Search

Additive Manufacturing (AM) is a revolutionary development that is being viewed as a core technology for fabricating current and future engineered products. While AM has many advantages over subtractive manufacturing processes, one of the primary limitations of AM is to swiftly evaluate precise part build orientations. Current algorithms are either computationally expensive or provide multiple alternative orientations, requiring additional decision tradeoffs. To hasten the process of finding accurate part build orientation, a data-driven predictive model is introduced by mapping standard machining features to build orientation angles. A combinatory learning algorithm of classification and regression is utilized for the prediction of build orientation. The framework uses 54,000 voxelized standard tessellated language (STL) files as input to train the classification algorithm for eighteen standard machining features using a nine-layer 3D Convolutional Neural Network (CNN). Additionally, a multi-machining feature dataset of 1000 voxelized STL files are evaluated in parallel by performing quaternion rotations to obtain build orientation angles based on minimization of support structure volume. A regression model is then developed to establish a relationship between the machining features and orientation angles to predict optimal build orientation for new parts.

To appropriately leverage the benefits of Additive Manufacturing (AM), it would be advantageous if a printing could be guaranteed before allocating the necessary resources. Further, when considering AM for an inventory of existing components traditionally fabricated through traditional means, such a guarantee could result in significant technical and economic advantages. To realize such advantages, this paper presents a platform that allows for a successful and efficient transition of part-inventories to AM. This is accomplished using a novel design recommender system supported by machine learning, capable of making suggestions towards effective design modifications. This system uses an automatic AM-feasibility analysis of existing parts and a clustering of the parts based on similarities in their AM-feasibilities to develop a set of recommendations for those part clusters whose current designs are deemed as infeasible and/or inefficient for AM. The design modifications leverage a re-design algorithm to address not only problematic geometric issues but also potential infeasibilities associated with resource consumption. The utility of the presented modification algorithm is demonstrated using a number of case studies.

In the wake of increasing proliferation of renewable energy and distributed energy resources (DERs), grid designers and operators alike are faced with several emerging challenges in curbing allocative grid inefficiencies and maintaining operational stability. One such challenge relates to the increased price volatility within real-time electricity markets, a result of the inherent intermittency of renewable energy. With this challenge, however, comes heightened economic interest in exploiting the arbitrage potential of price volatility towards demand-side energy cost savings. To this end, this paper aims to maximize the arbitrage value of electricity through the optimal design of control strategies for DERs. Formulated as an arbitrage maximization problem using design optimization, and solved using reinforcement learning, the proposed approach is applied towards shared DERs within multi-building residential clusters. We demonstrate its feasibility across three unique building cluster demand profiles, observing notable energy cost reductions over baseline values. This highlights a capability for generalized learning across multiple building clusters and the ability to design efficient arbitrage policies towards energy cost minimization. Finally, the approach is shown to be computationally tractable, designing efficient strategies in approximately 5 hours of training over a simulation time horizon of 1 month.

The substantial role of additive manufacturing (AM) in fabricating unique geometries is undeniable in the domain of design and manufacturing. However, the successful implementation of AM technologies requires a consistency between the geometric specifications of a component and AM manufacturability capabilities and constraints. Otherwise, AM could result in failed prints and a wasteful use of resources. The goal of this research is to provide geometrically feasible designs for AM processes by rectifying the potentially infeasible geometries. To this end, a novel design modification system is presented that addresses the problematic areas of an AM-infeasible component using appropriate redesign solutions. This system also includes a geometric assessment algorithm which identifies the potential problematic part features using a comprehensive evaluation. Based on the obtained manufacturability feedback, the detected problematic features are then modified through a holistic design modification system. The functionality of the presented system is illustrated using a case study, and the effectiveness of the implemented modification approaches is also demonstrated through an experiment.

Increasing the complexity of engineering design projects expands of the diversity of required topic knowledge. Multi-disciplinary design processes have the need for expertise from multiple fields of study. In the context of mass collaboration within engineering design, positioning key members within multi-disciplinary teams is of great importance. Determining how each discipline impacts the overall design process requires an understanding of the mapping between competency and performance. This work explores this mapping through the use of predictive models composed of various regression algorithms. Design performance of students working on their capstone design project is analyzed and the relationship between individual competencies is compared against their overall project performance. Each competency and project is represented as a distribution of topic knowledge to produce the performance metrics. Following the automated topic extraction of the textual data, the regression algorithms are applied. Three topic models and five prediction models are compared for their prediction accuracy. From this analysis it was found that representing both input and output variables as a distribution of topics while performing a support vector regression provided the most accurate mapping between ability and performance.

The use of additive manufacturing (AM) for fabricating industrial grade components has increased significantly in recent years. Numerous industrial entities are looking to leverage new AM techniques to enable fabrication of components that were typically manufactured previously using conventional manufacturing techniques such as subtractive manufacturing or casting. Therefore, it is becoming increasingly important to be able to rigorously evaluate the technical and economic feasibility of additively manufacturing a component relative to conventional alternatives. In order to support this evaluation, this paper presents a framework that investigates fabrication feasibility for AM from three perspectives: geometric evaluation, build orientation/support generation, and resources necessary (i.e., cost and time). The core functionality of the framework is enabled on voxelized model representation, discrete and binary formats of 3D continuous objects. AM fabrication feasibility analysis is applied to 34 various parts representing a wide range of manifolds and valves manufactured using conventional manufacturing techniques, components commonly found in the aerospace industry. Results obtained illustrate the capability and generalizability of the framework to analyze intricate geometries and provide a primary assessment for the feasibility of the AM process.

This work leverages the current state of the art in reinforcement learning for continuous control, the Deep Deterministic Policy Gradient (DDPG) algorithm, towards the optimal 24-hour dispatch of shared energy assets within building clusters. The modeled DDPG agent interacts with a battery environment, designed to emulate a shared battery system. The aim here is to not only learn an efficient charged/discharged policy, but to also address the continuous domain question of how much energy should be charged or discharged. Experimentally, we examine the impact of the learned dispatch strategy towards minimizing demand peaks within the building cluster. Our results show that across the variety of building cluster combinations studied, the algorithm is able to learn and exploit energy arbitrage, tailoring it into battery dispatch strategies for peak demand shifting.

Mass collaboration within the design engineering process supports the inclusion of unique perspectives when working on complex problems. Increasing the number of individuals providing input and support into these perplexing challenges can increase innovation, decrease product development times and provide solutions that truly encompass the needs of the market. One of the greatest challenges within mass collaboration engineering projects is the organization of individuals within these large design efforts. Understanding which projects would most effectively benefit from additional designers or contributors is paramount to supporting mass collaboration design networks. Within such networks, there exists a large number of contributors, as well as, a large pool of potential challenges. Matching individuals with the challenges that they can provide the greatest benefit to, or building a team of individuals for newly developed challenges requires the consideration of previous performance and an understanding of individual competencies and design abilities. This work presents a framework which recommends individual project placement based on individual abilities and the project requirements. With this work a pool of individuals and potential projects are simulated and the application of a hybrid recommender system is explored. Overall it was found that recommended team compositions greatly outperform the baseline team development, most notably as greater consideration is placed on collaborative recommendations.

Due to their volatile behavior, natural disasters are challenging problems as they often cannot be accurately predicted. An efficient method to gather updated information of the status of a disaster, such as the location of any trapped survivors, is extremely important to properly conduct rescue operations. To accomplish this, an algorithm is presented to control a swarm of UAVs (Unmanned Aerial Vehicles) and optimize the value of the information gathered. For this application, the UAVs are autonomously navigated with a decentralized control method. With sensor technology embedded, this swarm collects information from the environment as it operates. By using the swarm’s location history, areas of the environment that have gone the longest without exploration can be prioritized, ensuring a thorough search. Measures are also developed to prevent redundant or inefficient exploration, which would reduce the value of the gathered information. A case study of a flood scenario is examined and simulated. Through this approach, the value of the proposed swarm algorithm can be tested by tracking the number of survivors found as well as the rate at which they are discovered.

The potential for mass collaboration and concurrent engineering among individuals working on a design project allows for increased innovation in idea generation while also supporting the parallelization of project development. The former of these two benefits is highlighted by recent corporate initiatives to incorporate crowdsourcing in their design process. This is done in an effort to garner new and unique perspectives while also driving designs that carry greater acceptance from the community and/or intended market. The latter benefit is emphasized by the pronounced success of concurrent engineering projects which tend to reduce production costs, all while enhancing overall product performance. This work aims to exploit these two concepts in a structured and methodical crowd centered design framework. A simulated pool of individuals, representing the crowd, are used to form design groups using game-theoretic principles. The outcome of each sub-game aims to increase the overall improvement to the designed system by maximizing what each individual can potentially add to the project. The optimization of resource allocation is then performed where time spent on individual or collaborative efforts is determined. The proposed framework is applied to simplified design projects at their initial conception, before detailed mathematical design models are determined.

A critical task in product design is mapping information from consumer space to design space. Currently, this process is largely dependent on the designer to identify and map how psychological and consumer level factors relate to engineered product attributes. In this way, current methodologies lack provision to test a designer’s cognitive reasoning, which could introduce bias while mapping from consumer to design space. Cyber-Empathic Design is a novel framework where user-product interaction data is acquired using embedded sensors. To understand consumer perceptions about a particular product, a network of latent psychological constructs is used to form a causal model allowing designers to better understand user preferences. In this work, we extend this framework by integrating choice-based preference modeling to develop a Discrete Choice Analysis integrated Cyber-Empathic design framework (DCA-CED). We model user preferences and ultimately consumer choice by considering perceptions estimated by psychological latent variables and user-product interaction data. To demonstrate the effectiveness of the framework, a case study using a sensor integrated shoe design is presented where data to represent user demographics, sensor readings, and product choice is simulated. Using the DCA-CED method, the model parameters are recovered and compared with the original parameter values in the simulator. In addition, the ability of the framework to predict choice based on user product-interaction data is tested. The results show that the analytical method effectively captures the underlying data generation process thereby validating the proposed framework and the analytical method.

Optimization research on operational strategies of energy use in building clusters have generally marginalized the effects of uncertainty in favor of reduced computational expense. This however leads to a significant disconnect between the expected energy cost and the average cost observed under uncertainty. Bridging this divide requires the incorporation of uncertainty analysis which poses both technical and computational challenges. This paper addresses these challenges through the notion of a Pareto band, demonstrating its applicability towards developing resilient operational strategies in a timely and computationally efficient manner. Under the proposed approach, Monte Carlo simulations are leveraged to reveal an envelope of optimality contained within the energy cost solution space. This optimality envelope, formally introduced as a Pareto band, is then used to train generalized linear models (GLMs) enabling robust operational strategy predictions. The results obtained from this approach highlight significant improvements in energy cost performance under uncertainty.

Usage context is considered a critical driving factor for customers’ product choices. In addition, the physical use of a product (i.e., user-product interaction) dictates a number of customer perceptions (e.g. level of comfort, ease-of-use or users’ physical fatigue). In the emerging Internet-of-Things (IoT), this work hypothesizes that it is possible to understand product usage while it is ‘in-use’ by capturing the user-product interaction data. Mining the data and understanding the comfort of the user adds a new dimension to the product design field. There has been tremendous progress in the field of data analytics, but the application in product design is still nascent. In this work, application of ‘feature learning’ methods for the identification of product usage context is demonstrated, where usage context is limited to the activity of the user. Two feature learning methods are applied for a walking activity classification using smartphone accelerometer data. Results are compared with feature-based machine learning algorithms (neural networks and support vector machines), and demonstrate the benefits of using the ‘feature learning’ methods over the feature based machine-learning algorithms.

A unique set of cognitive and computational challenges arise in large-scale decision making, in relation to trade-off processing and design space exploration. While several multi-attribute decision making methods exist in the current design literature, many are insufficient or not fully explored for many-attribute decision problems of six or more attributes. To address this scaling in complexity, the methodology presented in this paper strategically elicits preferences over iterative attribute subsets while leveraging principles of the Hypothetical Equivalents and Inequivalents Method (HEIM). A case study demonstrates the effectiveness of the approach in the construction of a systematic representation of preferences and the convergence to a single ‘best’ alternative.

One of the critical tasks in product design is to map information from the consumer space to the design space. Currently, this process is largely dependent on the designer to identify and map how psychological and consumer level factors relate to engineered product attributes. In this way current methodologies lack provision to test a designer’s cognitive reasoning and could therefore introduce bias while mapping from consumer to design space. Also, current dominant frameworks do not include user-product interaction data in design decision making and neither do they assist designers in understanding why a consumer has a particular perception about a product. This paper proposes a new framework — Cyber-Empathic Design — where user-product interaction data is acquired via embedded sensors in the products. To understand the motivations behind consumer perceptions, a network of latent constructs is used which forms a causal model framework. Structural Equation Modeling is used as the parameter estimation and hypothesis testing technique making the framework falsifiable in nature. To demonstrate the framework and demonstrate its effectiveness a case study of sensor integrated shoes is presented in this work, where two models are compared — one survey based and using the Cyber-Empathic framework model. It is shown that the Cyber-Empathic framework results in improved fit. The case study also demonstrates the technique to test a designers’ cognitive hypothesis.

In this paper, we strive to combine concepts from the field of ergonomics with techniques from design analytics in order to inform the specifications of a product, such that a designer can create products that are both universal and well-received by consumers. While the ergonomic guidelines dictated by anthropometric data can provide a broad target range for design features, consumer reviews of previous iterations of similar products can narrow the design further, within that given range, to pinpoint a more effective design. We establish a general, ergonomically-centered set of cue-phrases to identify and extract useful information from online reviews, after which this information can be coupled with anthropometric measurements to produce more accurate product dimensions and features. A new metric, the Frequency and Accuracy Summation (FAS) Number is also introduced, as a means to predict the likelihood that a cue-phrase will yield useful results. The usefulness of these cue-phrases as well as the utility of combining consumer review content with anthropometric data sets are then tested in a case study centered on the design of a new and more universal “ear bud” headphone. Using the established cue-phrases, reviews of an existing product are processed such that problem components could be targeted in the new design. The information that was gleaned from those reviews is then paired with actual human ear measurements to propose a new, hypothetical product design that is more ergonomically sound and universal.Copyright © 2015 by ASME

In light of the growing strain on the energy grid and the increased awareness of the significant role buildings play within the energy ecosystem, the need for building operational strategies which minimize energy consumption has never been greater. One of the major hurdles impeding this realization primarily lies not in the lack of decision strategies, but in their inherent lack of adaptability. With most operational strategies partly dictated by a dynamic trio of social, economic and environmental factors which include occupant preference, energy price and weather conditions, it is important to realize and capitalize on this dynamism to open up new avenues for energy savings. This paper extends this idea by developing a dynamic optimization mechanism for Net-zero building clusters. A bi-level operation framework is presented to study the energy tradeoffs resulting from the adaptive measures adopted in response to hourly variations in energy price, energy consumption and indoor occupant comfort preferences. The experimental results verify the need for adaptive decision frameworks and demonstrate, through Pareto analysis, that the approach is capable of exploiting the energy saving opportunities made available through fluctuations in energy price and occupant comfort preferences.Copyright © 2015 by ASME

A sustainable solution should holistically optimize all objectives related to the environment and a product’s cost and performance. As such, it should explicitly address material selection, which significantly affects environmental impacts and other objectives of a product design. While Life Cycle Assessment (LCA) provides credible methods to account for environmental impacts, current methods are not efficient enough for use at the early design stages to prune the entire design space without requiring execution of costly LCA analysis for each design scenario. Alternatively, surrogate modeling approaches can facilitate efficient concept selection during early design stages. However, material properties consist of discrete data sets, thus posing a significant challenge in the construction of surrogate models for numerical optimization. In this work, we address the unique challenges of material selection in sustainable product design in some important ways. Salient features of the robust surrogate modeling approach include achieving manageable dimensionality of LCA with a minimal loss of the important information by the consolidation of significant factors into categorized groups, as well as subsequent efficiency enhancement by a streamlined process that avoids the construction of full LCA. This novel approach combines efficiency of use with a mathematically rigorous representation of any pertinent objectives across an entire design space. To this end, we introduce an adapted two stage sampling approach in surrogate model construction based on a feasible approximation of a Latin Hypercube design at the first stage. The development and implementation of the method are illustrated with the aid of an automotive disc brake design, and the results are discussed in the context of robust optimal material selection in early sustainable product design.

The motivation behind this work is to integrate economic and environmental sustainability into decision making at the early phases of design through the development of a hierarchical concept selection method. Life Cycle Assessment (LCA) is a frequently implemented technique used to assess the environmental impacts of products, but it does not provide a simple means for including preference at different levels that can be used for comparison across design alternatives. A method is proposed to accommodate this issue expanding the Hypothetical Equivalents and Inequivalents Method (HEIM) to handle multi-level and multi-attribute trade-offs. The selection of a coffee maker design is used as an example to illustrate the implementation of the method with actual LCA results. The example provides valuable insights into how preferences may be elicited at different hierarchical levels and then combined to create a single utility score that represents to what extent each design alternative is preferred by the decision maker.

Traditionally viewed as mere energy consumers, buildings have in recent years adapted, capitalizing on smart grid technologies and distributed energy resources to not only efficiently use energy, but to also output energy. This has led to the development of net-zero energy buildings, a concept which encapsulates the synergy of energy efficient buildings, smart grids, and renewable energy utilization to reach a balanced energy budget over an annual cycle. This work looks to further expand on this idea, moving beyond just individual buildings and considering net-zero at a community scale. We hypothesize that applying net-zero concepts to building communities, also known as building clusters, instead of individual buildings will result in cost effective building systems which in turn will be resilient to power disruption. To this end, this paper develops an intelligent energy optimization algorithm for demand side energy management, taking into account a multitude of factors affecting cost including comfort, energy price, Heating, Ventilation, and Air Conditioning (HVAC) system, energy storage, weather, and on-site renewable resources. A bi-level operation decision framework is presented to study the energy tradeoffs within the building cluster, with individual building energy optimization on one level and an overall net-zero energy optimization handled on the next level. The experimental results demonstrate that the proposed approach is capable of significantly shifting demand, and when viable, reducing the total energy demand within net-zero building clusters. Furthermore, the optimization framework is capable of deriving Pareto solutions for the cluster which provide valuable insight for determining suitable energy strategies.Copyright © 2014 by ASME

Wind turbine tower design looks primarily at the structural integrity and durability of the tower. Optimization techniques are sometimes employed to maximize the loading capability while reducing material use and cost. Still, the tower is a dynamic part of a complex wind energy conversion system. During system operation, the tower is excited and sways back and forth. This undesirable movement increases cyclical loading on the tower and drivetrain components. To minimize this motion the tower frequency must be offset from the natural frequency of other components. Hence, it is necessary to look at the relationships that exist between the tower and other wind turbine components, such as the rotor, nacelle, and foundation. In addition, tradeoffs between cost, structural performance, and environmental impact can be examined to guide the designer toward a truly sustainable alternative to fossil fuels. Ultimately, an optimal design technique can be implemented and used to automate tower design. This work will introduce the analytical model and decision-making architecture that can be used to incorporate greater considerations in future studies. In this paper, nine wind turbine tower designs with different materials and geometries are analyzed using Finite Element Analysis (FEA). The optimal tower design is selected using a multi-level variation of the Hypothetical Equivalents and Inequivalents Method (HEIM). Using this analysis, a steel tower with variable thickness has been chosen. The findings reaffirm that steel is a favorable choice for turbine tower construction as it performs well on environmental, performance, and cost objectives. The method proposed in this work can be expanded to examine additional design goals and present a higher fidelity model of the wind turbine tower system in future work.Copyright © 2014 by ASME

During the design of complex systems, a design process may be subjected to stochastic inputs, interruptions, and changes. These design impulses can have a significant impact on the transient response and converged equilibrium for the design system. We distinguish this research by focusing on the interactions between local and architectural impulses in the form of designer mistakes and dissolution, division, and combination impulses, respectively. We find that local impulses tend to slow convergence but systems subjected to dissolution/division impulses still favor parallel arrangements. The strategy to mitigate combination impulses is unaffected by the presence of local impulses.

When developing an artifact, designers must first capture and represent consumer needs. These needs can then be transformed into system requirements or objectives. This paper proposes a formalized affordance-based method for capturing consumer needs which leverages previous work in the area of Affordance-based design. The creation of an affordance basis and a relational structure (the Desired Affordance Model) facilitates the use of affordances to represent consumer needs. This formalization has the potential to improve model quality and consistency, while managing model creation time. Further, this formalization imposes a level of abstraction which forces solution independence but is capable of capturing the large range of consumer needs.Copyright © 2013 by ASME

Many engineering departments struggle to meet “the broad education necessary to understand the impact of engineering solutions in a global, economic, environmental, and societal context” (Outcome h) that is required for ABET. As a result, engineering students receive meaningful contextual experiences in piecemeal fashion and graduate with a lack of concrete competencies that bridge knowledge and practice in the global world in which they will live and work. By considering products as designed artifacts with a history rooted in their development, our product archaeology framework combines concepts from archaeology with advances in cyber-enhanced product dissection to implement pedagogical innovations that address the significant educational gap. In this paper, we focus on assessing elements of a sustainable and scalable foundation that can support novel approaches aimed at educating engineering students to understand the global, economic, environmental, and societal context and impact of engineering solutions. This foundation is being developed across a network of partner institutions. We present recent results from freshman, sophomore, and senior courses at two of the partners in the national network of institutions.Copyright © 2013 by ASME

A growing area of research in the engineering community is the use of data and analytics for transforming information into knowledge to design better systems, products, and processes. Data-driven decisions can be made in the early, middle, and late stages in a design process where customer needs are identified and understood, a final concept for a design is chosen, and usage data from the deployed product is captured, respectively. Design Analytics (DA) is a paradigm for improving the core information-to-knowledge transformations in these stages of a design process resulting in better performing and functioning products that reflect both explicit and implicit customer needs. In this paper, a simulator is used to model usage of a hypothetical refrigerator and generate artificial data driven by four different customer behavior profiles with variation. The population of customers is randomly divided among the four behavior profiles so that the underlying customer preferences are unknown to the experimenter prior to data analysis. The purpose of the simulation is to illustrate the use of DA in the late stage of a design process to improve the transition from an existing product to the next generation product. Metrics are developed to analyze the product usage data, and both prevailing and subtle usage trends are identified. After conclusions are made, the study proceeds to the early and middle stages of a subsequent design process where a hypothetical next-generation refrigerator is conceptualized.Copyright © 2013 by ASME

The evolution of design thinking has seen numerous challenges and advances in transforming information into knowledge for engineers to design systems, products, and processes. These transformations occur in three stages throughout a design process. In simple form, the early, middle, and late stages of a design process serve to develop an understanding of the customer’s needs, arrive at the final concept of the design, and analyze and support the performance and usage profile of the deployed product, respectively. The quality and accuracy of the input information and the effectiveness of each transformation determine the success or failure of the product. Capturing good information and converting it to knowledge are two important tasks that have motivated a long history of research in design processes and tools. In this paper, we propose Design Analytics (DA) as a new paradigm for significantly enhancing the core information-to-knowledge transformations. The overall aim is to capture, store, and leverage digital information about artifacts, their performance, and their usage. The information is transformed into knowledge in each of the three stages using various analytics and cyber-enabled tools such as design repositories and concept generators. The ultimate result is better performing and functioning products. As web analytics has transformed how companies interact with consumers on the internet, we expect DA to transform how companies design products with and for consumers. An illustrative case study is performed to demonstrate some of the foundations of DA in the redesign of a refrigerator.

There has been a significant amount of research investigating the design task of concept analysis, and much research on reconfigurable system design. Despite previous efforts, further research is still needed that explores how concept analysis should best be conducted for reconfigurable systems. Because reconfigurable systems have multiple configurations and different performance levels, additional information is required to understand each concept, making the design selection process more demanding. Desirable functions, concepts for each phase, and concepts for transition methods between configurations all need to be evaluated. In this paper, the fundamental challenges of reconfigurable system analysis are identified. A framework is then developed to evaluate the many options a designer may face when performing the concept analysis phase of reconfigurable system design, identifying the influence of reconfigurable systems on the process. A mini unmanned aerial vehicle case study is used to demonstrate how the framework is applied, and how existing concept analysis tools can be adapted to account for additional criteria. The paper concludes with a review of the contributions of this work and identifies areas of future work.

In this paper we present an approach for increasing the efficiency of a hybrid Genetic/Sequential Linear Programming algorithm. We introduce two metrics for evaluating the modality of the design space and then use this information to efficiently switch between the Genetic Algorithm and SLP algorithm. The motivation for this study is an effort to reduce the computational expense associated with the use of a Genetic Algorithm by reducing the number of function evaluations needed to find good solutions. In the paper the two metrics used to evaluate the modality of the design space are the variance in fitness of the population of the designs in the Genetic Algorithm and the error associated with fitting a response surface to the designs evaluates by the Genetic Algorithm. The effectiveness of this approach is demonstrated by considering a highly multimodal Genetic Algorithm benchmarking problem.

In this paper, we present a framework for the concept of flexibility in complex system design. This is one of the first of many steps toward developing new design methods for designers that will aid them in development of customizable systems that meet the requirements of multiple customers and multiple tasks. The hope is that this paper will provide both a starting point from which academia and industry can move forward in developing new design methods for flexible systems and a basis for establishing a standard lexicon for use when referring to flexible system design.

Many designers concede that there is typically more than one measure of performance for an artifact. Often, a large system is decomposed into smaller subsystems each having its own set of objectives, constraints, and parameters. The performance of the final design is a function of the performances of the individual subsystems. It then becomes necessary to consider the tradeoffs that occur in a multi-objective design problem. The complete solution to a multi-objective optimization problem is the entire set of non-dominated configurations commonly referred to as the Pareto set. Common methods of generating points along a Pareto frontier involve repeated conversion of multi-objective problems into single objective problems using weights. These methods have been shown to perform poorly when attempting to populate a Pareto frontier. This work presents an efficient means of generating a thorough spread of points along a Pareto frontier using genetic programming.

In this paper, we present an approach for the optimization of a racecar using vehicle dynamics simulation in a parallel-computing environment. The use of vehicle dynamics simulations in the automotive and auto racing industries is widespread. Complex vehicle simulations can include hundreds of parameters and be very computationally expensive to perform. This limits the number of design configurations that can be considered within a reasonable time, preventing thorough exploration of the design space. It also limits the usefulness of these simulations during the course a race weekend when time is of the essence. In this paper, we present the initial results from work to overcome this problem. By using distributed computing, the number of vehicle configurations that can be considered as well as the fidelity of the simulation can increase while keeping the execution time reasonable. The results focus on both the computational savings possible and the improved ability of the designer to make effective decisions due to the huge increase in the amount of design information available. To demonstrate the effectiveness of this approach the optimization of two vehicles will be considered, a two-design variable model and three-design variable model. These results are compared to previous results that were obtained executing the code on a single processor. Good speedup and excellent scalability is observed as the number of processors is increased from 2 to 32.

The ability to make rational decisions is one of mankind’s unique attributes. A characteristic of the formal techniques that have been used for decision making is the selection of the best alternative with respect to a certain figure of merit. One of the most critical problems in engineering design is making early decisions on a sound basis. However, the early stages of design are also the most uncertain, and obtaining precise information upon which to base design decisions is usually impossible. The need for a methodology to represent and manipulate imprecision is greatest in the early, preliminary stages of engineering design, where the designer is most unsure of the final dimensions and shape, material properties, and performance of the completed design. Utility Theory provides an analytical way to aid the decisions in engineering design. By exchanging from objective to attribute and expressing these attitudes mathematically, a utility based attribute function (Utility Function) can be set up to describe the attitude of a decision-maker with regard to his/her preference. In this work, the Aspiration-level Interactive Method (AIM), a goal-seeking method based on identifying non-dominated solutions, is used along with Utility Theory to compensate for the limitations of utility theory in forming a meaningful group preference. This work is an initial attempt to integrate two methodologies from the field of decision theory in order to provide rational decision support for design problems where a hierarchy of decision making is required. The hierarchy, in this paper, is characterized by multiple designers at the lower level who report to one manager. The designers each have different preferences and values, while the manager is driven by project goals and specifications. The approach presented generates feasible and preferred design combinations for further analysis in the detailed design phase. 1 ME Member ME Member, Corresponding Author NOMENCLATURE SUF: Single-Attribute Utility Function MUF: Multi-Attribute Utility Function AIM: Aspiration-level Interactive Method

This paper presents a new approach to solving the orthogonal packing problem (OPP). The OPP is shown to be related to one class of resource allocation problems, namely the land allocation problem. The proposed method differs from traditional methods in that component dimensions are not prescribed, allowing designers greater flexibility and giving them the ability to choose from a number of different solutions. We will show that in removing constraints from dimensions, we allow for the possibility for finding better solutions and this is verified in the case study.

In this paper we introduce a methodology to reduce the effects of uncertainty in the design of a complex engineering system involving multiple decision makers. We focus on the uncertainty that is created when a disciplinary designer or design team must try and predict or model the behavior of other disciplinary subsystems. The design of a complex system is performed by many different designers and teams, each of which only have control over a small portion of the entire system. Modeling the interaction among these decision makers and reducing the uncertainty caused by the lack of global control is the focus of this paper. We use well developed concepts from the field of game theory to describe the interactions taking place, and concepts from robust design to reduce the effects of one decision-maker on another. Response Surface Methodology (RSM) is also used to reduce the complexity of the interaction analysis while preserving behavior of the systems. The design of a passenger aircraft is used to illustrate the approach, and some encouraging results are discussed.

The Massachusetts Institute of Technology (MIT) Commission on Industrial Productivity, in their report Made in America, found that six recurring weaknesses were hampering American manufacturing industries. The two weaknesses most relevant to product development were 1) technological weakness in development and production, and 2) failures in cooperation. The remedies to these weaknesses are considered the essential twin pillars of CE: 1) improved development process, and 2) closer cooperation. In the MIT report, it is recognized that total cooperation among teams in a CE environment is rare in American industry, while the majority of the design research in mathematically modeling CE has assumed total cooperation. In this paper, we present mathematical constructs, based on game theoretic principles, to model degrees of collaboration characterized by full cooperation, approximate cooperation, sequential decision making, and isolated decision making. The design of a pressure vessel is given to illustrate the theory, and the design of a passenger aircraft is given to illustrate the application to a large-scale systems design.

Design models often contain a combination of discrete, integer, and continuous variables. Previously, the Adaptive Linear Programming (ALP) Algorithm, which is based on sequential linearization, has been used to solve design models composed of continuous and Boolean variables. In this paper, we extend the ALP Algorithm using a discrete heuristic based on the analogy of an animal foraging for food. This algorithm for mixed discrete/continuous design problems integrates ALP and the foraging search and is called Foraging-directed Adaptive Linear Programming (FALP). Two design studies are presented to illustrate the effectiveness and behavior of the algorithm.