A Framework Of Virtual Design For Product Customization - IEEE Xplore

A Framework of Virtual Design for Product Customization Mitchell M. Tseng, Jianxin Jiao, Chuan-Jun Su Department of Industrial Engineering & Engineering Management The Hong Kong University of Science & Technology Clear Water Bay, Kowloon Hong Kong thus engenders design variations and process changeovers, which seemingly contradicts the pursuit of low costs as in mass production. Such a setup therefore presents product development with a special challenge: it is vital to provide designers with feedback from production, quality, and tests early at the conceptual stage so as to maintain the integrity of the product family and the continuity of the infrastructure, hence leveraging existing design and manufacturing investments. This effectuates a strategy to achieve the economies of scale for mass customization by maximizing common denominators in product realization

Abstract - One major challenge of mass customization is how

to meet two seemingly conflicting goals, namely satisfying individual customers’ needs while keeping mass production efficiency. In order to help customers to select the appropriate product family so that the closest variant can be quickly generated and delivered, this paper proposes a virtual design approach by developing virtual prototypes to represent the product platform. With virtual prototypes, assessments of constant delivery can be made to facilitate the selection and negotiation process between sales and customers, thus enabling better informed customization design. The primary goal of virtual design for mass customization is to provide a multidisciplinary design definition at the product family level, and the closest product variant can be rapidly generated. Furthermore, a scenario-based simulation environment for design trade-off can be done in an integrated framework. Such a design environment facilitates capturing customer needs and appropriately informing them of the capability of the firm. Ultimately, through virtual design, customers can directly interact with the CAD systems to make trade-offs among factors that are important to them. These factors can be performance, aesthetics, value, cost, urgency, and others. This paper discusses the technical issues that need to be addressed in realizing virtual design and an example is presented to illustrate the proposed approach.

r31.

As one approach to these problems, concurrent engineering (CE) is well recognized for its natural focus on product design [4]. CE calls for the consideration and inclusion of product life cycle concerns such as aesthetics, ergonomics, marketability, manufacturability, etc. in the product design process. In CE implementation, the link between designs, represented as geometric information and manufacturing instructions, has been a major obstacle to CAD/CAM integration [ 5 ] . A number of techniques have been developed to bridge this gap, including feature-based design and feature extraction approaches. With the particular focus on keeping the economies of scale, this paper adopts an alternative approach, called Design by Manufacturing Simulation (DMS) [30]. The concept that manufacturing simulation could be used as a design tool was first introduced by Gossard [6]. In his approach, through designing parts by simulating manufacturing operations on the screen, the designers generate the manufacturing specification as they design. A similar approach to Gossard’s was used in the robotics world [7] to generate robot path. This approach demonstrated the concept of path generation through interactive graphics. Another related approach is virtual prototyping (VP) [8]. It is proposed to replace hardware prototypes with computational (virtual) prototypes of systems and the processes that they may undergo. By replacing hardware with computational prototypes, the potential is tremendous for greatly reducing product development time, manufacturing facility ramp-up time, and product development cost. On the other hand, a number of marketing studies [18, 19, 201 have pointed out the significance of understanding customer preferences regarding the appearance of new products. An attractive appearance draws customers to a product and adds value to the product by increasing the quality of the user’s experiences [21]. Therefore, identifying those elements that enhance the chances of customer acceptance represents an important issue for

I. INTRODUCTION In an age where consumers demand high-quality, lowpriced and customized products, the competition among firms has ceased to be strictly a price competition and is now a competition in product variety and speed to market. The current philosophy is to replace old products constantly with new versions of either an improved product or a new variation of the product. Differentiation in product variety, i.e., customization, has assumed an ever increasing importance as a marketing instrument. The duration of a product’s life depends on its acceptance by the consumers; a “failed” product could be out of the market in a matter of months. A short product development cycle is crucial to the survival of the company as it enables the company to deliver new products to the market quickly. On the contrary, alongside pursuing flexibility and quick response, manufacturers have to pursue a “dynamic stability” [ 11. That is they must keep mass efficiency to obtain the economies of scale, an advantage characterized by mass production. This oxymoron manifests the new production paradigm termed as mass customization [2]. Customization emphasizes the difference among or the uniqueness of the products. This product proliferation naturally results in the continuous accretion of varieties and 0-7803-4192-9/97/$10.00 0 1997 IEEE

7

engineering designers. With this in mind, effective use of customer preference data in engineering design helps integrate the perspectives of marketing professionals and designers. Accordingly, customization designs should emphasize the employment of customer preference data early in the design process. In terms of a product’s appearance, customer reactions appear to depend in part on how well the design conforms to aesthetic principles such as those developed by Gestalt theorists [22] and ergonomics criteria. Preferences also may be affected by how well a new design fits within the constellation of existing designs. Therefore, the ease with which customers may categorize a new design and its closeness to existing prototypes may play a significant role in marketplace acceptance [23]. Although Gestalt and prototypicality principles may apply widely across customers, there are also expected salient individual differences (i.e., customization) in product appearance preferences. As a result, it is important for design teams to explore the customers’ perceptions on the appearances of the target products. Traditionally, market analysis techniques are employed to investigate customer response to design features and price sensitivity. For example, conjoint analysis is widely used to measure preferences for different product profiles and to build market simulation models [24]. Related work by Louviere et al. [25] used discrete choice experiments to predict customer choices pertaining to design options. In a new approach, Turksen and Willision [26] employed fuzzy systems to interpret the linguistic meaning regarding customer preferences as an alternative to conjoint analysis. Researchers at IBM have used structured brainstorming to build customer requirements into the process of quality function deployment [27]. Others have taken a qualitative approach and used focus groups to provide a reality check on the usefulness of a new product design [28]. However, although a number of researchers agree on the importance of including customer preferences and other marketplace information in product designs, methodologies to capture product preferences and tastes are not evident in a concurrent engineering context [22], particularly in terms of the premises of mass customization. Our approach to mass customization is to combine the ideas of virtual prototyping with design by manufacturing simulation techniques. Our domain of VP consists more of manufacturing simulation than the modeling of interacting physical processes, which is the focus of most other VP investigations. That is, by constructing several virtual prototypes, accurate assessments of mass producibility will enable better informed customization designs. The primary goal of virtual design for mass customization is to provide a multidisciplinary design definition and rapid prototyping environment for concept development and a tailored, scenario-based simulation environment for concept evaluation within a single facility. This design environment facilitates the capture and utilization of information generated during the design phase, and the simultaneous generation, at design time, of manufacturing, materials,

costing, and scheduling data, to support the implementation of CE for mass customization, along with the visual evaluation of customer’s perception on the appearance of the target product. In the next section, the concept, feasibility and merits of VP are discussed with respect to the premises of customized product development, along with the key techniques associated with virtual design in the context of mass customization. Section I11 presents the considerations on the system architecture and configuration of virtual design environment (VDE) for mass customization. The idea and rationale of VDE, along with the integration of W,design iteration and engineering analysis are discussed in detail, as well as the interfaces between the virtual world and the physical world, and the interaction among designers, customers, and engineering database. An implementation example is presented in section IV. Finally, discussions and conclusions are drawn in section V.

11. VIRTUAL DESIGN It is important to effectively capture customer requirements for product success. In reality, customer and sales interaction is a two way street. Good products are resulted from the best match of the interests of the customers and those of the firms. Thus, it is equally important that customers are fully aware of the capability of the corporation including manufacturability, serviceability and other abilities that are often defined by knowledge, skill, equipment and other resources of a firm. The approach presented in this paper is to represent the capability of the firm in a set of Product Family Architecture (PFA) [3]. The architecture is further embodied in software prototypes. These prototypes are then used to facilitate the interaction between customers and the firm. It enables customers to “design” the products according to the best balance between their preference and the performance offered to them. The customers can exert their judgment based on implicit or explicit criteria, in the mean time the capability of the firm is best presented for economies. Since customers are conducting some form of mapping from functional requirements to high level design parameters [31], we refer this process to virtual design. A. Virtual Prototyping During product development, physical prototypes are frequently required for iterative evaluation to provide feedback for design modification such as to prove design alternatives, engineering analysis, manufacturing planning and often just to present a product. Even using conventional processes and highly skilled technicians, the time, effort and cost of constructing a prototype has been quite high [8]. Rapid Prototyping (Rp) systems are capable of making highly accurate prototypes in a very short time. The starting point for such systems is good quality 3D CAD where solid models are constructed and then post-processed in a layer format, using, for instance, stereolithography, to make them suitable for the prototyping machines [9].

The improvements from using rapid prototyping have been said to be found in design visualization, product functionality verification, iterative development, and testing for optimization [9]. There seems then some obvious overlap with the potential attributes of virtual reality (VR) [lo, 11, 121 for product design: intuitively VR could be used interactively with rapid prototyping technologies to enable even more effective use of RP and to planning better through very early visualization and fit and functionality testing. Gibson et al. [8] reported VR+RP had something different to offer and the notion of using “digital clay” allowed fuller and faster exploration of functional, aesthetic, and ergonomics design criteria. Considering RP still involves physical construction of prototypes that are costly, time consuming, and with limited “prototyping” usage, electronic prototyping based on product model data is an alternative to physical prototyping technologies in terms of the coherence of design and engineering analysis [SI.Recent development in computer graphics and computer simulation have now provided more sophisticated tools for electronic prototyping [121. Specifically, VR techniques offer the Possibility to experience virtual worlds with VerY high milism. Applying VR to electronic prototyping will greatly improve the quality of presentation. This intuitive approach to object manipulation opens new opportunities for prototyping based on CAD models. It is possible to take the data model of a product as a virtual prototype instead of a real one to model and analyze geometry, and manufacturability of the designing products interactively [13]. The virtue of VP lies not only in the reduction of the fabrication of physical therefore, shortening product development timeprototypes, and cost, but also in supporting

I“‘

Design World

,...

-

Recess -.-.... .-.....

Virtual World

World

Fig. 1 The principle of virtual design

environments, analytical virtual prototypes will not be used efficiently within the product development process until geometric representations are efficiently integrated with to be a significant analysis applications, which problem 114,161.

c. ~

~

~

~

~

Conventional design starts with customer needs identification, where designers play a major role in transfodng these needs into design though concurrent requirements engineering have proposed approaches to make suchthe as functional [3 11. been Various

product design and presentation by qualitative simulation and analysis, hence facilitating the discussion, manipulation and modification on the product data model directly among personnel with different technical backgrounds, which in the end facilitates CE.

transformation more efficient and effective. Virtual design advocates incorporatingcustomers in the center of product realization. With the assistance of a digital mockup (referred to as virtual prototype in the following sections), customers can interact directly with the capabilities and constraints of a manufacturer. Informed customers can then make trade-off in terms of cost, features, aesthetic appearance and delivery schedule, and so on. In order to realize such an approach, a virtual design environment is essential to provide the supporting infrastructure, which is based on the integration of computer supported modeling, simulation, and animation of

B. Types of Prototypes There are basically two types of v b ” prototypes, i.e., the immersive virtual prototype and the analytical virtual prototype. Recent Progress in the development of graphics hardware has allowed complex geometric representations to be rendered and in time* representations, when coupled with new hUman computer interfaces such as datagloves and headsets can help give the user a belief that the object actually exists. The Visual effects and tactile properties are of primary importance in these immersive virtual prototypes, which are in visualizing and interacting with the digital clay. A more useful fC”l of a virtual prototype, in the context of product development, is one that tells the user how it Will Perform and behave in its intended environment. This analytical virtual Prototype usuallY uses computing technology (mouse, keyboard and screen). It is thought that eventually immersive technology will be used. Leaving aside the problems of moving toward immersive

design alternatives, along with performance-related data. ~ i 1 ~shows . the virtual design process which starls from the design world where customers can freely express their needs and make decisions by balancing their budget, schedule and requirements. On the other hand, traditional designers switch their roles in the virtual world where their primary responsibility is to construct product technologies (not in the form of products). The interaction between the design wofld and the virtual world a virtual design environment where product definition, design engineering and manufacturing are integrated together.

D. K~~ Techniques Associated 9

virtual

~~~i~~

~

1 ) Product Representation and Model Generation: The Customer Needs Elicitation comprehensive modeling capability of products and the related engineering activities is essential for virtual design, Feedback from Customer Perception Task Clarification which integrates all product information and bridges different stages of product development. Underlying design Design by MCg. Simulation and simulation capabilities is the product representation. Process Design Product representation must evolve along the design timeComputer Simulauon Embodunent Design line, support design including decision making and concept Engineenng Analysis exploration, and be capable of readily generating For Life Cycle Concerns (CE) simulation models at an appropriate level of detail. Product modeling plays an essential role in product representation and simulation model generation. A product model is a generic model used for representing all types of artifacts that appear in the course of design and manufacturing. It represents target products, their materials and intermediate products, tools and machines, and any other manufacturing resources and production issues [14]. In addition, in order Fig. 2 Virtual design for mass customization to support downstream manufacturing activities, product representation must be converted from a central (design) processes that are required for representing product representation to a representation suitable for reasoning in a behavior and manufacturing processes. Activity models target activity. Computationally efficient methods for represent various engineering activities, whether human or recognizing features from solid models, i.e., conversion, by computer, for product engineering and production have been studied extensively [5, 15, 161. By following management, such as costing and scheduling. Some of the similar methods to feature-based component related works are given in [14, 15, 161. representations, converters can be developed for assembly 4 ) Product Library: Customization emphasizes the representations, specifically for generating VP models for uniqueness and variety of the products that inherently engineering analysis. engenders design variations and manufacturing 2) Human-Computer Interaction: A virtual prototype changeovers. To maximize mass producibility, it is means not only the product data model itself, but also important to reuse previous designs to generate new includes the potential customers’ imagination upon viewing products so as to maintain the integrity of the product and manipulating this model. It is essential that the human families and the continuity of the infrastructure, hence senses are provided with information that represents the leveraging the existing design and manufacturing data model as realistically as possible [13]. For this purpose investments. Therefore, a product family architecture (PFA) VR techniques are applied where graphical visualization of [3, 41 is called for to characterize the product platform. In and interaction with the virtual prototype are required. VR- essence, a PFA means the underlying architecture of a based presentations make all relevant data visible. The product platform, within which various product variants can visualization is based on the VR data model, which be derived from the basic product designs to satisfy a contains the actual geometry and lighting information spectrum of customer needs related to various market generated by the modeling and simulation tools. The niches. In other words, a good PFA provides a generic interaction interprets user actions and either changes the architecture to capture and utilize commonality, within viewing parameters or generates logical events for which each new product instantiates and extends so as to simulation and manipulation. The visual evaluation of a anchor future designs to a common product line structure. prototype by potential customers can contribute to the In the context of mass customization, the rationale of a PFA entire product development process, particularly, in cases lies in not only unburdening the knowledge base from of consumer products that feature aesthetic and ergonomics keeping variant forms of the same solution, but also design criteria. To explore customer response to a product’s modeling the design process of a class of products that can

l I

appearance, the rendering techniques are required to

provide a realistic product appearance, which involves the application of lighting and resulting shading. 3) Manufacturing Simulation: Design analyses are conducted based on prototype evaluation. Simulations make available the prediction information about the product behavior, manufacturing processes, and production planning, which enables prototype evaluation. As the kernel of computer-aided manufacturing simulation, in addition to product modeling, analytical virtual prototypes must involve process modeling and activity modeling [ 141. Process models are used for representing all the physical

widely variegate designs based on individual customization

requirements within a coherent framework [ 171. In addition, the functional view of a PFA performs as a taxonomy of design requirements for eliciting customer needs, which enhances customer recognition and the interaction between customers and engineers [29].

Et.VIRTUAL DESIGN ENVIRONMENT As shown in Fig. 2, the primary goal of virtual design for mass customization is to provide a multidisciplinary design definition and rapid prototyping environment for concept

development and a tailored, scenario-based simulation environment for concept evaluation within a single facility, namely, a virtual design environment (VDE). VP assists design for mass customization in two aspects, design by manufacturing simulation (DMS) and visual evaluation. As suggested in Section I, DMS is an effective approach to mass producibility. A well-established product model and the related process model not only facilitate the capture and utilization of information generated during the design phase, but also perform as the analytical virtual prototype for engineering analysis based on computer simulation. Therefore, product development life cycle concerns, such as manufacturing, assembly, production management and the economical scale, can be addressed at an early stage. On the other hand, the virtual prototype enables the visual presentation of the digital clay to potential customers by rendering the look and feel of the target product. This helps designers elaborate designs based on the customers’ perception and involvement, which is particularly useful for customized consumer products featuring aesthetic and ergonomics requirements. In such a way VDE enables the design, evaluation, and improvement of products by a CE team. In VDE, the sequential development process is significantly improved by bridging different stages of product development via virtual prototypes, and enabling the customer-in-loop through applying VR techniques. Fig. 3 illustrates a configuration of VDE, which is composed of the following subsystems. A. Visualization System It is well known that present 3D CAD systems have efficient modeling functions such as the parameteric and feature-based functions. The problem of displaying CAD geometry in a VR environment relates to the large number of polygons in the complex geometry of a part. Typically, engineering workstations and PCs do not have graphic engines. Therefore it takes a great deal of time to rotate 3D

models and to shade models and so on. Special VR hardware, for example a Silicon Graphics workstation, is necessary for a visualization system, as well as algorithms to render these complex parts in a virtual world while maintaining a reasonable frame rate. As for difficulty in reading CAD-dependent data directly, it is feasible to exchange 3D CAD data to IGES (Initial Graphics Exchange Specifications) files. Then the system reads IGES files and creates the original 3D model. The visual presentation can probe potential customers’ perceptions on the target products being designed for design elaboration. Furthermore, it is possible to provide immersion in and interaction with the virtual world through real-time simulation. In VDE, the visual evaluation of VP actually performs as a market survey tool to provide specific data about exact product appearance not available through other market research techniques such as conjoint analysis.

B. Simulation System Both the product design and the choice of the manufacturing process affects the assessment of mass producibility required by mass customization. The underlying issue is the cost of manufacturing. The purpose of the simulation system is to integrate prototypes and manufacturing activities directly into design decision making by defining appropriate product, process and activity models. Thus, results from a simulation are fed directly into a product design decision. Given a virtual prototype, CAD models provide the basis for a process plan, which is necessary before simulation. The process plan is developed based on CAD models and machine constraints and performs as the basis for manufacturing simulation. The assembly line will be laid out in terms of stations and operations, whether manual or automated, then the product will run down the line to simulate the process. The throughput, utilization, buffer sizes, and other simulation results can be determined. Virtual humans could

VisualizationSystem

% \\ Culllamer Perception

\

Visual Evaluation

Tnwrslon

Interaction

\ ‘J

Virtual Design System

SimulationSystem

Process Modellop.

4--+

AnemMy Planning

- Assembly System

Actidly

Modeling ~




Fig. 3 Configuration of virtual design environment (VDE) 11

Simvlaoon Model

b

be animated to simulate manual operators. In manufacturing simulation, each time a machine tool is actuated, the time and cost of that movement can be calculated from company standards. Similarly the material is selected and indexed into the simulated machine so the material costs and requirements can be calculated. Similar calculations can be performed for assembly simulation. All these partial cost, time and requirement data is assembled during the design process to give an indication of product cost and lead-time.

0 WorldTOolKit

CAD System

*

Virtual Design System

VR System (Vaualizauon)

Fig. 4 An implementationexample of VDE

C.Virtual Design System Design problem definitions and solutions will be developed in a virtual design system. That is, it should be possible to formulate and solve selection and compromise design specifications (DSPs) for product design. It is through these DSPs that information produced during prototyping, simulation, and evaluation is fed back to the designer to aid in decision making. For example, process plans and assembly plans are the contents of a DSP. Another important task of a virtual design system is the prototyping where virtual prototypes are interactively generated based on CAD models. In prototyping, the manufacturing process model, assembly model and scheduling model are prepared around the product data model. By prototyping, a virtual prototype is developed as an information model of a product with models of its manufacturing, assembly and production management, ready to be manipulated by a designer via computer simulation to instantiate such manufacturing information.

Fig. 5 The PFA for various customized souvenirs

D. Engineering Database

As the core of VDE, the engineering database system (ENDB) mainly manages 3D CAD data and product library, and has interfaces to other systems such as the master database and simulation systems. The retrieval of pertinent design related information is crucial to the customercentered approach. To this end, design data should be presented with the underpinning product family architecture (PFA) as the taxonomy. In such a way, an ENDB facilitates the customer’s ability to retrieve the products that are closest to their needs so that the variants from the PFA can be derived. The requirements for an ENDB involve (1) management of current CAD data including part data and configuration tree, (2) maintenance of the product library with a product family architecture built-in, (3) management of the engineering process, e.g. work flow, (4)interfaces to other systems, such as CAD/CAM/CAE systems, and ( 5 ) information retrieval functions, such as parts, assembly and DSPs. IV. AN IMPLEMENTATION EXAMPLE The design of a souvenir (a sundial model) is selected as the test-bed in our preliminary implementation of VDE. Fig. 4 shows the current configuration. AutoCAD is deployed for preparing 2D drawings. DXF data format is used to build interface from AutoCAD to the virtual design

Fig. 6 Visual presentation of a virtual prototype

system, where virtual objects are built for rendering and simulation. All the objects are built by first converting 2D geometry to 3D solid models, and then applying physical properties, such as color and texture. This is done through 3D Studio 4.0 for PC. All these object models are then converted to WorldToolKit 2.1 for SGI Onyx format (NFF) in order to be visualized in VR system. The visualization system is developed using Sense8’s WorldToolKit for PC and Unix as principal developing tool. The CrystalEyes glasses and Emitter are employed to obtain stereoscopic visual effects. In addition to the geometric representation, the underlying product model involving the visualization of prototypes relates to an appropriate product family architecture (PFA). The PFA comprises a product library with various types of components and the configuration tree to synthesize the end products. Product differentiation, i.e., customization and varieties, is captured and characterized by the PFA. Fig. 5 illustrates the PFA for the souvenir example, which manifests the diverse configurations of different customized sundials. The aesthetic visualization of the digital clay helps to explore the customer’s perception on the target product, as shown in Fig. 6. ProModel is an external simulation tool. Using ProModel, objects ranging from individual products to

complex factory scenarios can be modeled with functional properties and motion data. Process plans can be developed according to CAD drawings and machine constraints. In the souvenir example, the CNC machine is selected as the main machine tool. The rational consideration is that customized products are often characterized by large varieties with small lot size, which requires flexible manufacturing facilities. For each component of the souvenir, CNC programs are generated from CAD drawings using software MasterCAM, from which the machining process can be simulated to estimate machining time and cost. The assembly line can be designed in terms of stations, operations and tools according to the product model supported around a bill-of-material (BOM) type structure (Fig. 7). Fig. 8 gives the configuration of a manual assembly system designed in a virtual environment for the souvenir assembly. The manual assembly process can be simulated according to the virtual prototype, through which the appropriate assembly operations and sequences can be designed and analyzed. Fig. 9 illustrates the use of input devices (mouse or datagloves) to manipulate parts so as to imitate manual assembly operations. The performance of different assembly plans is evaluated using FYoModel simulation in conjunction with the virtual assembly model. Fig. 10 gives an example of assembly planning simulation in ProModel, in which the statistics suggest scheme 2 as the optimal one. According to the virtual prototype, the time and cost associated with manufacturing and assembly can be statistically summarized based on simulation. Adding data on labor rate and overhead, the manufacturing and assembly cost can thus be estimated. According to the process and assembly planning coherent with the virtual prototype, the material cost and requirements can be calculated, along with the tool costs and requirements. Finally, the overall cost and lead-time are estimated and scheduled.

LJ Souvenir

, I

Base with Label

Fig. 7 BOM-type product structure model for the sundial

Fig. 8 Assembly system design for souvenir assembly

Fig. 9 Virtual assembly of the souvenir (sundial)

V. DISCUSSION AND CONCLUSIONS In this paper, we have presented the idea of combining virtual prototyping with design by manufacturing techniques for the development of a virtual design environment. Our particular application is customized product development which is facing the challenge of keeping mass producibility. A central idea is by constructing virtual prototypes, accurate assessments of mass producibility will enable better informed customization designs. The primary goal of virtual design for mass customization is to provide a multidisciplinary design definition and rapid prototyping environment for concept development and a tailored, scenario-based simulation environment for concept evaluation within a single facility. We outlined the research issues associated with the development of such a environment including product representation and prototype generation, humancomputer interaction, manufacturing process simulation, and product library maintenance.

Fig. 10 Simulation on assembly plans for manual assembly system

Based on our work, we have these observations: (1) the combination of virtual prototyping and VR techniques has tremendous potential in providing advanced visualization and manipulation capabilities for customized product design by enabling visual evaluation and acquiring 13

H. Grabowski, R. Anderl and M.J. Pratt, Advanced Modeling for CAD/CAM Systems, Springer-Verlag,

customer’s perception on the target product; (2) the development of a virtual design environment shows great promise in providing intuitive guidelines for people engaged in design and (virtual) prototyping; (3) such an environment provides a unifying framework for bridging different stages of product development by facilitating the capture and utilization of information generated during the design phase, and the simultaneous generation, at design time, of manufacturing and production planning data; (4) product family architecture plays an important role in virtual design for mass customization by enabling previous designs and production plans to be mapped to a new design, potentially maintaining the integrity of the product family and the continuity of infrastructure, hence leveraging existing designs and manufacturinginvestment.

Berlin, 1991. T. Kjellberg and H. Schmekel, “Product modeling and information integrated engineering systems”, Annals of the CIRP, vol. 41, no. 1,1991, pp. 201-204. Mitchell M. Tseng and Jianxin Jiao, “Case-based Evolutionary Design for Mass Customization”, in Proceedings of the 21st International Conference on Computers & Industrial Engineering, Puerto Rico, USA, March, 1997. K. Beruand, “New product design: Shape of the future”, Business Marketing, 1991, pp. 18-19. Angela Schonberger, Raymond Lowery: Pioneer of American Industrial Design, Prestel, Munich; 1990. Marvin Berkowitz, “Product Shape as a Design Innovation Strategy”, Journal of Product Innovation Management, vol. 4, 1987, pp. 274-283. Philip Kotler and G. Alexander Rath, “Design: a powerful but neglected strategic tool”, Journal of Business Strategy, vol. 5 , Fall 1984, pp. 16-21. Robert W. Ji Veryzer, “Aesthetic response and the influence of design principles on product performance”, in Advances in Customer Research, vol. 20, Leigh McAllister and Michael Rothschele (eds.), Association for CustomerResearch, Provo, UT, 1993, pp. 224-231. Mita Sujan and Christine Dekleva, “Product categorization and inference making: some implications for comparative advertising”, Journal of Customer Research, vol. 14, 1987, pp. 372-378. G. Dobson and S . Kalish, “Heuristics for pricing and positioning a product line using conjoint and cost data”, Management Science, vol. 39, no.2, 1993, pp. 160-175. J. Louviere, D. Anderson, J.B. White and T.C. Eagle, “Predicting preferences for new product configurations: a high-tech example”, in Proceedings of the IFIP TC 7

VII. REFERENCES Andrew C. Boynton, “Beyond flexibility: building and managing the dynamically stable organization”, California Management Review, 1991, vol. 34, no. 1, Fall 1991, pp. 53-66. B.J. 11 Pine, Mass Customization: The New Frontier in Business Competition, Harvard Business School Press, 1993. Mitchell M. Tseng and Jianxin Jiao, “Design for Mass Customization”, Annals of the CIRP, vol. 45, no. 1, 1996, pp. 153-156. Mitchell M. Tseng and Jianxin Jiao, “Concurrent Engineering for Mass Customization”, in Proceedings of The 41h Intemational Conference on Computer Integrated Manufacturing, Singapore, October, 1997. B. Choi and M.M. Barash, “STOPP: an approach to CAD/CAM integration”, Computer-Aided Design, vol. 17, no. 4, 1985, pp. 162-168. D.C. Gossard, “Analogic part programming with interactive graphics”, Ph.D Dissertation, Massachusetts Institute of Technology,USA, 1975. C.D. Crane and J. Duffy, “An interactive animated display of man-controlled and autonomous robots”, in

Conference, Modeling the Innovation: Communications, Automation and Information Systems, Rome, Italy; 1990,

pp. 53-61. L.B. Turksen and I.A. Willson, “Customer preferences models: fuzzy theory approach”, in Proceedings of the

Proceedings of ASME Winter Annual Meeting and Symposium on Intelligent and Integrated Manufacturing Analysis and Synthesis, vol. 25,1986, pp. 1-10.

SPIE - International Society for Optical Engineering,

Boston, MA, 1993, pp. 203-211. Jani G. Byme and Todd Barlow, “Structured brainstorming: a method for collecting user requirements”, in Proceedings of the 37th Annual

I. Gibson, D.J. Brown and S.V. Cobb, “Virtual Reality and rapid prototyping”, in Virtual Reality in Engineering, IEE Computer series 20; 1993, pp.51-63. P. Jacobs, Rapid prototyping and manufacturing, Society of Manufacturing Engineers, The Netherlands: 10s Press; 1992. K. Pimental, Virtual reality: through the new looking glass, Intellwindcrest, New York; 1993, pp. 194-208. J.A. Adams, “Virtual Reality is for Real’’, IEEE Spectrum, ~01.30,no.10, 1993, pp. -29. L. Jacobson, ‘‘Virtual Reality: a status report”, A I Expert, vo1.6,no.8, 1991, pp.26-33. P. Astheimer and M. Gobel, “Virtual Design II: an advanced VR development environment”, in Proceedings of Eurographics Workshops, Monte Carlo, Monaco, 1995. F. Kimura, “Product and process modeling as a kemel for virtual manufacturing environment”, Annals of the CIRP, vol. 42, no. 1, 1993, pp.147-150.

Meeting of the Human Factors and Ergonomics Society,

Seattle, WA; 1993, pp. 427-431. Susan Lachance-Porter, “Impact of user focus groups on the design of new products”, in Proceedings of the 14th National On-line Meeting, Published by Learned Information, Medford, NJ; 1993, pp. 265-271. Mitchell M. Tseng and Jianxin Jiao, “A Variant Approach to Product Definition by Recognizing Functional Requirement Pattems ”, Computers & Industrial Engineering, September 1997, vo1.33. M.W.S. Jaques, “Design by Manufacturing Simulation”, Ph.D Thesis, 1993, University of Portsmouth, UK. N.P. Suh, The principles of design, Oxford series on advanced manufacturing; 1990. 14