to vision features as shown in Table 2. These products offer the opportunity to create a product family with the vision features as common functions that constitutes the product platform. Figure 6: Market Segmentation Grids for the Four Products Figure 5: Nokia N70 Series Products [33] Table 2: Vision Accessibility Features for Four Products 1 Vision features N73 N76 N78-1 N79-1 F1 Tactile key markers Yes No Yes Yes F2 Standard key layouts Yes Yes Yes Yes F3 Key feedback - tactile Yes Yes Yes Yes F4 Key feedback - audible Yes Yes Yes Yes F5 Audible identification of Keys - when pressed No No No No F6 Audible identification of Keys - feedback Yes Yes Yes Yes F7 Adjustable font style No Yes Yes No F8 Adjustable character size No Yes Yes Yes F9 Display Characteristics (color display) Yes Yes Yes Yes The objective in this case study is to determine a platform design strategy represented in a dynamic market. This case study focuses on modifying a current product platform through the addition of accessible modules. Benefits of the proposed product platforms are based on the marginal contributions of the proposed additional modules Phase 2: Develop Platform Design Strategies In Phase 2, we explore configuration for developing platform design strategies by identifying relationships between functions and market segments at a conceptual design phase. In Table 3, we relate different components to the features they do, or do not, provide accessibility for a user with reduced vision functioning. We consider that a cell phone consists of eleven components [34]. Among the components, we assume that a main board includes a program for supporting all features. Table 3: Features-Component Relationship To develop a new platform consisting of common modules and varietal modules, we need to determine marginal contributions of the varietal accessible modules (F1, F7, and F8). The marginal contributions of modules can help decide 7 Copyright 2009 by ASME

8 which the functions are included to a new platform for increasing benefits and accessibilities in a universal product family design. Table 4 shows functional configuration strategies that consist of the varietal modules for the phone product family. To determine the expected strategy cost as mentioned in Section 3.2.2, we considered the number of components that are related to vision features and use a unit additional cost, C a, for each component. For example, since the Tactile key marker is related to two components, Upper case and Keypad, the additional cost of the Tactile key marker is 2 C a. We assume that a factor of overhead cost and a volume penalty factor are 2 and 1, respectively. The expected strategy cost for the product family can be calculated by Equation (1). Table 4 shows the results of the expected strategy cost for the platform strategies. Table 4: The Expected Additional Strategy Cost for the Platform Strategies Strategy Additional component design cost Expected N73 N76 N78-1 N79-1 Total additional strategy cost S1-F1F7 4C a 2C a - 4C a 10C a 5C a S2-F1F8 4C a 2C a - - 6C a 3C a S3-F7F8 8C a - - 4C a 12C a 6C a S4-F1F7F8 8C a 2C a - 4C a 14C a 7C a In the vision impairment point of view, Table 5 shows a comparison of current market segments and the expected market segments for new platform design strategies with respect to vision features. Table 5: Comparison to Current Market Segments and the Expected Market Segments Platform N73 N76 N78-1 N79-1 strategy Current No Mild Mild, S1 No, Mild Mild Mild, S2 No, Mild Mild Mild S3 S4 No, Mild, No, Mild, Mild Mild, Mild, Mild, 4.3 Phase 3: Identify Design Quality Based on the platform design strategies, the expected strategy product qualities can be calculated. The Features- Universal Principles Matrix is used to determine the functional accessibility level for vision features as shown in Table 6. The degree of impairment (a) for accessibility by a user with reduced vision functioning and the weight of universal principles (!) for better allowing users with reduced vision to use the product were determined. These are shown in Table 6. Here, we assume that the values of the usability levels for the products are 5. Table 6: Functions-Universal Principles Matrix for Vision Features Table 7 shows the expected strategy qualities of the products with respect to vision features. The expected strategy qualities of products for the platform strategies are calculated by Equation (8) as mentioned in Section 3.3. For example, the expected strategy quality of N73 in S1 is 1135, because the functions of N73 consist of F1, F2, F3, F4, F6, F7, and F9 and their quality values are 50, 12, 50, 40, 15, 27, and 33, respectively. We performed normalization of the value of the expected strategy quality for a product to compare current quality with strategy qualities as shown in Table 7. Next, we will apply the proposed coalitional game to develop a universal product family architecture. Table 7: The Expected Strategy Qualities of the Products (Normalization) Strategy N73 N76 N78-1 N79-1 Current 1000 (1.000) 1020 (1.000) 1270 (1.000) 1135 (1.000) S (1.135) 1270 (1.245) 1270 (1.000) 1270 (1.119) S (1.135) 1270 (1.245) 1270 (1.000) 1135 (1.000) S (1.270) 1020 (1.000) 1270 (1.000) 1270 (1.119) S (1.270) 1270 (1.245) 1270 (1.000) 1270 (1.119) 4.4. Phase 4: Determine a Platform Strategy For the case study, we assume that the expected demands for the products are determined by the result of market analysis as shown in Figure 7 and the total demands for the cell phone products are 100,000. For example, the market demands of Low Price and No Impairment are 50% and 50%, respectively. And then, the amount of demands for the product is 25,000. We also assume that the product cost of each product is 80% of the market price in Figure 6. We consider the current price of the product as the coefficient of the price to obtain a new product price along with the normalized strategy quality. Revenue for each product family based on platform strategies and the expected strategy qualities can be calculated by Equation (10) as shown in Table 8. To determine a platform strategy, the proposed coalitional game was applied to obtain the marginal contributions of vision features. 8 Copyright 2009 by ASME

9 Figure 7: The Expected Market Demands for Market Segments Table 8: Revenue of Current and the Proposed Product Families (Unit: ) Strategy N73 N76 N78-1 N79-1 Additional Cost Total Current 1,750,000 1,200, , ,000 4,302,800 S1 3,810,625 1,335, ,800 2,392, ,000C a 7,990, ,000C a S2 3,810,625 1,335, , , ,500C a 6,498, ,000C a S3 6,991, , ,800 2,392, ,000C a S4 6,168,750 2,225, ,800 2,392, ,000C a 10,436, ,000C a 11,239, ,000C a The game between three varietal modules for platform design of this product family is defined as the proposed coalitional game that is described in Section 3.4. Table 9 summarizes the coalitional game for determining vision features with three players. To determine marginal contributions for each varietal module, the coalitional benefits of the design strategies were calculated by Equation (11) as shown in Table 10. Since there is no benefit in a single module design strategy according to the definition of a coalitional game, we defined four collaborations as the combination of three varietal modules for design strategies. Therefore, the payoff vector of the game is (0, 0, 0, 0, C a, C a, C a, C a ). Table 9: The Proposed Coalitional Game for Platform Design Game Functions in vision features Players (N) F1, F7, F8 Coalition (G) G1("), G2(F1), G3(F7), G4(F8), G5(F1, F7), G6(F1, F8), G7(F7, F8), G8(F1, F7, F8) Table 10: Coalitional Benefit of Design Strategies Coalition Design Strategy Benefit () G5 F1, F7 3,688, ,000C a G6 F1, F8 2,195, ,500C a G7 F7, F8 6,133, ,000C a G8 F1, F7, F8 6,936, ,000C a To determine the marginal contribution of each varietal module, we used the Shapley value as mentioned in Section 3.4. The Shapley values of the varietal modules (F1, F7, F8) are ( C a, C a, C a ). Based on the marginal contributions of the varietal modules, we can decide a platform strategy for a universal family according to company s market strategy and market situations. Figure 8 shows marginal benefits based on various additional design cost with respect to vision features. Since the contributions of F1, F7, and F8 are depended on the redesign cost, C a, additional vision features for a new platform can be selected by design constraints. For example, if C a is less than 10.85, we can consider the Adjustable Font Style Function as the first candidate that will be included in a new platform to increase vision features based on common components. Figure 8: Marginal Benefits of Vision Features with respect to Redesign Cost Through the case study, we demonstrated that the proposed coalitional game could be used to determine a platform strategy by selecting functions that provide more benefits with respect to vision features in product family design. This coalitional game framework can provide a quantitative method to facilitate universal product family design in dynamic market environments. 5. CLOSING REMARKS AND FUTURE WORK In this research we present new quantitative methods for universal design. These methods are based on approaching universal design as a field in which mass customization via product family design provides an approach that can better serve the needs of those with a disability. By extending concepts from product family design to universal design, we have introduced a method for developing a universal product family through a game theoretic approach in a dynamic market environment. Modular product architecture is used to allow a range of trade-offs in determining the specific configuration for a platform at a conceptual design phase. Additionally, modular product architecture may be used to better leverage accessible and universal modules across multiple economies of scale. To evaluate and measure accessibility of a product, we introduced a strategy quality function using a well-established and thorough representation and rating method from the ICF. We considered a platform level selection problem as a strategic module selection problem under collaboration. A coalitional game is used to model module sharing situations 9 Copyright 2009 by ASME

10 regarding market segments and decides which modules provide more benefit when included in the platform based on the marginal contribution of each module. Via a case study, we have applied the proposed method to determine a platform for a family of mobile phones. Though the case study was analytic, we demonstrated that the proposed method can be used to determine appropriate features for a platform within different market segmentation strategies. Thus, the method can facilitate universal product family design in dynamic market environments. To improve the proposed method, we need to develop a method to better identify universal, accessible, and typical modules upon which a product family can be built. Additionally, since the product cost and the expected strategy cost are sensitive to players payoffs in a game, future research efforts will be focused on improving product cost models. Also, the proposed method will be compared to other decisionmaking methods for determining a design strategy in a universal product family. REFERENCES [1] Beecher, V. and Paquet, V., "Survey instrument for the universal design of consumer products," Applied Ergonomics, vol. 36, no. 3, pp , [2] Smith-Jackson, T. L., Nussbaum, M. A., and Mooney, A. M., "Accessible cell phone design: development and application of a needs analysis framework," Disability and Rehabilitation, vol. 25, no. 10, pp , [3] Field, M. J. and Jette, A. M., The Future of Disability in America. Washington DC: The National Academies Press, [4] Mace, R., Universal Design: Barrier Free Environments for Everyone. Los Angeles, CA: Designers West, [5] Preiser, W. F. E. and Ostroff, E., Universal Design Handbook. United States: McGraw-Hill Inc, [6] Silveria, G. D., Borenstein, D., and Fogliatto, F. S., "Mass Customization: Literature review and research directions," International Journal of Production Economics, vol. 72, no. 1, pp. 1-13, [7] Simpson, T. W., "Product Platform Design and Customization: Status and Promise," Artificial Intelligence for Engineering Design, Analysis, and Manufacturing, vol. 18, no. 1, pp. 3-20, [8] Shooter, S. B., Simpson, T. W., Kumara, S. R. T., Stone, R. B., and Terpenny, J. P., "Toward an Information Management Infrastructure for Product Family Planning and Platform Customization," International Journal of Mass Customization, vol. 1, no. 1, pp , [9] Osborne, M. J. and Rubinstein, A., A Course in Game Theory. Massachusetts, MA: MIT, [10] Connell, B. R., Jones, M., Mace, R., Mueller, J., Mullick, A., Ostroff, E., Sanford, J., Steinfeld, E., Story, M., and anderheiden, G., Center for Universal Design. Raleigh, North Carolina: North Carolina State University, [11] Vanderheiden, G., Design for people with Functional Limitations Due to Disability, Aging or Circumstances, in GavrielSalvendy (ed.), Handbook of Human Factors and Ergonomics: John Wiley & Sons, [12] Danford, G. S., "Universal Design People with Vision, Hearing, and Mobility Impairments Evaluate a Model Building," Generations, vol. 27, no. 1, pp , [13] Feathers, D., "Digital Human Modeling and Measurement Considerations for Wheeled Mobility Device Users," SAE Transactions, vol. 113, no. 1, pp , [14] Langdon, P. M., Clarkson, P. J., and Robinson, P., Designing Inclusive Futures: Springer Verlag, [15] Clarkson, J. P., Coleman, R., Keates, S., and Lebbon, C., Inclusive Design: Design for the Whole Population: Springer, [16] Bowe, F. G., Universal Design in Education: Bergin and Gavey, [17] Simpson, T. W., Siddique, Z., and Jiao, J., Product Platform and Product Family Design: Methods and Applications. New York, YN: Springer, [18] Simpson, T. W., Maier, J. R. A., and Mistree, F., "Product platform design: method and application," Research in Engineering Design, vol. 13, no. 1, pp. 2-22, [19] Gonzalez-Zugasti, J. P., Otto, K. N., and Baker, J. D., "A Method for Architecting Product Platforms," Research in Engineering Design, vol. 12, no. 2, pp , [20] Siddique, Z. and Rosen, D. W., "Product Family Configuration Reasoning using Discrete Design Spaces," presented at ASME Design Engineering Technical Conferences Proceedings, Baltimore, MD, ASME, Paper No. DETC00/DTM-14666, [21] Rai, R. and Allada, V., "Modular product family design: agent-based Pareto-optimization and quality loss function-based post-optimal analysis," International Journal of Production Research, vol. 41, no. 17, pp , [22] Johannesson, H. and Claesson, A., "Systematic Product Platform Design: A Combined Function-Means and Parametric Modeling Approach," Journal of Engineering Design, vol. 16, no. 1, pp , [23] Thevenot, H. J., Alizon, F., Simpson, T. W., and Shooter, S. B., "An Index-based Method to Manage the Tradeoff between Diversity and Commonality during Product Family Design," Concurrent Engineering: Research and Applications, vol. 15, no. 2, pp , [24] Moon, S. K., Park, J., Simpson, T. W., and Kumara, S. R. T., "A Dynamic Multi-Agent System Based on a Negotiation Mechanism for Product Family Design," IEEE Transactions on Automation Science and Engineering, vol. 5, no. 2, pp , [25] Zacharias, N. A. and Yassine, A. A., "Optimal Platform Investment for Product Family Design," Journal of Intelligent Manufacturing, vol. 19, no. 2, pp , [26] Meyer, M. H. and Lehnerd, A. P., The Power of Product Platforms: Building Value and Cost Leadership. New York, NY: The Free Press, [27] Magrab, E. B., Integrated Product and Process Design and Development: The Product Realization Process. Boca Raton, NY: CRC Press, [28] Demirbilek, O. and Demirkan, H., "Universal product design involving elderly users: a participatory design model," Applied Ergonomics, vol. 35, no. 4, pp , [29] WHO, International Classification of Functioning, Disability and Health: ICF Short version. Geneva: World Health Organization, [30] Story, M. F., "Assessing Usability: The Principles of Universal Design," Assistive Technology, vol. 10, no. 1, pp. 4-12, [31] Dyer, R. F. and Forman, E. H., "Group decision support with the Analytic Hierarchy Process," Decision Support System, vol. 8, no. 2, pp , [32] Shapley, L. S., "Cores of Convex Games," International Journal of Game Theory, vol. 1, no. 1, pp , [33] Nokia, accessed December, [34] Holtta-Otto, K. and de Weck, O., "Degree of Modularity in Engineering Systems and Products with Technical and Business Constraints," Concurrent Engineering: Research and Applications, vol. 15, no. 2, pp , Copyright 2009 by ASME