Theories and Models of Design: A Summary of Findings

Abstract

This chapter introduces the goals of the book and provides a historic overview of theoretical developments in design. The main focus of the chapter is an attempt to answer the three main questions addressed in this book: (1) What is a theory or model of design? What is its purpose: what should it describe, explain or predict? (2) What are the criteria it must satisfy to be considered a design theory or model? (3) How should a theory or model of design be evaluated or validated?. The answers are derived from the contributions of the various authors in this book and from the results of the International Workshop on Models and Theories of Design that gave rise to this book. Taken together, the contributions and the workshop outcomes showcase the rich but varied tapestry of thoughts, concepts and results. At the same time, they highlight the effort still required to establish a sound, generally accepted theoretical and empirical basis for further research into design.

Appendices

Appendix A: Summary of Discussions from the International Workshop on Models and Theories of Design

Discussions in the workshop, carried out primarily in three, parallel breakout sessions that continued through the days of the workshop, and culminated in a subsequent, common, final discussion session on the last day, focused on the four questions discussed below. This appendix provides a summary of the outcomes from these discussion sessions, which, we hope, will add to the richness of the knowledge already encapsulated in the individual chapters. As will be seen, while it is far from being conclusive, some major similarities in (lack of) understanding about theories and models, their purposes and criteria, and as to how they should be validated, have already began to emerge, and a number of common directions for further activity in this area have been proposed.

1. What is the distinction between a theory and a model?

Team 1: Rapporteur: John Gero; Scribe: Sonal Keshwani. The team took a broad approach of decomposition, and looked at the elements that constituted a model. A model was taken as a representation (i.e. away in which a language is used to describe something) of some observable phenomena. It had been noted that some phenomena may not be observable, and observation of phenomena may sometimes change the phenomena themselves. It was noted that the point of view of the observer plays an important role in what will be observed and how it will be interpreted: ‘what you come up with is always limited by how you see the world and your output is evaluated by how the world looks at it’. All representations, it was felt, are limited, ideally by the purpose of the representation; hence, all models are also purposively limited. Models have generality and causality. Models project or predict, and can be used to explain. The team defined a theory to be an abstract representation of a generalisation of phenomena; a theory may have axioms that explain how a world behaves. Three views on the distinction between a model and a theory emerged: (i) a theory may be composed of multiple models; (ii) a model may be more concrete and specialised in its context than a theory, which is more abstract and general; (iii) a model may embed explanation of phenomena, while a theory may allow for such explanation. A theory may be represented by different models. There may be theory-driven and phenomena-driven models.

Overall, the team summarised its findings as follows. A model is a representation of some phenomena and relationships among these phenomena. With features that are operationalisable, a model provides some generality with respect to the phenomena, which can be causal, speculative and dynamic, and independent from theory. A theory is an abstract generalisation of phenomena, which can be modelled in multiple ways. Models, but not theories, can change with time. Phenomena are things that have regularity and are directly or indirectly observable, and are interpretable. A representation is an externalisation of a description of phenomena. Any representation leads to a reduction in some aspects of the phenomena and its granularity. What is represented is limited by the purpose or intention of the representation.

Team 2: Rapporteur: Udo Lindemann; Scribe: S Harivardhini, Praveen Uchil. The team distinguished between two types of models: research-based (driven by truth) and practice-based (driven by utility). The team raised the question: should models and theories in design be able to explain only (as in natural sciences) or should they also be useful, since the purpose of design research is to improve knowledge to improve design practice? The team also discussed what constituted goodness of a model, and argued that the goodness of a model depends on understanding of its system boundary, i.e. the context and purpose of the model. The team felt that there is an overlap in meaning between models and theories. A model may simulate a part of the world, but does not necessarily explain it. A model could be a subset of a theory, in that a theory provides explanation at a higher level than a model does.

Team 3: Rapporteur: Lauri Koskela; Scribe: Boris Eisenbart. The team distinguished between two types of models: models of design (i.e. of outcomes of design activity), and models of designing (of design activity). The latter is often used synonymously to theories of design. The team distinguished between a model and theory in the following. A model is an abstraction of reality created for a specific purpose, and the purpose includes representation of a theory; a model is helpful: it may serve multiple purposes and may be applied in multiple ways. A theory, on the other hand, may involve a number of hypotheses, each of which should be possible to be falsified. They recognised that describing something as a theory is sometimes a cultural issue; for instance, in some fields of research, less comprehensive approaches, frameworks etc. are called theories for the only reason that the term ‘theory’ added some kind of value to the proposition. The team recognised that while taxonomies are typically not considered theories in natural sciences, design research should consider theories as a spectrum with various levels of maturity in its context and purpose of use.

Overall, the team felt that a model and a theory have several aspects in common: both models and theories serve a (set of) specific purpose(s) that are useful for researchers and/or practitioners; both are explanatory in character which facilitates prediction and prescription. A goal of theories that is distinct from those of models is to provide an explanation of what design and designing mean within the context of use of the theory.

2. What is a model or a theory expected to describe, explain or predict? What criteria must it satisfy?

Team 1: Rapporteur: John Gero; Scribe: Sonal Keshwani. The purpose of a model is to transform something (e.g. produce an output given an input, which can form a prediction), to explain something. Explanatory power of the model comes from the result produced when using the model. A theory is a set of beliefs that are proposed as a generalisation of some phenomena, which are intended to give an explanation for the phenomena. Models have to be useful; theories have to be falsifiable. A model may help in prediction or exploration. A theory has to be testable/refutable. A model has to be usable in design, if this is a model for design. A theory cannot be evaluated directly, but can be evaluated only after its implementation. Theories contain rules and principles which together form their explanatory framework; this characteristic (i.e. of being constituted of rules and principles) is one of the criteria that a theory should satisfy.

Team 2: Rapporteur: Udo Lindemann; Scribe: S Harivardhini, Praveen Uchil. The team argued that a major distinction in the nature of phenomena dealt with between natural sciences and design research is that, design research focuses on design processes that are unique and operate within incomplete information and uncertainty. It is important to distinguish between different models in terms of their system boundary (i.e. scope of application) and their purpose. The purpose can be truth (in research) or utility (in practice). For a model to be good for truth, it should be true at least with the scope of its application. Goodness criteria for models for utility include: usability, ease of use, how quickly it can be used, system boundary, and limits of the model. Many theories and models are not used well in practice because it is hard for practitioners to understand the terms used in these theories and models. A theory or a model should be able to provide insight. A theory must be falsifiable.

Team 3: Rapporteur: Lauri Koskela; Scribe: Boris Eisenbart. The team felt that theories need to be useful: they can be curiosity-driven where the goal is to understand the nature and characteristics of objects, entities and their relationships, or problem-driven where the goal is to support practitioners and provide utility, or to support education. Understanding is necessary for predicting an outcome, and eventually prescribing how to perform design to achieve an expected outcome. Theories in design may be more probability-driven rather than being strictly causal, given the large number of influences, and may take the form of narratives rather than strict propositions. The team asked for whom theories are to be developed, and felt that these would be primarily for researchers or managers. The team discussed what phenomena a theory should address. While it noticed there may not be a single phenomenon of designing, there might be something fundamental to designing that every designer or design team does or shares, e.g. similar activities, aspects etc. appear across different design projects and disciplines. Overall, it was agreed that there are similarities and differences across designing in different contexts, and a theory of design should explain both similarities and differences across the contexts. It was strongly felt that ‘We do not have a thorough understanding of all the assertions we make about designing. We ought to have theories about how to differentiate between different types of design’.

The team felt that phenomena of designing essentially refer to ‘how design works’; various aspects (e.g. people, process, product, knowledge etc.) play a role in this, and therefore, designing may look very different as these aspects change. There are also many partial activities within designing (e.g. the work of an FEM engineer), i.e. there is ‘designing within designing’, which theories currently do not capture. Design processes are seen as a major aspect, and therefore, need to be comprehensively understood. Since human reasoning is an essential part of the phenomena of design, and since there is a variety of different kinds of reasoning that exist in design (e.g. logical, informal etc.), a theory should account for these differences and their influences.

Overall, the team argued that the criteria which a theory should satisfy is its amenability to validation and testing, where correspondence between what can be concluded from the theory and the phenomena it tries to explain are assessed. Another criterion is that a theory helps prediction which is useful; this can also be in the form of justification in a historical context. Theories are evolutionary rather than stationary. All assumptions underlying a theory should be made explicit, and one should be aware, as a researcher, about the process by which is a theory is developed.

3. How should a theory or model be evaluated or validated?

Team 1: Rapporteur: John Gero; Scribe: Sonal Keshwani. The team felt that all theories have to be falsifiable. The team defined evaluation as assessment of usefulness, and validation as assessment of consistency. It noted that a model that has so far always given correct results can still give incorrect results: theories are never tested to be true, but with more evidence, confidence in the theory grows. A model has to be validated (checked for internal consistencies) followed by evaluation (checked for usefulness). A difference between models and theories is that, ‘hypotheses are derived from theories, while hypotheses are derived from application of models’. A causal model is a network of hypotheses. In evaluating, one has to test each of these hypotheses. To evaluate a theory, one has to operationalise its hypotheses and test these.

Two aspects are critical to pay attention to, when discussing validation: the first is, what should be taken as true and false, and what the process of refutation is whereby truth and falsity should be adjudged. According to this team, validation involves application of the theory or model in design, checking for their internal and external consistencies, and checking them against other, already validated theories or models.

Team 2: Rapporteur: Udo Lindemann; Scribe: S Harivardhini, Praveen Uchil. Validation, the team argues, is about finding the limits of a theory. A major difficulty in validating theories and models of design is that, unlike much of natural sciences, being able to carry out repeatable experiments is hard to impossible. The team proposes that one way of validating a model or theory would be in terms of the level of reliability of the model or theory to achieve its purpose. The team proposed several ways of validation e.g. by comparative studies, by comparing and reducing gaps between research and practice models, by comparing multiple practice based models, or by referring to an existing theory which is already validated.

Team 3: Rapporteur: Lauri Koskela; Scribe: Boris Eisenbart. No design is ever repeatable; however for many areas of natural sciences too. There are various levels of variation across so called repeatable phenomena (e.g. the breaking stress of no two samples of the same material is exactly the same, the effect of the same medicine on no two people is exactly the same, etc.). If the discipline looks into a vast number of design projects in various fields, it might find the phenomena at some level of repeatability (as both material science and medical science already do by taking a statistically large set of samples or subjects). However, two distinct challenges for our discipline are: (i) comparable data in our discipline is currently missing, and (ii) such data is hard to generate. For instance, designers may not be aware of what they do during designing, or may distort certain aspects of their work (e.g. to hide failure, due to miscommunication, post facto rationalisation, forgetting, etc.).

A major issue in validation is that, while some researchers develop theories and others develop empirical results, the two rarely discuss their results with one another to bootstrap their work. A platform to support such discussion is necessary. Another issue is that, many empirical studies are carried out with students only; as a consequence, what can be learnt from these about design in practice is relatively limited. In these studies, and even more so for studies of practice, sample sizes are small due to lack of availability of subjects and constraints on time for detailed analyses. There is a strong need for developing appropriate design research methods to tackle these issues. Another issue is the lack of information of the contexts in which a theory of design is applicable. Given the complexity and variety of designing, it may be too ambitious to develop one theory of design; the community needs to develop many theories, each of which applies in a particular context for a particular purpose. These may then form the basis for developing more comprehensive theories. Another challenge is the difficulty of validating prescriptive theories in practice, e.g. asking practicing designers to change their thinking or process of designing may be hard. Validation need not be done only via practice, but also via teaching, training budding designers into preferred ways of thinking and processes of designing. A possible, new direction for validating theories is theory-driven prediction of new, hitherto non-existing, types of design or design fields.

Overall conclusions about these three questions

Regarding the definition of models and theories, two main points emerged. One is that the term ‘model’ has multiple meanings. In one meaning, models are used as a means to carry out design, e.g. a digital model of the product; we may call these models for design. In the other meaning, models describe, explain or predict how designs and designing are, and how aspects of these are related to various criteria that are of importance to practice, e.g. how designing relate to costs of designs. We may call these models of design.

The second point is that there is considerable overlap between the meanings of models of design and theories of design. A spectrum of meanings emerged, starting from having ‘no distinction in how these terms are currently used in our area’, to one where ‘Theory defines a framework from which multiple models could be derived’. A consensus emerged that there is need to understand ‘theory as a spectrum’, with terms such as taxonomies, models and theories having varying degrees of maturity in context, purpose and explanatory capacity.

The purpose of the need for understanding these terms was also discussed. It was felt that for a practitioner, it made no difference as to what these terms meant. However, for a discipline of research such as design, understanding of these terms is crucial, since this forms the basis for research. Overall, it was agreed that a clear understanding of the terms model and theory in the context of design research is necessary. It is also felt that any proposal for a model or theory should be accompanied with its purpose and context.

A strong consensus arrived at across the teams is in the criteria to be considered a theory: theories should be testable and refutable (i.e. falsifiable), and this should be possible to be carried out within the context and purpose of the theories, i.e. where it applies, and how well.

Validation was seen to be testing the limits of a theory or a model. Validation, too, emerged to have a spectrum of meanings, from testing for internal consistency, to truth and usefulness, in terms of providing explanation or insight in the form of predictions or post-dictions.

Several challenges to validation were identified: difficulty or lack of repeatability of phenomena, the large number of factors blurring clear and identifiably strong influences, difficulty of finding statistically large number of appropriate subjects or cases, and difficulty of generating reliable data about the phenomena under investigation.

4. What are Gaps in our Current Understanding and What are the Directions for Further Research?

Several directions emerged.

One major issue identified in the discussions is the general lack of a common understanding that can act as the underlying basis for the discipline. One symptom or a possible cause of this lack is the poor citing of each other’s work in the discipline. A need for an overview, or even consolidation, of research carried out so far was strongly emphasised. As a discipline, we need good ‘demarcating theories’ that provide a clearer understanding of what constitutes (and what does not constitute) part of the phenomena of designing (e.g. designing is demarcated by intentionality), the different types of designs and designing that form our discipline; and position the models and theories with respect to these.

This base, it was suggested, might be initiated by including these:

• The philosophies of the discipline, including what design means, and what the ‘phenomena of designing constitute’. ‘We need a philosophy of design, like a philosophy of science’.

• A list of ‘demarcating theories’ that provide an understanding of the different types of designs and designing that form our discipline.

• A list of terms that are used within the discipline, including theory and model, along with their contexts and purpose.

• A list of research methodologies and methods within the discipline, along with their contexts and purpose.

• A list of empirical results, along with their context and purpose.

• A list of models and theories of design, along with their context and purpose.

• A list of influences of results of design research on practice.

Another major point was the need to clarify the common purpose of design research, and identify what the pressing, concrete questions are that the discipline needs to address. Also emphasised was the need for investigating the specific characteristics, benefits and complementarities across the various theories and models, rather than discussing only about which one among these.

A further major point was the challenge of validating theories of models of phenomena of design, which pointed to the need to develop research methods that are appropriate for scientific studies within the constraints and expectations of design research: how to develop and validate testable, refutable theories and models of adequate accuracy within the constraints of complexity of the phenomena observed and within the low availability of appropriate cases and subjects?

Towards addressing the above directions, several suggestions were made:

• Have more discussion events at various levels, e.g. students, researchers, educators, etc., to discuss these issues. Getting together is the first step to ‘form the discipline’. Developers of theories and empirical results should interact more with one another.

• Like in other disciplines, teach the common understanding to those (intending to be) in this discipline. This knowledge should be taught in a context-specific manner, i.e. ‘make explicit what is applicable in which specific situation’.

• Interact with other disciplines with similar goals, such as management, and learn from their perspectives.

• Carry out more empirical studies that are unbiased, of high value, high-quality, and are clearly explained, as we still do not understand in sufficient depth why design processes happen the way they do.

• Have ‘grand debates’ where specific models are discussed and contrasted together.

• Work more on developing research methods that are appropriate for serving the specific needs of design research. A starting point can be to propose Special Interest Groups (SIG) to work on these, e.g. on research methodology.

Appendix B: Major Theories and Models not Contained in this Book

This appendix provides a summary of some of the major theories not contained in this book, but are necessary to point to for the sake of completeness. The summaries are not meant to be comprehensive, but only as a pointer to more detailed sources.

General Design Theory (GDT) was proposed by Yoshikawa [86] and later expanded by Tomiyama and Yoshikawa [79]. It is one of the first design theories at the knowledge level—a concept originally proposed by Newell [56] in the context of computational theories. GDT describes design as a transformation between two spaces—function and attribute, and discusses the nature of this transformation in relation to availability of complete and incomplete knowledge.

Axiomatic Design Theory was proposed by Suh and colleagues [74, 75]. It describes design as a transformation between functions and parameters, and argues that good designs can be described by two axioms: axiom of independence and axiom of information content. According to Axiomatic Design Theory, the less coupled the functions are in a design and the less information content the design has, the better it is.

Another Knowledge Level theory—${\text{K}}^{\text{L}} {\text{D}}^{\text{E}}_{0}$—was proposed by Smithers [69, 70]. This theory was tested by the author on design of a new font that the author himself designed. ${\text{K}}^{\text{L}} {\text{D}}^{\text{E}}_{0}$ distinguishes six types of knowledge needed in design: 1. knowledge needed to form requirements, knowledge of the requirements descriptions actually developed, and their associated justifications; 2. knowledge of how to develop well-formed problem descriptions and knowledge of the well-formed problem descriptions developed and their justifications; knowledge needed to solve well-formed problems, and the knowledge of the solutions and justifications actually formed; 4. knowledge needed to analyse and evaluate problem solutions, knowledge of the analyses and evaluations actually performed together with their justifications; 5. knowledge needed to form design descriptions, and the knowledge of the actual design descriptions and justifications; 6. knowledge needed to construct design presentations, and the knowledge of the presentations actually formed and their justifications.

A quest for a Universal Design Theory (UDT) was made by Grabowski et al. [37, 53]. UDT is attempted to be a design theory containing findings and knowledge about design from different engineering disciplines in a consistent, coherent and compact form [52]. It is aimed at serving as a scientific basis for rationalizing interdisciplinary product development. The aim of UDT is to provide models of explanation and prediction of artefacts and away of designing them. The theory takes the ‘process of design as the mapping of a set of requirements onto a set of design parameters’ that constitute a design solution. The process is proposed to be carried out in by transition through four linked, abstraction levels: modelling requirements, modelling functions, modelling effective geometry, and embodiment design. A design solution is a specification of information sets associated with levels of functions, effective geometry, and embodiment. UDT proposes three axioms: the first states that there is a finite number of levels of abstraction; the second axiom states that the ‘the set of well-known basic elements on each level of abstraction is finite at a certain point of time’; the third axiom states that ‘the number of transitions between the different levels of abstraction is also finite’. Based on these axioms, the authors considered that ‘Elements of a design theory…can only include the components currently known to us whereas the invention of new effects etc. has to be the concern of research work’. In line with this, they hypothesised the following: ‘The invention of a product is always a new combination of known basic elements’, and that ‘Discovery, achieved through research, is defined as the finding of new basic elements’. In this sense, the scope the universal design theory is limited to those types of design where new designs can be seen only as a combination of old basic elements.

Based on the methodological framework used for the development of Grabowski’s universal design theory [52], Lossack [50, 51] proposes the foundations of a Domain Independent Design Theory. The theory describes design knowledge, design process knowledge and system theoretical approaches for processing this knowledge system. The underlying concept consists of three elements: object patterns, process patterns and design working-spaces. Lossack emphasises that ‘design is not a workflow […] workflows represent processes in a deterministic manner, whereas design is intrinsically indeterministic’. He therefore proposes an approach based on solution patterns to support indeterministic design processes, which include solution finding processes and creativity. A solution pattern is an aggregation of an object and a process pattern, although an object pattern can be used without process patterns. Object and process patterns describe design knowledge with which a mapping between properties of the design stages is defined. To define the design context, design working-spaces are introduced [36]. A design working space is a system (with elements, relationships and boundaries) which builds a framework to support the solution finding processes with object and process patterns. The approach is regarded to be general enough to support designing in mechanical, electrical and software engineering.

The theory of synthesis by Takeda et al. [76] focuses on the properties that the synthesis process should have as a thought process and propose a theory for synthesis. Knowledge for synthesis in design, they argue, ‘needs physicality, unlikeness, and desirability’. Physicality ensures possibility of existence, while unlikeness and desirability ensure newness and value. The theory is based on the assumptions that a design process is an iterative logical process of abduction and deduction on design solutions, their properties and behaviours, and knowledge of objects. The synthesis theory for design is defined as a process of reconstruction of design experiences, where each experience contains a logical design process having three steps: ‘collecting design experiences, building a model that includes the collected design experiences, and minimizing an element that designers want to find newness’.

Infused design [66] is an approach for ‘establishing effective collaboration between designers from different engineering fields’. Infused design provides representation of the design problem at a mathematical meta-level that is common to all engineering disciplines. The problem solving is carried out by using mathematical terminology and tools that, due to generality, are common across design disciplines. The meta-level proposed consists of general discrete mathematical models termed combinatorial representations (CR). In particular, Infused design demonstrates ‘how methods and solutions could be generated systematically from corresponding methods and solutions in other disciplines’, and ‘guarantees the correctness of results by relying on general ontology of systems that is embedded in the different representations’. Taura and Nagai [78], in their systematised theory of creative concept generation in design, proposed a theory on the thinking process at the ‘very early stage of design’, they define as the phase that ‘includes the time just prior to or the precise beginning of the so-called conceptual design’. They segregate concept generation into two phases—the problem-driven phase and the inner sense-driven phase. They found that the concept generation process could be categorised into two types: first-order concept generation, which is related to the problem-driven phase, and high-order concept generation, which is related to the inner sense-driven phase.

Appendix C: Overview of Theories, Models and Key Concepts Proposed by the Authors

As discussed in Sect. 1.3.5 some authors have proposed ontologies for the development of their theories and models, others have defined their main concepts but not yet put these together into an ontology. In this section, we summarise the proposed theories or models and the related key concepts. What is immediately visible is the differences in concepts used, as well as the difference in their number. Some overlap in key concepts exists. As expected, this is the case where a theory or model has been built on other theories and models. The differences suggest that the phenomenon of design is (as yet) too large, or maybe its boundaries not fixed enough, to be treated as a whole, as also suggested by Eckert and Stacey [28], Chap. 19.

Agogué and Kazakçi [1], Chap. 11: Concept-Knowledge-theory of C–K theory, a theory of creative design reasoning.

Key concepts: K-space, C-space, logical status, properties, restrictive and expensive partitions, co-evolution of C- and K-spaces through operators (conjunction, disjunction, expansion by partition/inclusion, expansion by deduction/experiments), d-ontologies, generic expansion, object revision, preservation of meaning, K-reordering.

Albers and Wintergerst [2], Chap. 8: Contact and Channel (C&C) Model and Approach to integrate functions and physical structure of a product in a shared representation using product models that are widely spread in practice.

Key concepts: Channel and support structures, working surface pairs, connectors, Wirk-Net, Wirk-structure, operation mode, input parameter characteristic, environmental conditions system state property.

Andreasen et al. [6], Chap. 9: Domain Theory as a systems approach for the analysis and synthesis of products.

Key concepts: Activity, organ, part, structure, elements, behaviour and function, state, property, characteristic, technical activity, need, operands, effects, surroundings, use function, wirk function transformation.

Badke-Schaub and Eris [7], Chap. 17: Understanding the role intuitive processes play in the thinking and acting of designers, to inform their Human Behaviour in Design (HBiD) framework which aims to understand the complex interplay between the designer, the design process, design output, and the related patterns and networks of influencing variable.

Key concepts: Intuition (physical, emotional, mental and spiritual), un/sub-consciousness, reasoning.

Cavallucci [21], Chap. 12: Inventive Design Method based on and an extension of TRIZ theory, to rapidly arrive at a reasonable number of inventive solution concepts to evolve a complex initial situation that is currently unsatisfactory.

Key concepts: Contradiction (administrative, technical, physical), problem, partial solution, action parameter, evaluation parameter.

Culley [24], Chap. 18: An information-driven, rather than task-driven, design process to manage and control design activity.

Key concepts: ‘Information as thing’, knowledge (embedded, encoded, encultured, embrained, embodied).

Eckert and Stacey [28], Chap. 19: Identifying the causal drivers of design behaviour as a first step to generate partial theories of design.

Key concepts: Constraints (problem, process, solutions and meeting constraints), causal drivers (characteristics of classes of products or processes, conditions in which they are created), and requirements.

Eder [29], Chap. 10: Theory of Technical Systems and an engineering design methodology based on this theory.

Key concepts: Transformation process (operands and related states, effects, operators, technology, assisting inputs, secondary inputs and secondary outputs, active and reactive environment) and Technical System (function, organ, organ connector, constructional parts and their relationships: functional structure, constructional structure), life cycle of a technical system (a sequence of transformation systems), properties of transformation processes and technical systems (observable, mediating, elemental) and their related states.

Gero and Kannengiesser [33], Chap. 13: The Function-behaviour-structure (FBS) ontology to describe all designed things, irrespective of design domain, the FBS and the situated FBS (sFBS) frameworks to represent the process of designing, and its situatedness, respectively, irrespectively of the specific domain or methods used.

Key concepts: Function, behaviour (expected, derived from structure), situatedness (interactions between external, expected and interpreted world), interaction (interpretation, focussing, action), function, requirements, structure, design description, transformation (formulation, synthesis, analysis, evaluation, documentation, reformulation types 1–3), comparison.

Goel and Helms [34], Chap. 20: A knowledge model of design problems called SR.BID, derived from the Structure-Behaviour-Function knowledge model, and grounded in empirical data about biologically inspired design practice to capture problem descriptions more deeply than with the SBF knowledge model.

Key concepts: Function, performance criteria, solution, deficiencies/benefits, constraints/specification, and operating environment, structure, behaviour and function.

Goldschmidt [35], Chap. 21: A model of the role of sketching in the early, search phase of design.

Key concepts: Problem, search space, internal and external representations, rapid sketch, cognitive benefits and affordances (time effective/fluent, minimal cognitive resources, minimally rule-bound, transformable/reversible, tolerant to incompletion, tolerant to inaccuracy/lack of scale, provides unexpected cues).

Koskelaet al. [46], Chap. 14: The first theory—proto-theory—of design proposed by Aristotle based on the claim that design is similar or analogous to geometric analysis.

Key concepts: Analysis (theoretical and problematical), synthesis, deliberation, science of production, causes (efficient, formal, material and final), types of reasoning (regressive, transformational, decompositional or configurational).

Lindemann [49], Chap. 6: Definition and nature of the variety of models used for design, discussion on quality and requirements for modelling based on important characteristics like transformation and reduction, purpose and subject, and nature of the process of modelling.

Key concepts: Transformation, reduction, pragmatism (purpose, users, time frame), modelling conventions (accuracy, clearness, profitability, relevance, comparability, systematic settings), process of modelling (intention, modelling, validation, usage).

Maier et al. [54], Chap. 7: A cybernetic systems perspective to understand designing as a self-regulated modelling system, i.e. to consider the synthetic role of models in designing.

Key concepts: Sensoring, actuating.

Ranjan et al. [59], Chap. 15: Integrated Model of Designing’ (IMoD) for describing task clarification and conceptual design, and for explaining how various characteristics of these stages relate to one another, by combining different views (or models).

Key concepts: Activity view (generate, evaluate, modify, select), outcome view (phenomenon, state change, effect, input, action, organ, part, other), requirement-solution view (requirement, solution, associated-information), and system-environment view (relationships, elements, subsystem, system and environment).

Sonalkar et al. ([72], Chap. 3): Two-dimensional structure for design theory: describing the theoretical constructs and relationships between them, and providing the perceptual field and action repertoire that makes a theory relevant in situations of professional practice.

Key concepts: Perceptual field, action repertoire, event, relationship,

Taura [77], Chap. 4: A framework composed of the Pre-Design, Design, and Post-Design stages is introduced to allow the explicit capture of the motive of design, as an underlying reason for the design of highly advanced products, that links the Post-Design and Pre-Design stages.

Key concepts: Pre-Design, Design, Post-Design, deductive, inductive and abductive processes, personal/social motive, inner/outer motive, need, problem, personal inner sense, inner criteria, function (visible/latent), force of a product, standard, field (physical/scenic/semantic; visible/latent).

Weber [85], Chap. 16: The CPM/PDD approach to modelling products and product development based on characteristics and properties (CPM: Characteristics-Properties Modelling, PDD: Property-Driven Development).

Key concepts: Characteristics, properties (current, desired), relations, external conditions, analysis, synthesis, solution elements/patterns.