Engineering design: perspectives, challenges, and recent advances

Abstract

This paper provides a review of papers published in the Journal of Engineering Design from 2007 to 2008. Included are the different perspectives on key challenges existing within, and state-of-the-art technologies developed as solutions to current design issues. The intention of this paper is not to provide detailed coverage of all the papers published over the last two years, but rather to provide an overview of research efforts focusing upon key research streams and to illustrate trends within design research and practice.

Keywords:

• design review
• perspectives
• challenges
• recent advances

1. Introduction

Within the global product supply chain that typifies modern product development, engineering design faces a number of challenges. First, the need remains to attract and retain customers. Second, there is a need to be competitive within the marketplace through maintaining or increasing market share and profitability. Third, there is a need to ensure that the requirements of diverse communities and governments existing within global supply chains are satisfied. To meet these needs, research efforts within engineering design seek to develop research outputs that can effectively address customer requirements, comply with government regulations, and address community issues such as environmental concerns, and strengthen their competitiveness within the marketplace.

A review of the research published within the journal since 2007 shows that these efforts have varied in their adopted perspectives, considering engineering design from customer, designer, and community perspectives. In addition, cross-cutting research efforts have looked at the development of tools, techniques, and methods that can support engineering design from all of these perspectives, such as the development of decision support tools and methods for achieving integration within design. Section 2 presents an overview of such cross-cutting work published within the Journal of Engineering Design (JED) in the last two years. Sections 3–5 provide an overview of the design challenges and advances from the designer, customer, and community perspectives, respectively. Section 6 briefly reviews three Special Issues of the Journal that focus upon cost engineering, computational methods, and global platforms, and, finally, in Section 7 concluding remarks are offered on the overall trends of design research. Figure 1 illustrates the classification structure for recent design research published within JED over the past two years and how they are organised in this paper.

Figure 1. Classification structure of recent design research.

2. Cross-cutting support for design

This section discusses fundamental support required for engineering design irrespective of whether design is taking place from a designer/manufacturer, customer, or community perspective. Recent research efforts have primarily shown advances in the following three areas: decision support, virtual reality (VR) tool support, and integration support.

2.1. Decision support

There has been continuing interest in research on decision support for engineering design decision-making over the last two years. Research efforts have sought to understand and support decisions of not only a strategic nature (Choi et al. 2008, Fu 2008), but also those of a tactical and operational nature. New methodologies and approaches have been developed for traditional areas, such as the selection of product design projects (Wei and Chang 2008), material selection (van Kesteren et al. 2008), resource allocation (Reich and Paz 2008), and risk management (Derelov 2008). In addition, decision support research has expanded into new areas such as durability choice (Saleh 2008a, 2008b). While much of the research has focused on the designer/manufacturer's perspective alone, there have been promising advances in integration of other perspectives, for example, business or customer aspects (Fu 2008, Steele et al. 2008), and environmental considerations (Choi et al. 2008). Furthermore, work has sought to support decision-making that occurs not only during conceptual design (Reich and Paz 2008, Steele et al. 2008), but also during the concept to detailed design bridge (Ruder and Sobek 2007). Overall, a wide variety of work has been performed within the realm of decision support, as detailed in Table 1.

Table 1. Overview of research on support for design decision-making.

Publication	Decision problem area	Design stages/decision levels	Methods/approaches	Key contribution
Choi et al. (2008)	Selecting design for environment strategies	Strategic decision	Integrating lifecycle model and multi-criteria analysis Analytic Hierarchy Process (AHP)	A systematic approach with both qualitative and quantitative analysis enables decision makers to examine strength and weakness of each strategy by comparing appropriate criteria
Derelov (2008)	Potential failure evaluation	Early design stage	Qualitative modelling of weak spots and errors	A generic approach of modelling failure behaviour and identification of potential failures
Fu (2008)	Selecting refuelling strategy for scooters	Strategic decision	Sensitivity analysis of product life-time, fuel price, and user time value	A framework with links between design parameters and performance costs, flexible in response to condition changes
Reich and Paz (2008)	Resource allocation	Early design stage	Mathematical extension of Quality Function Development (QFD)	Resource Quality Function Development (RQFD) model allows redistributing resources to maximise the quality under new circumstances
Ruder and Sobek (2007)	Components, subsystems and their configuration	Concept to detailed design bridge	Experimental groups	The decision tool developed can help users come to a deeper understanding of the system-level design issues earlier
Saleh (2008a)	Rational choice of durability	Decisions at detail design stage	Quantitative analysis of the marginal cost of durability	Metrics and cost analysis from customer's perspective
Saleh (2008b)	Optimal durability choice under uncertainty	Decisions at detail design stage	Quantitative modelling of risk of obsolescence and system net present value (NPV)	Addressed the durability choice problem from the customer's perspective (i.e. an optimal durability is sought to maximise the NPV for the customers)
Steele et al. (2008)	Tennis ball design solution	Preliminary design stage	User-centred experimental study	Techniques to integrate measurable performance properties to customer-perceived performance
van Kesteren et al. (2008)	Material selection	Decisions at early design stage	User-centred experimental study	Identified critical factors as material selection criteria and available information about materials
Wei and Chang (2008)	Selecting design solutions	Decisions at upper-level management	Integrating analytic network process and goal programming	A systematic problem solving methodology that have both multi-criteria and interdependence properties

2.2. VR tool support

VR has attracted wide research interest over the last decade. The use of VR environments can support product design through achieving reductions in development time and increasing customer's satisfaction. Previous VR applications within engineering design have been primarily restricted to use by engineers and designers, for example, for illustrative purposes, and for the study of user behaviour and interaction with products. However, recent publications within JED over the past two years seem to indicate that VR research is expanding.\enlargethispage*{6pt}

In product design, it is important (as with capturing customer needs) to accurately transform customer needs into the actual forms and functions of the products. Further, it is essential for the people involved in the development process to visualise the design as effortlessly as possible and to gain a comprehensive understanding of the functional behaviour of the product. Traditional physical prototyping as well as computer-aided design (CAD) software cannot reflect the functional behaviour of the product. A VR environment was reported to satisfy such requirement through the case of design evaluation of a digital consumer product (Park et al. 2008). The key feature of the VR environment is that it combines VR-based interaction with functional behaviour simulation. The VR tool allows users to conduct functional evaluation and usability test before engaging in the costly and time-consuming process of building physical prototypes.

VR tools have also been applied within the manufacturing domain. Dukic et al. (2007) detailed a case study focusing on the verification of visual demands in car assembly work using virtual tools. Computer mannequins were created for the analysis of the ergonomics of assembly operations. This use of VR within the product development process allowed the early identification of ergonomic problems using only virtual mock-ups. Cappelli et al. (2007) detailed a virtual environment for disassembly that could accept a virtual CAD assembly prototype as an input and produce necessary information for the identification of the disassembly path.

In the above research VR was used as a local tool to aid product development. To render a VR environment on the Internet, Ottosson and Holmdahl (2007) detailed an approach that combined a content management system and transformed VR files, so that the VR application could be used as an ordinary web application. The major contribution of this research is that by transformation of VR to a web-based tool, it allows more project stakeholders, not just engineers and designers, to get involved in the decision-making process of a new product design, a new production line, or simply the purchase of a product.

Virtual 3D representations of geometric models are in fact 2D presentations that attempt to mimic 3D appearance through the use of different visual cues. Consequently, the risk exists that what people see in a VR 3D representation may not represent an accurate representation when compared to physical reality, depending upon the visual quality appearance (VQA) of the virtual 3D representation. The research described by Wickman and Soderberg (2007) focused upon the evaluation of the VQA of industrial design concepts. In particular, Wickman and Soderberg concentrated on the issues of gap and flush, which could affect people's judgement within VR environments.

2.3. Integration support

Much literature has highlighted the need for holistic design. Integrated design has been emerging as a holistic design methodology that can achieve better design performance. Recent research on integrated design includes system integration, requirements integration, knowledge integration, and method and process integration.

System integration has been a key area of integrated design because one of its main motivations is the concept of “whole system” design, that is, to view a design target as a “system” interdependently as opposed to its separate components. Moscoso's (2007) discussed the integration of social and technical systems within a design-oriented framework for modelling production management systems. In essence, the integrated design approach does not artificially separate the production management system from the shop-floor it manages, but rather considers the two as intertwined and complementary elements. Generally, systems can be linked together through information, material, energy, etc. However, system integration should not be understood as leading to a straightforward design solution. Integration can actually complicate system design in many ways. Two good examples are robustness and durability. An integrated system is only considered as robust when its functional performance is intensive to selected design parameters and environmental changes, while it is still satisfying design and customer requirements. To improve the integrated system robustness, Zakarian et al. (2007) explored integration from a system configuration viewpoint to minimise system-to-system interactions and overall system sensitivity to noise factors. With regard to integrated system durability, Saleh (2008a) described how a cost profile could be determined as a function of durability through quantifying the impacts of durability on individual systems within a spacecraft and then integrating those impacts to provide the cost profile for the complete spacecraft.

Requirements integration in product design considers the “horizontal” integration of requirements from different sources and the vertical integration of requirements through different design stages. The problem of modelling and mapping of “difficult to quantify” customer requirements to technical (functional) requirements and subsequently to design parameters (i.e. solutions) is discussed by Guenov (2008). Integration between customer's technical requirements is achieved through the consideration of co-relations that are modelled in quantitative models. This work bridges an important gap between engineering design and marketing, as it incorporates the definition and the seamless decomposition and integration of customer and functional requirements to design parameters. Recently, environmental requirements have been of growing interest within engineering design as a result of public pressure and stricter legal regulations. Choi et al. (2008) developed a framework for the integration of environmental and business requirements toward sustainable product development.

Engineering design is a knowledge-intensive process and as a result of this knowledge integration has been regarded as crucial by designers. Indeed much research effort has been devoted to developing different knowledge integration strategies, such as knowledge mapping, sharing, and access. However, given that design knowledge can assume many forms, knowledge integration remains an ongoing challenge. Condoor and Kroll (2008) discussed the integration of one specific type of knowledge (that of knowledge embodied in design principles) and proposed a formal methodology for integrating design principles within the design process to ensure that appropriate principles are applied when required. Thus, it can support, for example, the identification of appropriate design principles for novice designers who may not be aware of their existence.

Methods are often used to assist individual tasks within the design process. Integration of such methods provides the potential to increase overall design performance across a series of design activities. Wei and Chang (2008) discussed method integration for an optimal product design solution and proposed a systematic methodology for integrating three methods: fuzzy Delphi, analytic network process (ANP), and zero-one goal programming (ZOGP) methods. In this case, the fuzzy Delphi method was used to identify essential selection criteria for design solutions, the ANP method was used to determine the weights of these criteria and proposed design alternatives. Finally, the ZOGP model was used to perform multi-objective evaluation in order to reach an optimal solution. The integration of these three methods provides a combined qualitative and quantitative analysis against single or multi-criteria for the design solution selection. Haussler and Albers (2007) focused upon the integration of a set of numerical methods for shape optimisation in dynamic mechanical systems. Facilitated by the integration of methods for performing finite element analysis, hybrid multi-body system simulation, fatigue analysis and shape optimisation, an integrated structural optimisation process could be fully automated. However, a feature of method integration is that it normally requires highly skilled users and involves the management of large amounts of data.

3. Designer perspective of design

For a manufacturing concern, to maintain and increase their market share and profitability, product designers need to adopt new technology that supports them to produce high-quality products in a time- and cost-effective manner. This section discusses recent advances in robust design, design optimisation, and design cognition and how this research can enhance their competitiveness.

3.1. Robust design

Robust design is concerned with the development of methods intended to make a product's function more consistent in the face of variations in downstream processes, environments, and customer use patterns (Jugulum and Frey 2007).

Herrmann (2007), Condoor and Kroll (2008), and Xue et al. (2008) developed models that can improve robustness through consideration of design parameters. Xue et al. (2008) described a Taguchi method-based model to minimise the impact of design parameter changes on implementation processes by considering possible design parameter changes, and the design parameter changes associated with even or uneven probabilities. Condoor and Kroll (2008) recommended the use of design parameters and design principles (together with ideas) to quantitatively evaluate and improve the robustness of design solutions during the conceptual design phase.

One of the key challenges of robust design is evaluating robustness in the face of the dynamic nature of design. Yassine (2007) defined robustness as “the ability of a process to absorb design changes” and proposed using a design structure matrix methodology to model and simulate the performance of design processes in terms of robustness. Zakarian et al. (2007) focused on developing system robustness, rather than individual product robustness. Their methodology consisted of three components: system modelling, integration analysis, and quality engineering. Implementing the methodology can prove problematic, however, for example, lack of system engineering expertise in companies can limit the ability to develop appropriate system models.

Earlier work on robust design can be found in the journal's 2005/2006 review (Sheldon 2006).

3.2. Design optimisation

Optimisation has been a focus of design research for many years and has mostly focused on evaluation of design parameters. Such research has assumed that objective functions can be expressed algebraically and that it is possible to differentiate the functions that assist algorithms that try to perform gradient descent. However, recent research efforts have extended to designs of such complexity that their effectiveness can only be evaluated through computational methods and new processes.

Computational methods have proved to be highly efficient and effective in shape optimisation (Horvath et al. 2007), which is used to improve a given shape by means of surface/or feature modification. For parts in dynamic mechanical systems, shape optimisation is a complex task because in many cases it is not possible to deduce a limited set of load cases from load-time histories, and traditional optimisation approaches do not take fatigue into account. Haussler and Albers (2007) developed an integrated structural optimisation model targeted at addressing dynamic fatigue design problems. In addition, models have also been developed to enable the coupling of optimisation operation and dynamic systems. Vajna et al. (2007) describe a case study on the application of evolutionary methodologies to shape optimisation of a bow riser. The key idea of the evolutionary methodology is that good properties of preceding selections are passed onto succeeding solutions. The implementation of the evolutionary methodology consists of a mixed process of searching, adoption of existing knowledge, learning, selecting, and combining. By using the evolutionary methodology, the mass of the bow riser was reduced by 22% while the stiffness remained unchanged. The Taguchi method has been widely used within the design fraternity to optimise product design. Herrmann (2007) employed the Taguchi method as a testing methodology to ensure the proper optimisation of the design of a printer registration system. Huang et al. (2008) investigated a genetic algorithm-based optimisation method for product family design. Their research considered multi-objective commonality in product family design in the sense that the feature or component could be common only among some product variants; therefore, the method is suitable for addressing multi-optimisation problems. Also targeting multi-objective optimisation problems, Zhao et al. (2007) adopted a data-driven design optimisation methodology, combining simulation and experimental data to generate surrogate models for the objective functions. The method was applied to the design of a cooling system for electronic components.

Cappelli et al. (2007) focused upon optimising processes associated with a product rather than its physical design. Particularly, they developed a method for optimising the disassembly sequence of mechanical systems in which disassembly sequences are automatically generated from a product CAD model associated with an index and evaluated (e.g. in terms of cost, time, number of operations, etc.) to identify the optimal sequence.

Although the objective of optimisation can be summarised simply as to maximise or minimise objective functions across a set of parameters, realising this is a considerable challenge. Difficulties can stem from the objective functions (in particular their non-linear nature and the high degree of noise they can contain) and the design parameters (which can be continuous or discrete depending on the nature of the underlying problem). Given that no single optimisation method is likely to work for all problems, it remains likely that design optimisation will remain a key research area in the future.

3.3. Design cognition

In recent years the study of design cognition, that is, understanding the rationale behind design problems and how they are solved, has steadily increased. Existing research covers a wide range of topics in the area such as cognitive processes, the cognitive behaviour of designers, research techniques used to capture such behaviours, and examples of how these techniques have been implemented. Coley et al.’s (2007) review of existing work on the cognitive behaviour of designers revealed that a considerable amount of research had begun to identify the role of cognition in design, such as that described by Houseman et al. (2008), but that the tools and techniques being used to produce this valuable research (such as “thinking aloud” and protocol analysis) were highly subjective. When applying the thinking aloud technique to identify the cognitive actions of cost estimators, Coley et al. noted that its reliability was questionable given its dependence upon the effectiveness of the participant's verbalisation and their willingness to participate in the cognitive analysis activity. In terms of progressing work in this area, it would therefore seem prudent to focus on developing techniques being used to capture the cognitive behaviours of designers. However, Coley et al. do illustrate the benefits of such research, for example, comparing the cognitive actions existing in different domains can facilitate the identification of necessary training requirements that can subsequently help achieve workforce flexibility.

4. Customer-centred design

Customer-oriented design seeks to address the issue of how to satisfy customers from two dimensions. One is to make sure that customer needs and requirements are properly captured and realised through product functionality. The other is to ensure customers’ safety, health, and comfort when they use the products.

4.1. Requirements management

Within axiomatic design (AD), the customer domain is the first of four that make up the world of design (the remaining three are the functional, physical, and process domains) (Gumus et al. 2008). Customer requirements are, therefore, the origin of what needs to be achieved and the process of mapping to the other domains represent how these needs are achieved. Thus, requirements management includes both the initial capture and evaluation of customer requirements, and the transformation of these requirements into functional requirements that then guide the development of the design solutions within the physical domain. A number of recent research efforts have focused upon this area of requirements management.

Bjornfot and Stehn (2007) explored the use of a design structural matrix (DSM) as a holistic approach for systematic consideration of customer requirements through a case study in the construction domain. The DSM was found to be effective through its ability both to provide standardised system and detailed element views, and to illustrate the physical and functional interactions among the elements. Yan et al. (2007) developed an effective measure for evaluating demographic customer characteristics and detecting demographic customer requirements differences in product conceptualisation.

AD theory also indicates that it is important to understand that design is an interplay between what we want to achieve (based on customer requirements) and how we achieve it. Thus, captured and evaluated customer requirements need to be transformed into product functional requirements and it is these functional requirements that guide development within the physical domain. Guenov (2008) developed a synthetic procedure, named structural equation models (SEM), for the design of engineering systems. The SEM incorporates a confirmatory and a structural component. The former is used for the decomposition and mapping of qualitative customer requirements, modelled as latent variables, onto a generally large number of measurable technical requirements. The structural component then maps the technical requirements to design parameters. However, in the case study presented the model was restricted to a consideration of linear dependencies between the variables. In general, the approach can handle a number of non-linear relations through variable transformation. The SEM represents a sufficiently rich and generic structure capable of bridging the gap between the customer, functional, and physical domains of AD.

4.2. Ergonomic design

In engineering design, ergonomics is concerned with designing products according to human (including customers and users) needs in order to optimise their well-being and overall system performance (Dukic et al. 2007). Ergonomic considerations have been recognised as an important aspect ensuring that designed products satisfy customer requirements from ease of use and safety viewpoints.

Liu et al. (2008) discussed ergonomic design within the development of helmets, focusing upon ensuring that they were comfortable to wear for extended periods of time. Through their case study, Liu et al. demonstrated that developing a helmet fitting design through consideration of head shape could significantly reduce the weight and enhance the stability of a helmet, thus making it more comfortable to wear. Traditionally, ergonomic design has not been considered in traditional helmet design because of the rigid design and manufacturing process that was unsuitable for incorporating individually customised helmet shapes. Liu et al. developed a rapid preliminary design method (and a software implementation of it) for designing the helmet shell. The software uses 3D anthropometric head scans to generate head models that are representative of the intended user group. Thereafter a semi-parametric surface modelling tool is applied to generate quickly the helmet shell through the input of a number of parameters relating to helmet protection, size, and shape requirements and the adjustment of key curves within the design. Subsequently, an integrated helmet head model makes it possible to evaluate a design against ergonomic considerations. This method, therefore, supports the customised design of products and can result in higher customer satisfaction through increased wear comfort and protection.

Ergonomic design can also play a role in reducing the possibility of user injury through product use. Numerous CAD mannequin models exist that are based upon an articulated human representation. During design, such models can be manipulated and positioned to assess design options ergonomically. Such models need to incorporate a wide range of motions to support ergonomic design effectively and typically different postures have to be created individually through the separate positioning of joints and model animation needs to be defined a priori for each design modification. Consequently, design success is highly dependent on the design engineer's judgement and knowledge of relevant postures, and their ability to incorporate sound ergonomic principles into their designs. Mitchell et al. (2007) introduced the use of constraint-based mannequin models to effectively predict movement patterns for a range of potential designs, which eliminated the need to manipulate CAD mannequins in response to design environment changes and reduced uncertainty associated with mannequin manipulation. Application of their models to machine tools users demonstrated that they could reduce a machine operator's likelihood of injury through successfully predicting movement patterns across a wide range of conditions.

5. Eco-design: community perspective

Serious consideration of environmental requirements has only become an important feature of design in the last decade or so as a result of growing pressure from the public and stricter government regulations. Such requirements are fundamentally impacting the way new products are designed and launched. An early example of this impact has been the development of new methods for evaluating pollution emissions within the early stages of design. Such approaches form the focus of eco-design, and seek to ensure that products address environmental requirements in addition to those related to technical and economic consideration. This section provides an overview of two eco-design methods: design for environment (DfE) and design for disassembly (DfD).

5.1. Design for environment

DfE involves the systematic consideration of design performance with respect to environment, health, and safety objectives over the full product and process lifecycle. DfE involves design procedures that aim to minimise material and energy consumption while maximising the possibility for reuse and recycling. During the last decade many companies have established their own DfE programmes to address environmental concerns in their product development process.

DfE practices reveal that conformance to environmental regulations often depends on whether a stakeholder within a DfE programme can cooperate effectively with other key eco-conscious design stakeholders (such as design groups, suppliers, etc.). Ge and Wang (2007) developed a two stage systematic approach to help ensure the success of these DfE programmes. Stage 1 (problem formulation) is concerned with characterising corporate strategy, the DfE programmes, and the stakeholders within it to form an assessment basis for applying a systematic evaluation method. The evaluation method (Stage 2) quantitatively evaluates the effectiveness of corporate strategies in terms of the resulting environmental impact, product quality, and cost performance. The approach allows the evaluation and comparison of the effectiveness of different corporate strategies and the selection of a sustainable one that allows the eco-conscious design products. Further work is needed, however, to automate the evaluation process through software development and integration with a design methodology. More recently, Choi et al. (2008) developed a framework for integrating environmental and business aspects within sustainable product development. The environmental aspect of the product is captured by lifecycle assessment, and the result is directly introduced to the selection of DfE strategies prior to multi-criteria decision-making.

5.2. Design for disassembly

DfD is concerned with improving the effectiveness of a product's disassembly throughout its life cycle in order to reduce maintenance costs or to facilitate cost-effective reuse, remanufacturing, and recycling. Disassembly is a sequence of actions by which a product is disassembled, and the main goal of DfD is the exact choice of the “optimal sequence”. Cappelli et al. (2007) developed a virtual disassembly environment that included a methodology for identifying the optimal disassembly sequence. This methodology provides a theoretical basis for the creation of a tool that assists the designer during the virtual design phase.

The operational life of a product is extended through its maintenance programme, while the reuse and remanufacture operations can extend the operational life of some parts of the product. Remanufacturing methods provide manufacturers with the ability to comply with current and forthcoming environmental legislations. Alongside the optimisation of the remanufacturing processes, it is important to understand the properties of products that increase the value of remanufacturing and reuse, and to develop methodologies that support the design of sustainable products. Thus, requirements for remanufacturing should be addressed throughout the design of a product, and should focus on the ease with which a product at the end of its life can be disassembled, cleaned, tested, and reused. Zwolinski and Brissaud (2008) discussed remanufacturing strategies to support product design and redesign. Their study, based upon an analysis of a wide range of remanufactured products, identified a number of factors affecting the success of a remanufacturing operation and culminated in the development of 11 remanufacturable product profiles (RPPs) and their use to develop easily remanufacturable products. Extensions of this work will be to focus on environmental aspects linked to the RPPs in order to help designers integrate environmental criteria into their decision-making process.

6. Special Issues

Over the past two years, there have been three Special Issues within JED that have focused on research areas of particular interest. These Special Issues have highlighted current research on cost engineering, computational methods, and product platforming. This section gives only a brief overview of these three issues. More detailed information on each issue is available from their respective editorial summaries (Horvath et al. 2007, Roy 2008, Thevenot et al. 2008).

6.1. Cost engineering

Cost is one of the most fundamental criteria applied during the evaluation of design proposals. Cost engineering is concerned with cost estimation, cost control, business planning and management, profitability analysis, cost risk analysis, and project management, planning, and scheduling (Roy 2008). The Special Issue on cost engineering covered a wide range of topics that included modelling of whole-life cost prediction (Newnes et al. 2008), a framework for cost-based producibility assessment (Elgh and Cederfeldt 2008), cognitive actions between experienced designers and experienced cost engineers (Houseman et al. 2008), incorporating cost analysis tools within a design and engineering architecture (van der Laan and van Tooren 2008), and application of cost engineering methods for product development (Dimitrellou et al. 2008, Mauchand et al. 2008). Despite new advances in this area, more attention is needed to understand the uncertainties and cost risks involved in cost estimation (Roy 2008).

6.2. Computational methods

Intensive research has been conducted on the development of efficient computer-oriented methods and tools able to support conceptualisation, detailing, and optimisation of shapes from the point of view of both industrial and engineering designers. The Special Issue on computational methods to support sketching, reverse engineering, and optimisation of shapes focused on three key issues: shape sketching, reconstruction, and optimisation. Cheutet et al. (2007) and Mengoni et al. (2007) sought to support shape sketching through semantic-based operators and reverse engineering of heterogeneous design representations, respectively. Within shape construction, Ruiz et al. (2007) focused on supporting curve reconstruction, Pernot et al. (2007) on mesh reconstruction, Vukasinovic et al. (2007) on surface reconstruction through fast reverse engi-neering of CAD files, and Langerak and Vergeest (2007) on free form feature recognition. Two papers on shape optimisation discussed the optimisation of parts in dynamic mechanical systems with regard to fatigue (Haussler and Albers 2007), and the optimisation of a bow riser using autogenetic design theory (Vajna et al. 2007). A more detailed summary of the papers published in this Special Issue can be found in (Horvath et al. 2007).

6.3. Global platforms

In today's highly competitive global marketplace, companies use product families as a means of customising their products to achieve increased product variety, reduced costs, and shortened lead times. Product platforming has been adopted to facilitate more effective product family design. This Special Issue on product platforming for the global marketplace focused on recent research seeking to address the following key issues within product family design and platform-based product development:

• Managing trade-offs between product family commonality and performance of individual product variants (Thevenot and Simposon 2007, Huang et al. 2008, Ye and Gershenson 2008).

• Handling customer needs and behaviour, and technical requirements (Zhang et al. 2007, Gupta and Okudan 2008, Park et al. 2008).

• Developing platforming processes (Gunzenhauser and Bongulielmi 2008).

• Consolidating existing product families (Dolan and Lewis 2008).

A more detailed summary of the papers published in this Special Issue can be found in (Thevenot et al. 2008). Despite diligent efforts within the field, the editors acknowledge that further research is required to effectively support product family design and product platforming.

7. Conclusions

This paper presents a review of new advances in engineering design primarily from designer, customer, and community perspectives. Comparing this review with its 2005/2006 predecessor, a number of topics such as robust design, design optimisation, computational methods, and tool support, continue to receive attention. Recent work, however, illustrates that research interests are also shifting to new topics, including eco-design, ergonomic design, cognitive design, requirements management, decision support, and integration support. The development of research in these areas indicates that engineering design is expanding to consider a greater number of factors, for example, environmental concerns, that influence both the design process and the product being designed.