The goal of this work is to provide industrial designers with an intuitive mechanism for transferring brand-identity elements from one object to another, and to allow manual modifications of physical prototype details to be scanned and reincorporated back into original digital models. An algorithm to copy-and-paste freeform details from one freeform surface to another using displacement field and planar parameterization operators is presented. A tool based on the algorithm is developed such that, from the designers viewpoint, details from one surface can be mapped to another in two steps: the selection of the detail to transfer (copy); and the choice of the surface onto which the detail is to be mapped (paste). A proof of concept has been implemented and is used to demonstrate the capabilities of the framework.
'), new pubdetails('05-caoq-asme-cie', 'full document', '297 kB', '05-caoq-asme-cie/dontindex/05-caoq.pdf', '', '', '', 'MIT CIPD', 'DOME, distributed design', 'DOME', 'New emerging modeling and simulation environments have the potential to provide easy access to design models and simulations over the Internet, much as the World Wide Web (WWW) has provided easy access to information. Parametric simulation models accessible through such Internet-based environments can be seen as providers of computational services. As Internet-based computational services become more common, an effective mechanism for identifying functionally appropriate computational services will be necessary to support distributed simulation-based design.
There are ongoing efforts to develop ontological descriptions of computational services, with the goal of explicitly defining the functional or semantic role of a computational service. However, much in the way web search engines operate, the goal of this work is to use patterns found within the syntactic information in interfaces for computational model services to identify functional roles. Typical, pre-existing syntactic information expected to be available¡ªsuch as model and parameters names, parameter units, and input/output structure¡ªare identified. Given a model with a known functional role, a Fuzzy Attributed Relational Graph representation of its syntactic interface information is used as a pattern or template for this known role. Then, using a graph-matching algorithm based on attribute similarity metrics, the interface information from other models can be classified against the functional template. New information from functionally equivalent model interfaces that have been found can be incorporated into the original template, thereby both generalizing the pattern for the functional role and strengthening the most critical aspects of the pattern. Test results for a prototype implementation of the proposed algorithm are provided.
'), new pubdetails('01-kraines-jie', '', '', '', '', '', '', 'Alliance for Global Sustainability', 'holistic design, environmentally-conscious design, DOME', 'DOME, ECD', 'The evaluation of tradeoffs between technologies and policies for mitigation of environmental problems requires a systematic investigation of effects over the entire region under consideration. When attempting to model such large complex systems usability, maintenance, and computing efficiency often become major modeling barriers. In this work a software prototype for integrating the services of computational models over the Internet, called DOME (distributed object-modeling environment) is used to facilitate the construction of "virtual Tokyo"-a simulation platform for evaluating holistically the tradeoffs between various technologies for reducing the emission of greenhouse gases. In making steps towards this ultimate goal, two models that encapsulate data defining spatial land-use distributions and the flows of goods (input-output analysis model) for an urban region have been developed. Integrated, these models form a preliminary virtual Tokyo model when applied to Tokyo specific databases. Given this platform, process models are applied to examine the effectiveness of using photo-voltaic (PV) cells on the demand side to reduce conventional electric power generation, and thereby also reduce emissions of CO2. Study results of introducing PV cells on the rooftops of buildings in Tokyo under various installation conditions are presented as a working example of the prototype. For full deployment on usable rooftop space, PV power generation could reduce CO2 emissions from electric power generation by almost 18%. Cost, a critical determinant in the actual feasibility of PV cell installation, will be examined using the same methods as presented here in future work.
'), new pubdetails('00-senin-asme', 'full document', '566', '00-senin-asme/dontindex/00-senin.pdf', '', '', '', '', 'DOME, object-based modeling, distributed design, integrated design, service marketplace', 'DOME', 'Predictive integrated system modeling is now a pressing issue in the design of complex products ranging from home air conditioners, to automobiles, and aircraft. While product development organizations have official top-down development processes, in practice individual participants perform their work in an informal marketplace, bartering service exchange relationships to get what they need to resolve their part of the problem. Ideally, these activities would also create a set of heterogeneous and distributed models representing the complete product which could then be used for the rapid exploration of design tradeoffs and global optimization. The authors envision a distributed simulation service marketplace running in parallel with the activities of participants. Participants autonomously form electronic service relationships with others, and in doing so drive underlying predictive simulations. Ultimately, networks of distributed service exchange relationships and underlying simulations emerge as a computational system to predict the integrated behavior of the product.
This paper develops an object model representation to enable this vision for a decentralized design simulation marketplace. Agent-like service-objects are provided to embody different types of engineering data and to allow the propagation of these data between object models over the Internet. Relation objects allow participants to define functional interactions between services locally within a given object model, and managers solve and coordinate the firing of relationships within local modeling scopes. Finally, there are mechanisms to create plug-in objects so participants can define service interfaces for models in third party software or custom code without additional programming, or to make the services of tools for decision support or optimization available in the marketplace. A prototype implementation, called DOME (Distributed Object-based Modeling Environment), is used to illustrate the concept on a beverage container design problem.
'), new pubdetails('00-sousa-jie', '', '', '', '', '', '', '', 'integrated design, surrogate modeling, design for the environment', 'ECD', 'Parametric life-cycle assessment (LCA) models have been integrated with traditional design tools, and used in prior work to demonstrate the rapid elucidation of holistic, analytical tradeoffs between detailed design variations. However, a different approach is needed if analytical environmental assessment is to be incorporated in very early design stages. During early design stages there may be competing concepts with dramatic differences. Additionally, detailed information is scarce, and decisions must be made quickly. Thus, both the overhead in developing parametric LCA models for a diverse range of concepts, and the lack of detailed information make the integration of traditional LCA models impractical. This paper explores an approximate method for providing preliminary life-cycle assessments. Learning algorithms trained using the known characteristics of existing products might allow the environmental impacts of new products to be approximated quickly during conceptual design without the overhead of defining new models. Artificial neural networks are trained to generalize on product attributes and environmental inventory data from pre-existing LCA studies. The product design team then queries the trained artificial model with new high-level product attribute data to quickly obtain an impact assessment for a new product concept. Foundations for the learning system approach are established, and then an application within the DOME integrated modeling environment is provided. Tests have shown it is possible to predict impacts on life-cycle energy consumption, and that there is a basis for the method to be used in predicting solid material, greenhouse effect, ozone layer, acidification, eutrophication, winter smog, and summer smog. Issues for further research are also identified.
'), new pubdetails('00-senin-basys', '', '', '', '', '', '', '', 'optimization/search, design automation, design for manufacture', 'DOME', 'A modular approach for modeling, simulation and control of production systems, embracing multiple domains from the shop floor level to the single workstation is presented. Applications and knowledge are integrated over a computer network in a concurrent engineering environment involving both software and human interaction. Integration is achieved by using Distributed Object Technology. To demonstrate the approach, a tool for shop floor simulation based on stochastic timed Petri Nets, and a tool for simulation of machine tool motions during part program execution are integrated using a framework based on Distributed Object Technology, to build a multiple domain model of a manufacturing cell. The implemented prototype can be used for process parameter configuration, system simulation, and eventually real time system supervision and control in a concurrent engineering environment.
'), new pubdetails('00-andersson-asme-dt', 'full document', '142 kB', '00-andersson-asme-dt/dontindex/00-andersson.pdf', '', '', '', 'Linkoping University, Sweden', 'design automation, optimization/search, Pareto optimality, concept selection, simulation', 'DOME', 'Many real world engineering problems are characterized by the presence of several conflicting objectives. In this paper, a new multi-objective genetic algorithm is employed to optimize two different concepts of hydraulic actuation systems. The different concepts have been modeled in a simulation environment to which the optimization strategy has been coupled.
The outcome from the proposed optimization strategy is a set of Pareto optimal solutions elucidating the tradeoffs between competing objectives. By comparing Pareto frontiers for competing concepts, valuable insights about the properties of the different concepts can be gained. Depending on how the decision-maker values the different objectives, different design solutions are more appropriate. This is exemplified in the hydraulic actuation systems, where the acceptance of a large control error results in a design with low energy consumption.
In many cases, multiple objective problems are aggregated into one single overall objective function. Optimization is then conducted with one optimal design as the result. This paper presents a method where the solution space is searched for a set of Pareto optimal solutions, from which the decision-maker may choose the final design.
The method has been applied on two concepts of hydraulic actuation systems. The resulting Pareto optimal frontiers elucidate the advantages of the different concepts and, advice the decision-maker which concept to choose depending on his or her preferences.
'), new pubdetails('00-smyth-asme-dt', 'full document', '1.3 MB', '00-smyth-asme-dt/dontindex/00-smyth.pdf', '', '', '', 'Ford Motor Company, Bose', 'conceptual design, product design, form synthesis, geometric modeling, styling, aesthetics, industrial design', 'CAID', 'A method for the synthesis of aesthetic product form is proposed. The approach attempts to bring computation to bear on the notion of brand DNA, a term which is used in industry loosely to refer to the aesthetic form elements that contribute to brand identity. Increasingly, external suppliers are contracted to design, manufacture and deliver entire product subsystems. This fragmentation of aesthetic design execution has potentially negative implications for the cohesiveness of the overall product image. Once defined, the product form DNA can be distributed to all suppliers of design services so that a consistent product image can be maintained.
The model strives to advance aesthetic form design by proposing an archetype approach, evolving product forms and supporting design in both digital and physical media. It emphasizes the modeling of a re-usable form archetype ?the skeleton - as opposed to the skin approach favored by existing geometric modeling tools. This would allow the designer to specify certain key parameters by defining the archetype of the product and perhaps character lines and apply corporate or style-specific DNA. The system will suggest skins and the designer will guide the development of the skin, but will not need to understand the complex underlying mathematical representation. Simulating the evolution of form has the benefits of hiding the process of geometry creation and allowing designers to quickly and thoroughly explore form space. The evolutionary algorithm acts as a suggestive tool and designers provide evaluative directions on the aesthetics of generated surfaces. Such a capability would not only free the designer from the repetitive, laborious construction of detailed geometry, but also allow her to experiment with a greater number of feasible concept variants per unit time.
A prototype implementation of the core form synthesis engine, FormDNA, is underway and examples of the synthesis process for procedurally generated and deformable parametric surfaces are described. The system is being incorporated into an industrial design cycle that supports the rapid alternation between digital and physical representations of product geometry.
'), new pubdetails('00-eisenhard-asme-dt', 'full document', '108 kB', '00-eisenhard-asme-dt/dontindex/00-eisenhard.pdf', '', '', '', 'Alliance for Global Sustainability', 'integrated design, product design, artificial neural networks, approximate life-cycle assessment, design for the environment ', 'ECD', 'Prior work has demonstrated the integration of detailed life-cycle assessment into a traditional design modeling process. Integration in this sense means the expert services of different design participants are linked while the content of the individual proprietary models remain distributed. This allows real-time tradeoff analysis without the typical barriers to communication—establishing and maintaining adequate synchronization of information. Although the method has been demonstrated as effective in supporting integrated environmentally conscious product design, it is still of limited value for conceptual design. While a detailed life-cycle model provides insight into a product\'s potential impact on the environment, it is often too time consuming for analysis during conceptual product design, where ideas are numerous and information is scarce.
The work presented in this paper explores an approximate method for preliminary life-cycle assessments without detailed modeling requirements, called a learning surrogate life-cycle assessment. Learning algorithms trained on the known characteristics of existing products allow the environmental impacts of new products to be approximated quickly during conceptual design. Artificial neural networks train on product attributes and environmental impact data from pre-existing life-cycle assessment studies. The product design team queries the trained artificial model with new high-level product attribute data to quickly obtain an approximate impact assessment for a new product concept.
An abbreviated life-cycle inventory list is developed for the output of the surrogate model. The list consists of key inventory elements most influential to different impact categories, balancing demands on the learning algorithm with the ability to predict categorized impacts. Tests based on this simplified inventory data have shown it is possible to predict impacts on life-cycle energy consumption, and that there is a basis for the method to be used in also predicting solid material, greenhouse effect, ozone layer depletion, acidification, eutrophication, winter smog, and summer smog impact categories.
'), new pubdetails('00-sosa-asme-dt', 'full document', '', '00-sosa-asme-dt/dontindex/00-sosa.pdf', 'slideshow', '', '00-sosa-asme-dt/dontindex/00-sosa-asme-slides.pdf', 'CIPD', 'design interfaces, integrative systems, modular systems, product architecture, systems engineering, team interactions', 'PD', 'This paper describes a method that allows us to enhance our understanding of the difference between designing modular systems and integrative systems. Modular systems are those whose interfaces are well defined and shared with only a few other systems. Integrative systems are those whose interfaces may be more complex and shared across the product. Our approach is illustrated by analyzing the development of a large commercial aircraft engine. We document both the product\'s design interfaces and the technical interactions between design teams. We found statistically significant differences in the ways modular and integrative design teams handle design interfaces. We focus our analysis on studying the effects due to organizational and system boundaries, and to the existence of various types of design interfaces. By identifying modular and integrative systems and by understanding the differences in designing those systems, development organizations can improve the integration process for complex designs.
'), new pubdetails('00-senin-pd', '', '', '', '', '', '', '', 'DOME, optimization/search, assembly', 'DOME', 'This work investigates the application of Genetic Algorithm (GA) based search techniques to concurrent assembly planning, where product design and assembly process planning are performed in parallel, and the evaluation of a design configuration is influenced by the performance of its related assembly process.
Several types of GA\'s and an exhaustive combinatorial approach are compared, in terms of reliability and speed in locating the global optimum. The different algorithms are tested first on a set of artificially generated assembly planning problems, which are intended to represent a broad spectrum of combinatorial complexity; then an industrial case study is presented. Test problems indicate that Genetic Algorithms are slightly less reliable than the combinatorial approach in finding the global, but are capable of identifying solutions which are very close to the global optimum with consistency, soon outperforming the combinatorial approach in terms of execution times, as the problem complexity grows. For an industrial case study of low combinatorial complexity, such as the one chosen in this work, GA\'s and combinatorial approach perform almost equivalently, both in terms of reliability and speed.
In summary, GA\'s seem a suitable choice for those planning applications where response time is an important factor, and results which are close enough to the global optimum are still considered acceptable; such as in concurrent assembly planning, where response time is a key factor when assessing the validity of a product design configuration in terms of the performance of its assembly plan.
'), new pubdetails('00-kraines-CEE', '', '', '', '', '', '', 'Alliance for Global Sustainability', 'DOME, environmentally-conscious design', 'DOME, ECD', 'Much of scientific research in the past has been driven by the belief that essentially random, undirected, and uncoordinated experiments result in serendipitous discovery that justifies the investment in that research. Recently, however, there is an increasing awareness that a more directed research structure is necessary for the following reasons:
In Chemical Engineering, the concept of unit operations has been used to structure and modularize knowledge about chemical processes and reactor systems. At the department of Chemical System Engineering in the University of Tokyo, we have been involved in developing tools and methodologies for an iterative directed research approach which combines experiments, computer simulation, theoretical modeling and knowledge structuring in a complementary fashion to address complex system analysis. We need now tools and methodologies to integrate the knowledge and research efforts of different fields of study. DOME, the distributed object-based modeling environment provides an infrastructure for this approach.
In this paper, we first examine the emergence of a "need" for technologies and methodologies to manage information. Then we introduce the concepts of interfacing and modularization that provide the basis for our approach. Next, we present the software integration infrastructure DOME, describing the concepts and motivation behind DOME\'s design. Finally, we show some of the actual system modeling applications of DOME to large-scale, complex problems.
'), new pubdetails('00-wallace-CAD', 'full document', '', '00-wallace-CAD/dontindex/service_marketplace.pdf', '', '', '', 'NSF CIPD, Ford ', 'distributed design, integrated design, design service marketplace', 'DOME', 'This paper presents a service marketplace vision for enterprise-wide integrated design modeling. In this environment, expert participants and product development organizations are empowered to publish their geometric design, CAE, manufacturing, or marketing capabilities as live services that are operable over the Internet. These services are made available through a service marketplace. Product developers, small or large, can subscribe to and flexibly inter-relate these services to embody a distributed product development organization, while simultaneously creating system models that allow the prediction and analysis of integrated product performance. It is hypothesized that product development services will become commodities, much like many component-level products are today. It will be possible to rapidly interchange equivalent design service providers so that the development of the product and the definition of the product development organization become part of the same process. Computer-aided design tools will evolve to facilitate the publishing of live design services. A research prototype system called DOME is used to illustrate the concept and a pilot study with Ford Motor Company is used in a preliminary assessment of the vision.
'), new pubdetails('00-borland-journal-IE', 'full document', '', 'http://mitpress.mit.edu/journals/JIEC/JIE3.23_Borland33-46.pdf', '', '', '', 'MIT Center for Environmental Health Sciences, NSF CIPD, Alliance for Global Sustainability ', 'DOME, design for the environment', 'ECD', 'This paper proposes a computer-based method for providing product designers with real-time environmental impact assessment. In this concurrent modeling approach, environmental experts build life cycle models, define their interfaces, and publish them as distributed objects on the Internet. Traditional designers integrating these objects into their design models have access to the impact assessment methods provided by the environmental expert. In this paradigm, the focus shifts from providing techniques that let non-expert designers perform environmental impact assessments to tools that provide designers with timely assessments from appropriate environmental experts. Establishing real time communication between the product design models and the environmental life-cycle models is the focus of this paper. A beverage container design example illustrates how this collaborative approach can use environmental and traditional design goals to elicit effective tradeoffs between design alternatives.
'), new pubdetails('99-senin-tr9901-dome', 'full document', '', '99-senin-tr9901-dome/dontindex/99-senin-tr9901.pdf', '', '', '', 'NSF CIPD', 'optimization/search, DOME, integrated design, distributed design', 'DOME', 'Product designers typically use a combination of custom, semi-custom, and standard components from catalogs during the design process. Likewise, mathematical models used to represent and analyze various sub-elements of a design may be custom or off-the-shelf. These models may be developed by different experts or vendors and operate in different computational environments. Ideally, designers would like to create a system model that conveniently links this set of heterogeneous and distributed models to represent the complete product, facilitating the rapid exploration of design tradeoffs and global optimization. A representation for this purpose is developed. Problems are decomposed into modules representing different aspects of a design problem, such as a physical element or an analysis viewpoint. Modules behave according to object-oriented principles, hiding the details of their implementation and providing a set of services, either locally or over the Internet, as a means of interaction with them. As the different aspects represented by modules are usually closely coupled, a synthesis mechanism is also provided. Modules can be linked to each other through their services to form a service exchange network. Each design participant can define local mathematical relations which coordinate the propagation of model state changes so that an integrated system behavior emerges from a set of distributed modules.
The module and service representation is formalized and specific modules for product design are outlined: variables, solvers, catalogs, decision support, and wrappers for third party software. A genetic optimization module is also detailed. The efficacy of the modeling and search method is tested on a variety of design problems using a software prototype implementation called DOME (distributed object-based modeling and evaluation). Optimization results using the representation and several genetic algorithms are provided.
'), new pubdetails('99-wallace-asme-dt', 'full document', '', '99-wallace-asme-dt/dontindex/DTM8780.PDF', '', '', '', 'NSF CIPD', 'DOME, design service marketplace', 'DOME', 'Refinement of product design processes is recognized as a means for organizations to impact lead times, cost, and quality. Although methods such as the design structure matrix (DSM) exist to analyze design processes, their applications have been limited by overhead in collecting data and maintaining accurate representations. Representation timeliness and resolution issues have typically led to static, high-level process views. In this work, the design process is derived dynamically from an evolving network of design resources, inter-related by service dependencies.
The design process is viewed as a system of transforming actions allocated to individual participants, individual computational tools or models, teams or organizations, and integrated computational tools. Elicitation and management of the process involves obtaining an ongoing, detailed, timely understanding of the complex transactions between these actions. In abstract, one can view the result of an action, performed by an individual, computational tool or organization, as a service. Thus, a transaction between individuals, computational tools, or organizations is a service relationship, and an entire process is a network of entities providing related services. If an integrated computational design environment is used to mediate service interactions, then it will also be possible to automatically extract the process structure.
A design structure matrix (DSM) module was constructed for a research integrated modeling system called DOME, which is based on this service exchange concept. The DSM module automatically creates a process visualization from the relationships in service networks and updates as new resources or services are added to the model, providing real-time process evolution feedback. The module provides a matrix visualization of service relationships and mechanisms to sort resources according to service priority. This allows for real-time evaluation of design processes as models of the product evolve, providing information to guide the strategic addition of resources. In the future, a product design process may no longer be viewed as a static institutionalized structure, but rather as a dynamic network that is constructed in real-time as design goals, needs, priorities, and resources evolve.
'), new pubdetails('99-abrahamson-asme-dt', 'full document', '', '99-abrahamson-asme-dt/dontindex/CIE9084.PDF', '', '', '', 'NSF CIPD', 'DOME, integrated design', 'DOME', 'This paper describes an integrated product design study conducted with Polaroid Corporation for a liquid crystal display video projector, applying a research system called DOME (Distributed Object-based Modeling Environment). Integrated models are created by building a computational service exchange network, thereby interconnecting the input and output services of different design participants. The services of distributed objects—encapsulating CAE simulations, component catalogs, manufacturing cost models, geometric and configuration models, customer preference models, and environmental life-cycle assessment—are mathematically related to form an integrated product system model. Software objects providing optimization and decision support are also included in the model to create a design tradeoff environment. As such, designers can obtain sales predictions based upon configuration changes and make tradeoffs with other requirements.
Benchmarking suggests there would be approximately a 30% increase in the time to fully evaluate the first design configuration due to the overhead of creating the integrated system model. However, the time to fully evaluate subsequent alternatives may be reduced from months to minutes, possibly reducing overall design time by up to 50%.
'), new pubdetails('99-sousa-lifecycle', 'full document', '', '99-sousa-lifecycle/dontindex/lifecycle.pdf', 'slideshow', '', '99-sousa-lifecycle/dontindex/lifecycle-slides.pdf', 'Alliance for Global Sustainability', 'Environmentally-conscious design, integrated design', 'ECD', 'It is important to trade-off environmental effects and product performance, preferably using integrated analytical models, when designing a product. One approach to collaborative modelling for this purpose is to view sub-models, including life cycle assessment models, as objects which exchange services. This communicating object architecture for integrated environmental assessment has been demonstrated successfully in prior research. However, it is often prohibitively time consuming to build detailed LCA models when designs are evolving rapidly. This paper explores the feasibility of an approach to address this issue, using surrogate LCA models based upon learning systems which extrapolate results from detailed product LCA analyses available in databases. First, a list of key product attributes or descriptors is identified based on existing LCA studies and databases. Then, a database of LCA results, mapped to these attributes, is developed. A neural network is trained using these data to generalize the relationships between product attributes and LCA indicators. The services of this surrogate LCA model are then made available to other models so that it can provide preliminary predictions of LCA indicators for a new design without requiring a new detailed LCA model. A preliminary implementation is described and tests to assess the feasibility of the approach are provided. The trained surrogate LCA model is used to predict LCA indicators of products and compared with traditional LCA results and issues requiring further research are outlined.
'), new pubdetails('99-design-seminar', 'full document', '', '99-design-seminar/dontindex/CIRPfinal.pdf', 'slideshow', '', '99-design-seminar/dontindex/Conference_slides.pdf', 'NSF CIPD', 'DOME, distributed design, integrated design', 'DOME', 'Product design typically involves a mix of custom, semi-custom, and standard components. A mathematical model of a design problem may also mix custom models and cataloged models that represent standard elements or families of elements. Further, design problem models are often decomposed into subsystems and delegated to appropriate experts, each of whom may use their own preferred modeling and analysis tools. A degree of coupling usually exists between these subsystems so they cannot be optimized independently, yet it can be difficult to understand their integrated system performance. Ideally, it would be possible to link this set of heterogeneous and distributed models to represent the complete product in a way that facilitates rapid exploration of design tradeoffs and global optimization.
This paper describes a general object-based system modeling formalism forproduct design. It allows the convenient integration of custom models, models from catalogs, and a wide variety of software modeling and analysis tools distributed over the Internet. Subsystems operating in different locations using different modeling tools are integrated, and mixed variable optimization using custom, semi-custom and standard elements is made possible. A bottle design problem is implemented using a software prototype called DOME (Distributed Object-based Modeling and Evaluation), linking models in ProEngineer, Excel, TEAM (life-cycle analysis software), and DOME custom models.
'), new pubdetails('99-senin-GECCO', 'full document', '', '99-senin-GECCO/dontindex/senin-GECCO.pdf', 'slideshow', '', '99-senin-GECCO/dontindex/gecco-senin-slides.pdf', 'NSF CIPD', 'DOME, distributed design, integrated design, optimization/search', 'DOME', 'DOME (Distributed Object-based Modeling Environment) is a software framework for product design system modeling where designers are distributed geographically and make use of different software tools. Designers develop their own local software components, and distributed object technology is used to integrate their services via the Internet to form an overall system model. Designers can then explore alternatives by making changes to local models or remote services while observing how the entire model responds.
This exploration is amenable to automated search, which involves both continuous parameters (changing the value of services) and discrete changes (selecting different objects to substitute entire local models). A genetic optimization object and appropriate direct representation genomes and operators were developed for this purpose. The effectiveness of several genetic algorithms was compared and a new variation of restricted tournament selection (RTS) was developed. The RTS variation, called the Struggle GA, most reliably located the global optima and the most local optima. Other crowding algorithms reliably located the global optima but were less successful identifying multiple local solutions. The global algorithms (simple and steady state) did not reliably locate the global optima in mixed variable problems. Finally, a realistic beverage container design example is presented.
'), new pubdetails('98-borland-asme-dt-conference', 'full document', '', '98-borland-asme-dt-conference/dontindex/98-borland.pdf', 'slideshow', '', '98-borland-asme-dt-conference/dontindex/98-borland-slides.pdf', 'MIT Center for Environmental Health Sciences, NSF CIPD, Alliance for Global Sustainability', 'DOME, design for the environment, best paper award', 'ECD', 'Every product introduced into the market has some impact on our environment: each is made from raw materials, uses energy, or creates some waste. However, designers and companies can make every attempt to minimize the environmental impacts of their products. This is not an easy task, because many complicated tradeoffs are involved, and the most significant improvements come from rethinking and redesigning products from the ground up.
This paper proposes a computer-based method for providing product designers with real-time environmental impact assessment. Although many environmental impact assessment programs exist, few are readily accessible to product designers ?and those that are available tend to be too simplistic to give realistic answers. By contrast, in the concurrent modeling approach proposed, environmental experts build life cycle models, define their interfaces, and publish them as distributed objects on the internet. Traditional designers integrating these objects into their design model have access to the impact assessment methods provided.
It is hypothesized that this method makes environmental impact assessment a more accessible and realistic option for product designers during the critical early stages of design. A product design example is used to illustrate how this collaborative approach can be effective in determining tradeoffs between design alternatives.
Recipent of the ASME International Design Enginering Division, Design for Manufacturing Committee IBM Best Paper Award.
'), new pubdetails('98-kim-tr9801-semivariance', 'full document', '', '98-kim-tr9801-semivariance/dontindex/98-kim.pdf', '', '', '', 'Leaders for Manufacturing', 'decision making, tradeoffs', 'DOME', 'Design decision making under uncertainty often uses probabilities to represent and quantify uncertainties associated with early design alternatives. Expected value, the most compact statistic to summarize a probability structure, is used to compare design candidates each of which may have distinct uncertainty structures. However, often there is a discrepancy between the decisions based upon expectation and the final realization after all uncertainties are resolved.
In order to understand and partially address this issue, expectation-based decision making frameworks are reviewed and compared with mean-variance analysis used in finance theory. With the insights gained from the comparison, a descriptive set of metrics is suggested to supplement an expectation-based decision rule. The concepts of risk and opportunity will be introduced and mathematically defined. These variability related metrics may provide insight into improvement potential or payoff at higher risk.
'), new pubdetails('98-wetice-wallace', 'full document', '', '98-wetice-wallace/dontindex/98-wetice-wallace.pdf', 'slideshow', '', '98-wetice-wallace/dontindex/98-wetice-wallace-slides.pdf', 'NSF CIPD', 'DOME, distributed design, integrated design', 'DOME', 'This paper presents an open product development environment for distributed, collaborative, and integrated design. The web-based framework, called DOME, allows designers to build integrated models using both local and distributed resources, and to collaborate by exchanging services in a network-centric environment. An integrated model can be created while each participant focuses on their own area of expertise. A design problem model is created by connecting modules, each of which can represent specific components, analysis capabilities/software, disciplines, or organizations relevant to the problem. The resulting module network forms a concurrent model in which changes propagate through service exchanges. A Java applet-based user interface provides cross-platform and distributed user access to DOME module servers throughout the network.
'), new pubdetails('97-weiner-web', 'text-only document', '', '97-weiner-web/dontindex/97-Weiner_text_only.pdf', '', '', '', 'NSF ESCEL Program, NSF CIPD', 'design education, electronic education', 'DEd', '
This paper investigates how traditional lecture time might be used given the availability of effective web-based courseware for delivering materials typically presented in the classroom. After providing an overview of web-based educational materials, the effectiveness of two teaching approaches is compared using a product design lecture on visual prototyping. One group of students prepared for class using web-based materials and then received a lecture-style second coverage of the topic in class (web+class lecture). The other group of students prepared for class using the same web-based materials and then worked with the faculty, applying only a small portion of the subject matter, on illustrative examples (web+limited experience). The two groups then completed an assignment based on the subject matter. The average grade performance of the web+limited experience group was 10.8% higher than for the web+class lecture group. An achieved significance level of 0.001 provides very strong evidence to reject the hypothesis that the two groups performed equally. Analysis of how the two student groups used the web-lecture resource showed that the form of class instruction had a strong influence on student motivation for independent study. The average time spent in web-based preparation by the web+limited experience group was 1.6 times greater than the average preparation time for the web+class lecture group. Sixty percent of the web+class lecture group prepared less than the least prepared student in the web+limited experience group. These findings suggest that, if codified materials are primarily delivered using a medium such as the WWW, traditional classroom time might be liberated for potentially higher value-added activities such as mentoring and experiential activities.
'), new pubdetails('98-pahng-dome', 'full document', '', '98-pahng-dome/dontindex/98-pahng-dome.pdf', '', '', '', 'Leaders for Manufacturing, NSF CIPD', 'DOME, distributed design, integrated design', 'DOME', 'This paper proposes a framework for the modeling and evaluation of product design problems in a computer network-oriented design environment. The framework is intended to integrate designer-specified mathematical models for multi-disciplinary and multi-objective design problems. The goal is to provide the ability to rapidly construct integrated design problem models to facilitate collaborative design work, improve product quality and reduce development time. Ultimately, it should allow specialized engineering applications and design problem models to operate under a common design environment.
A product design problem is modeled in terms of interacting objects, called modules, each representing a specific aspect of the problem. Modules interact with one another through services that allow the exchange of information. Modules can encapsulate engineering models and data or software applications. The method is extended using a standard network communication protocol to create a distributed object-based modeling and evaluation framework for design problems.
'), new pubdetails('98-pahng-asme-dt-conference', 'full document', '', '98-pahng-asme-dt-conference/dontindex/98-pahng-eim.pdf', 'slideshow', '', '98-pahng-asme-dt-conference/dontindex/98-pahng-eim-slides.pdf', 'NSF CIPD', 'DOME, distributed design, integrated design', 'DOME', 'This paper presents an integrated but open product development environment for distributed and collaborative design. The web-based framework, called DOME, allows designers to build integrated models using both local and distributed resources and to collaborate by exchanging services. Thus, an integrated model can be created while each participant focuses on their own area of expertise. A design model is created by connecting modules, each of which can represent specific components, analysis capabilities/software, disciplines, or organizations relevant to the problem. The modules interact with each other using service exchanges based upon the CORBA standard communication protocol for distributed objects. The resulting module network. forms a concurrent model in which changes propagate through service exchanges. Modules can simultaneously function as both clients (using services from other modules) and servers (providing services to other modules). A Java applet-based user interface provides cross-platform and distributed user access to DOME module servers throughout the network.
'), new pubdetails('97-senin-tr9702-dome', 'full document', '', '97-senin-tr9702-dome/dontindex/97-senin-tr9702.pdf', '', '', '', 'NSF CIPD', 'optimization/search, DOME, integrated design, distributed design ', 'DOME', 'In this paper an integrated modeling and search framework is proposed to optimize product design problems which involve simultaneously making choices from discrete sets of alternatives (catalog-based design) and the setting of design parameters defined over continuous intervals (parametric design).
The modeling framework is based on the decomposition of problems into modules that interact by exchanging information services through interfaces. Modules contain embedded models composed of continuous variables, mathematical models or software programs and represent an aspect of a design problem, such as a component or an analysis viewpoint. Modules can be stored in catalogs so that they may be reused and substituted in a problem model to explore alternatives. The framework is used to optimize both module selections from catalogs and continuous parameters within modules. The optimization engine uses a genetic algorithm, and a genome representation and genetic operators are developed for the purpose. A high-level model definition software language is implemented for the framework so that the modeling and search approach may be readily applied to a wide variety of design problems.
The efficacy of the modeling and search approach is tested for several genetic search algorithms, including simple, steady state and crowding techniques. The crowding algorithms most reliably located the global solution. A particular variation, referred to as Struggle2, offered the best overall performance, reliably locating global solutions in a relatively low number of evaluations while simultaneously locating many local solutions.
'), new pubdetails('97-kim-asme-dt', 'full document', '', '97-kim-asme-dt/dontindex/97-kim.pdf', 'slideshow', '', '97-kim-asme-dt/dontindex/97-kim-asme-slides.pdf', 'MIT Leaders for Manufacturing', 'DOME, decision making, tradeoffs', 'DOME', 'In contrast to their importance, decisions during early design stages are often made using less information than is available in later stages, especially for new products. Additional information and data becomes available through the design process. Early design activities, upstream of detailed design, require decision-making based upon limited information. This paper attempts to address this issue through the articulation of design goals and the consideration of uncertainty.
A prescriptive goal-oriented design evaluation model is presented. The approach uses goals or targets as is commonly observed in design practice. After providing motivation for a goal-oriented approach in the early stages of product design the concept and method, based upon subjective probabilities of acceptance, are formalized. A unidimensional acceptability function is defined and the mathematical construction of a multidimensional acceptability function is presented. It is hoped that the method provides an intuitive design evaluation tool, balancing operational simplicity and mathematical accuracy.
Additionally, the structure of uncertainty in an evaluation is analyzed to provide insight into the potential for later design improvement. Designs with high improvement potential may initially not have the highest expected acceptability based upon preliminary design performance estimates. The use of conditional standard deviations in acceptability is proposed as a qualitative measure of improvement potential.
Finally, an engine selection problem is presented to illustrate the approach. In this example, a bootstrap resampling method is used to construct probability mass functions to estimate the performance of design alternatives. Acceptability functions are constructed and the performance estimates are used to evaluate design alternatives.
'), new pubdetails('97-jackson-shielding', '', '', '', 'slideshow', '', '97-jackson-shielding/dontindex/97-jackson-shielding-slides.pdf', 'MIT/Volvo Environmental Research Award', 'DOME, design for the environment', 'ECD', '
Designing products for reduced environmental impact and improved performance requires that a diverse range of characteristics be considered. Such tradeoffs require a practicable method for combining specialized tools into an integrated analytical system. The paper describes how an integrated design tool can be constructed from modules which correspond to different aspects of the problem, such as geometry, materials and mathematical subproblems. An electronics housing design example (subject to electromagnetic, heat transfer, environmental and cost considerations) illustrates the use of the design methodology. The tool is then used to demonstrate the ability to investigate and understand interactions between different design goals.
'), new pubdetails('97-jackson-asme-dt', 'full document', '', '97-jackson-asme-dt/dontindex/97-jackson-asme.pdf', 'slideshow', '', '97-jackson-asme-dt/dontindex/97-jackson-asme-slides.pdf', 'MIT/Volvo Environmental Research Award', 'DOME, design for the environment', 'ECD', 'The environmental and health impacts resulting from the production, use, recycling and disposal of consumer products is increasingly perceived a product design issue. As a result, there has been a concerted research effort to develop procedures and tools to quantify, evaluate and manage environmental impact. This effort has focused on assessing the effects of products and their predefined life-cycles, primarily based upon static inventory data. However, a productís demand may display pronounced variability over time due to its seasonal nature, its marketable life-cycle or its reliability characteristics. Environmental impact is often dependent upon emission rates and accumulation, which also may vary with product demand.
This paper presents a method for simulating the time-based fluctuations of resource flows in a product life-cycle using a modular approach intended to facilitate the incorporation of LCA into product design. These resource flows may be used for time-dependent inventories for use in environmental impact assessment. A general process module is defined relating resource inputs and outflows, based upon an embedded mathematical model. Then, a parametric model to represent the average performance of manufacturing processes is proposed for use within modules. Different parameter values may be used to represent a variety of life-cycle processes. Individual modules are combined to form product life-cycle networks. A designer specifies the required system output (product demand) as a function of time, and the integrated network calculates the necessary time-dependent resource flows throughout the network.
The generalized modules are used to create a simple life-cycle network for a product involving the manufacture and assembly of a plastic and an aluminum part. The relevant production processes are implemented using generalized process modules which are linked to form a product life-cycle. Illustrative time-based resource flows for the life-cycle are presented.
'), new pubdetails('97-pahng-asme-dt', 'full document', '', '97-pahng-asme-dt/dontindex/97-pahng.pdf', 'slideshow', '', '97-pahng-asme-dt/dontindex/97-pahng-asme-slides.pdf', 'MIT Leaders for Manufacturing', 'DOME, distributed design, integrated design', 'DOME', 'This paper proposes a framework for the modeling and evaluation of product design problems in a computer network-oriented design environment. The framework is intended to integrate designer-specified mathematical models for multi-disciplinary and multi-objective design problems. The goal is to provide designers with the ability to rapidly construct integrated design problem models that reflect interactions between subproblems, facilitating collaborative design work and reducing development time while improving product quality. Ultimately, it should allow specialized engineering analysis applications, product data management systems, CAD models and design problem models to operate under a common design environment.
| |
| (a) | (b) |
| Figure 1. Product topology of a battery-powered cordless drill and its decomposition into modules | |
A product design problem is modeled in terms of interacting objects, called modules, each representing a specific aspect of the pro blem. Modules interact with one another through services that allow the exchange of information. Modules can encapsulate engineering models and data or software applications. The method is extended to create a distributed object-based modeling and evaluation framework, enabling the modeling and evaluation of design problems in the distributed design environment. Interaction between modules within the designerís local representation of the problem and remote modules (distributed objects) available in the networked environment is achieved using standard communication protocols. After reviewing the elements of the proposed framework, its preliminary software implementation is described and demonstrated for a cordless drill design problem.
'), new pubdetails('97-senin-borland-dome', 'full document', '', '97-senin-borland-dome/dontindex/97-senin-borland.pdf', 'slideshow', '', '97-senin-borland-dome/dontindex/97-senin-borland-slides.pdf', 'NSF CIPD', 'DOME, distributed design, integrated design', 'DOME', 'This paper describes a framework for the distributed modeling and evaluation of product design problems in a network-oriented design environment. The framework, intended to facilitate integrated product life-cycle design, uses distributed computer-based models and is applicable to collaborative design. The goal is to provide designers, manufacturers and customers with the ability to rapidly construct integrated design problem models accounting for interactions between different viewpoints and to incorporate specialized engineering analysis applications, CAD models and design problem models under a common design environment. An overview of the framework is provided and an application to the integrated design and environmental assessment of cardboard packaging is described. The example illustrates how design models can be used to configure products to meet customer performance expectations and also to diagnose potential problems.
'), new pubdetails('97-tomiyama-cirp-life-cycle', 'full document', '', '97-tomiyama-cirp-life-cycle/dontindex/97-tomiyama.pdf', '', '', '', 'Alliance for Global Sustainability', 'environmentally-conscious design, holistic design', 'ECD', 'This paper describes an in-progress project to develop a framework for holistic life cycle design. Environmentally-conscious design requires the balance of a variety of deeply coupled issues. Thus a holistic approach which addresses both the design of the product and its life-cycle is needed. A design support system for modelling, simulating and evaluating product life cycle is described and an example of modular product design is illustrated. The results show that product design significantly impacts product life cycle design and provide a foundation for the concept of holistic product life cycle design. Finally, a prototype system for integrating distributed design models to enable holistic design is outlined.
'), new pubdetails('96-mutooni-web', '', '', '', '', '', '', 'NSF ESCEL Program, NSF CIPD', 'design education, electronic education', 'DEd', '
This paper investigates the hypothesis that the world-wide-web (WWW) presents a new opportunity to interactively present and disseminate curricula. The WWW appears to have the flexibility needed to let students order the material and choose the presentation format that best suit their preference. The objective of this paper is twofold: to develop a general web-lecture structure that provides such flexibility, and to compare its efficacy with a classroom style lecture. The average grade performance of the students receiving web instruction was higher than for those receiving traditional classroom instruction. An achieved significance level of 0.063 provides reasonably strong evidence to reject the hypothesis that the two groups performed equally. Analysis of web-lecture use patterns revealed that the web-group students spent roughly the same amount of time on-line as the classroom group spent in lecture. However, the web group typically covered the materials in 3 short 30 minute sessions as opposed to one 90 minute lecture. Different students in the web group showed different preferences for either video or text-with-image presentation of materials. In a focus group, web students indicated that they valued the ability to pace their learning and review materials as needed. The positive results of this preliminary experiment raise the question of how teaching could change to exploit information technology and provide more effective and economical engineering education. Primary learning might take place individually through a medium such as the WWW, thus liberating classroom time for experiential activities.
'), new pubdetails('97-weiner-wheat', 'full document', '', '97-weiner-wheat/dontindex/97-wheat.pdf', 'slideshow', '', '97-weiner-wheat/dontindex/97-weiner-wheat-slides.pdf', 'MIT Libraries, NSF ESCEL Program', 'design education, product design, information literacy', 'DEd', 'In a senior MIT Mechanical Engineering course on the Product Engineering Process, students work in large teams of approximately 24 individuals to design and build working prototypes of new products. The effort spans the product development process, including: generating ideas, gathering customer and market data, selecting ideas and devising concepts, building and testing mockups, customer evaluation/focus groups for mockups and embodiment design/working prototype construction. The large teams must work effectively to realize this task, learning about group dynamics, role definition, consensus building and the value of communication.
In the course it has become apparent that students are deficient in their ability to gather, analyze and utilize information or understand the economic impacts of engineering decisions. To address this problem, students are now formally introduced toinformation collection as a component of the product design process. The students are required to gather and analyze market, product and technical information obtained through both primary (customer surveys/interviews) and secondary (published) sources. Students utilize information technology to locate data, transfer knowledge and coordinate their activities, thereby increasing their awareness of the vast quantities of information available in published literature, on-line databases and the World Wide Web (WWW).
This paper describes how students are introduced to information gathering, analysis, and utilization. Students are provided with a WWW-based guide for marketing research and individual help from a librarian. Students are exposed to the various data sources through a team based information treasure hunt. A design professional presents a benchmarking case study representing 10 hours of information collection, illustrating the process of gathering and distilling information to benchmark and cost products. In order to assess the effectiveness of the effort, a survey was given at both the beginning and end of the course. The initial survey measures student perceptions about the relative importance of the information gathered in the various phases of the product development process. The surveys have been used to evaluate how these perceptions changed through exposure to information gathering. The survey results indicate an increased ability to anticipate information needs and an increased awareness of information resources.
'), new pubdetails('96-senin-tr9602-ga', 'full document', '', '96-senin-tr9602-ga/dontindex/96-senin.pdf', '', '', '', 'NSF CIPD', 'optimization/search', 'DOME', 'This paper investigates the extension of genetic algorithms to multimodal optimization (niching methods). After reviewing current niching methods, a new variation of a crowding technique, named the struggle genetic algorithm, is introduced. Replacement occurs only between similar individuals based upon a similarity measure if an offspring wins the competition. Using a suite of test problems, the performance of the struggle genetic algorithm and three other niching methods ?deterministic crowding, restricted tournament selection, and fitness sharing - is empirically examined. For each test problem the struggle GA consistently located a more complete set of optimal solutions. The struggle GA has also performed well when compared to global methods (simple and steady state GA). Additionally, crossover\'s adaptive mechanism based upon the similarity measurement of parents is investigated for commonly used representations. Empirical investigations suggest that real-coded parameters are superior to the traditional binary or Gray-coding of continuous variables. For real-coded variables a newly designed crossover operator, the sphere-crossover, is introduced and tested. Preliminary test results illustrate its adaptive power.
'), new pubdetails('96-senin-asme-dt', 'full document', '', '96-senin-asme-dt/dontindex/96-senin.pdf', 'slideshow', '', '96-senin-asme-dt/dontindex/96-senin-asme-slides.pdf', 'Sloan foundation', 'design automation, optimization/search, DOME', 'DOME', 'In this paper a general design and catalog representation is proposed and implemented. The representation is applied to mixed continuous variable and catalog search using genetic algorithms. The representation addresses a number of common catalog search design scenarios. The capabilities of the model are demonstrated through an object-oriented computer implementation that uses a genetic optimization and search algorithm.
The variables associated with a design or subdesign are represented by graph-like structures of isolated or interrelated nodes. Any node has the ability to modify its state according to changes in its environment. A design or subdesign also may have design criteria which evaluate performance variables against specifications. Each subdesign or design has an interface, through which it may pass information to its external environment or receive information from its environment. Conceptually, there is no distinction between a design and a subdesign, and any design or subdesign can be composed of numerous subdesigns. A catalog is a collection of records, each containing a subdesign. A catalog may contain subdesigns with different topologies. Catalog hierarchies may be defined to represent different levels of data abstraction or classifications.
The representation has been applied to the optimization of mixed continuous variable and catalog search problems using a genetic algorithm. In addition to searching catalogs of static data (e.g., different sizes of pipes), it is possible to search subdesigns with internal models that can provide performance data based upon operating conditions (e.g., a DC motor). Thus, models for subdesign topologies are modularized and it is possible for the topology of subdesigns to be altered during an optimization. The use of catalog hierarchies allows the search to simultaneously consider catalogs from different vendors or determine how precise a selection must be in order to meet the design goals (e.g., will any structural steel suffice or is a specific alloy required?). Although this representation is very general and flexible, the number of subdesigns and the connections between subdesigns are defined by the designer and fixed during the optimization. Thus, at present, the macro-topology of the problem is constant.
Preliminary tests indicate that the genetic algorithm performs reasonably for this type of combinatorial design problem. However, further tests to assess the choice of parameters for the genetic algorithm and evaluating the scalability of the approach are required.
'), new pubdetails('96-shu-ieee-dfe', 'full document', '', '96-shu-ieee-dfe/dontindex/96-shu.pdf', '', '', '', 'Eastman Kodak', 'environmentally-conscious design, decision making', 'ECD', 'This paper summarizes our progress in using probabilistic methods to enable both single-domain and multiple-domain considerations in life-cycle design. As an example of a single-domain application, a probability-based reliability model that describes the effect of remanufacture on the reliability of parts and systems is outlined and experimentally verified. This reliability model is applied to fastening systems and used to calculate fastening-related remanufacture costs. These costs are integrated with other life-cycle fastening-related costs, thus enabling the simultaneous consideration of multiple domains. The combined life-cycle cost is optimized using genetic algorithms. The implementation and general results of genetic algorithm-based optimization of a product fastening and joining plan are described.
'), new pubdetails('96-wallace-journal', '', '', '', '', '', '', 'Carl G. Sontheimer Fellowship', 'DOME, decision making, tradeoffs, optimization/search ', 'DOME', 'In this paper a specification-based design evaluation method is presented and applied to optimization using genetic algorithms. The evaluation model is intended to emulate how specifications are used by product designers in a concurrent design environment. The approach allows designers to formulate optimization problems in terms of design specifications. The method is intended for a wide variety of design applications, including the location of feasible designs or starting points for other mathematical programming techniques. The model is demonstrated through a computer implementation that uses a genetic optimization and search algorithm. The penalty-based objective formulation strives for designs which are acceptable according to specifications.
'), new pubdetails('93-dighe-structural-synthesis', '', '', '', '', '', '', '', 'design automation, design for manufacture', 'CAID', 'Currently, the abstract for this paper is unavailable.
'), new pubdetails('93-wallace-automated-ieee', 'full document', '', '93-wallace-automated-ieee/dontindex/93-automated.pdf', '', '', '', 'Leaders for Manufacturing', 'design automation, design for manufacture, aesthetics, industrial design', 'CAID', '
Computer technology has provided many artistic fields with creativity enhancing tools. Music and cinema are just twoexamples. In contrast, the impact of computer-aided industrial design seems to be lagging. In the past, the technology required for interactive three dimensional computer graphics has been prohibitively expensive for most industrial designers. This is no longer true. Powerful graphics workstations are now within the reach of most design firms. The under-utilization of computer technology in industrial design can no longer be attributed to the cost of hardware.
Early attempts at computer-based industrial design were generally not well received, and have created a negative bias towards computers. Designers will not incorporate new tools unless they improve upon ?not merely replicate ?familiar design methods. We are working on design tools that actively assist conceptual product design. Thus far our work is directed towards the design of consumer electronics products with injection molded housings. Product design is considered from a consumer\'s perspective. Our research software can generate alternative design concepts based upon manufacturing, ergonomic, aesthetic, and style considerations. Unlike analytical engineering design tools, we concentrate upon less quantitative aspects of product design.
We hope such systems will join conceptual engineering design and industrial design into a unified process. We feel that modern design suffers from a duality of approach, depending upon whether a design is to be judged as useful or beautiful. Broadly speaking, industrial design is biased towards visualization, while engineers focus upon performance. The efforts of industrial and engineering designers are often at odds, yet clearly the successful design of consumer products requires the skills of both the industrial designer and the engineer.
Axiomatic design is a general method that helps designers structure and understand design problems, thereby facilitating the synthesis of suitable design requirements, solutions, and processes. The approach also provides a consistent framework from which the merits of designs can be quantified. In this paper we briefly review the tenets of axiomatic design, and demonstrate how information content can be applied to environmental design problems–an issue that is becoming increasingly important. A prototypical computer program is used to illustrate the approach.
'), new pubdetails('92-wallace-design', '', '', '', 'slideshow', '', '92-wallace-design/dontindex/92-wallace-slides.pdf', '', 'design education, industrial design, product design', 'DEd', '
The responsibilities of industrial and technical designers are highly delineated and often at odds. Broadly speaking, industrial design is biased towards visualization, while engineers focus upon performance. This division is reinforced by current educational practice. We believe that technological advances in design are beginning, and will continue, to knock down this separation. The ever increasing scope of design in concurrent design, design for manufacture, design for assembly, design for recycling, etc., calls for a new breed of multi-disciplinary designers with integrative capabilities. We begin the paper with research that illustrateshow technology might reshape future industrial design practice. We then propose a hypothetical design program intended to meet the needs of the future. Three major aspects of design education are discussed: curriculum, teaching methods, and faculty. A multi-disciplinary undergraduate curriculum is proposed. The program is intended to provide designers with knowledge that promotes flexibility of both in design capabilities and design tools or methods. A new approach to teaching, which emphasizes design research and the development of design tools, is recommended.
'), new pubdetails('91-wallace-caid', '', '', '', '', '', '', 'Leaders for Manufacturing', 'design for manufacturing, aesthetics', 'CAID', 'Currently, the abstract for this paper is unavailable.
') ); for (i=0; i