Modelling Products and Product Development Based on Characteristics and Properties

Abstract

About a decade ago, a new approach of modelling products and product development processes, based on product characteristics and properties, was presented for the first time—called ‘Characteristics-Properties Modelling’ or CPM (as a base for modelling products) and ‘Property-Driven Development’ or PDD (as concept for modelling product development processes). In a series of subsequent publications, the approach was confronted (and, thus, tested) with several questions so far unanswered in Design Theory and Methodology. This contribution gives an overview over the backgrounds of the CPM/PDD approach, explains its core elements and asks what came out of it.

Appendix: Existing Approaches to Design Theory and Methodology—A Brief Overview

In this appendix, a brief overview over Design Theory and Methodology (DTM) findings and approaches is presented. Although this overview has a slight bias towards those approaches that influenced and/or are related to the work presented here, it might be of interest for others. Therefore, the editors of this book agreed to put it into an appendix.

Quite dominant in DTM are the early procedural models; almost all of them stem from mechanical engineering and were developed in Europe, mainly but not entirely in German-speaking countries. Groundbreaking work has to be attributed to authors such as P.J. Wallace, Hansen, Kesselring, Eder & Gosling, Rodenacker, French, Hubka, Koller, Pahl & Beitz, Roth, Ehrlenspiel, Roozenburg & Eekels [15, 18, 21, 27–29, 32, 33, 36, 38, 47, 51 ³, 53, 54, 67]. Hubka's work was later extended into [16, 17, 34, 35]. The book of Pahl and Beitz was very competently and quite early translated into the English language [48], therefore is well-known in the DTM community world-wide.

Although the initial DTM concepts displayed some differences in background and focus, the researchers succeeded quite early in formulating a common view under the auspices of the VDI (Verein Deutscher Ingenieure, German Association of Engineers). The outcome was VDI-guideline 2222 [65], later developed into the VDI-guideline 2221 as a basic framework for design processes [63]. VDI 2221 was also translated into English [64] and is—alongside the book of Pahl and Beitz—an important reference for many international research activities to this day.

The core of all these approaches is a generic phase model of product development processes which starts with the task clarification, goes through functional and principle considerations and ends with layout and detail design. The aim of the approaches is to provide support for product developers, in particular for the systematic development of solution alternatives, their evaluation (and subsequent decision making) and the systematic detailing of the chosen solutions. In most cases, the books and guidelines focus on original design (or ‘new product development’)—despite the fact that this type of development task is not very common in practice where so-called variational and adaptive tasks prevail.

All approaches provide models of the development/design process; models of the product are not explicitly covered, with the notable exception of Hubka [32], later integrated into Hubka and Eder [34, 35] and nowadays Eder and Hosnedl [16, 17]. Methods and tools to model products and support processes using computers—today of great importance in practice—are not discussed; this is, of course, natural in the older publications, but even more recent articles on DTM do not go very deep into the issue.

Hubka and Eder [35], Blessing [9], Wallace and Blessing [66] and Heymann [31] present very profound insights into these European/German-speaking approaches to DTM and their impacts.

Although the approaches mentioned above basically share similar phase models there are several different ‘schools’ of DTM; they differ in focus, but also in terminology. Therefore, the unification or consolidation still is a frequently discussed topic in DTM (see e.g. [7, 24]).

Based on findings of Hubka [32] and Andreasen [4] introduced, among other things, the concept of distinguishing between structural characteristics, which define or specify the constituents of a technical system, and behavioural properties.

While the models and procedures of developing products/designing according to the European/German-speaking ‘schools’ proved very successful in teaching, in the 1990s a—still ongoing—discussion started why there is less acceptance and application in engineering practice (see e.g. [6, 14, 20]).

A completely different approach to DTM was presented by Suh [57, 58]. His Axiomatic Design Theory is intensively discussed in the academic as well as the practical world of product development/design, especially in the USA. The core of Axiomatic Design is to see designing as a transformation (or ‘mapping’) of information from the ‘functional space’ (represented by a list/vector of functional requirements, FRs) into the ‘physical space’ (represented by a list/vector of design parameters, DPs).

Suh also proposed a very elegant mathematical formalisation of his approach—not covered here. Finally, Suh formulates certain axioms (plus corollaries and theorems deduced from them) that define an optimal design solution; the most commonly known axiom is the independence axiom which requires that in an optimal case each FR shall only be dependent of one DP (leading to a ‘decoupled design solution’).

The ‘European School’ of product development/design and Suh’s Axiomatic Design approach usually are considered as rather incompatible.

Already originating in 1990, John Gero proposed the Function-Behaviour-Structure (FBS) model of designing [22, 23]. In this model, product development/design is a transformation of requirements (function F) into the description of the solution (structure S). Between the two there is the category of ‘behaviour B’ which can be split up into expected behaviour (Be) and structure behaviour (Bs)—i.e. (actual) behaviour as derived from actual structure. A fifth relevant element is the documentation of the result (D). Between these five categories (F, S, Be, Bs, D), there are 8 basic relations of which the activity of developing products/designing is constituted. As the FBS model was developed from considerations towards Artificial Intelligence (and first published in an AI journal) it was only later recognised in the DTM community.

Sándor Vajna, later together with Tibor Bercsey coined the term ‘Autogenetic Design Theory’ (first publication Vajna and Wegner [61], advanced and more detailed version in Vajna et al. [59]). This approach describes product development in analogy to evolution processes in nature, in particular as a continuous co-development of objects, techniques and technologies. Developing products is considered as a constant optimisation process, starting with one or more base solutions (the original population) that are defined by chromosomes (i.e. product characteristics); the process then consists of varying the chromosomes (mutation), thus producing new solutions that have to be evaluated against the requirements (selection).

In the 1980s and 1990s—basically decades after the first approaches to and concepts of product development/design had been published—so-called empirical design studies emerged as a new branch of design research. In laboratory experiments (sometimes using experienced practitioners, more often using students as test persons), the actual process of designing was analysed and studied; experiments in real industrial settings would be even more desirable, but often face many practical and methodical problems. Studies of this type originate in the UK (University of Cambridge), in Germany (TU Darmstadt, TU München) and the Netherlands (TU Delft); they are often performed in co-operation with psychologists and social scientists who can contribute vast methodical know-how in conducting experiments involving humans and in interpreting the results. The outcomes of these studies have always been interesting and challenging, sometimes surprising. One of the earliest contributions came from Hales [25]; the relative large number of publications in this field since that time cannot be cited here in detail.

Before empirical design research, research in DTM had the habit of inventing new procedures, methods, tools and methodologies, but very rarely measuring or proving their necessities and impacts. Blessing and Chakrabarti [10] saw this deficit and presented their Design Research Methodology (DRM). DRM sees the traditional prescriptive activities (i.e. developing procedures, methods, tools, methodologies) framed by two descriptive, often empirical studies: One up front to find out what is needed, one at the end to check whether the measures taken show any improvement. By using the DRM framework, design research is given more rigour, its results become refutable in a scientific sense.

In industrial practice, during the 1980s companies—large and small—started to equip themselves with computer tools to support design, simulation and product data management (CAx-systems). This process is still ongoing, the number of tools increasing, the tools getting more and more complex, gaining considerable influence on product development processes. Already in the late 1990s, Spur and Krause [56] coined the term of the (completely) virtual product and (completely) virtual product development.

However, apart from very few exceptions (e.g. [44, 46, 55]) research in DTM on one hand and in computer support of product development/design processes on the other hand have been very weakly interlinked. DTM tended (and tends) to concentrate on the ‘early phases’ of product development (e.g. functional and principle reasoning), while CAx-systems are particularly successful in the ‘late phases’ (embodiment design, numerical simulation, optimisation). The separation of DTM and CAx development is negative both ways: DTM has largely bowed out of discussing computer methods and tools, therefore has lost competences in this field. At the same time computer methods and tools are being developed, introduced and applied without comprehensive methodical background—not always to the benefit of product developers/designers in practice.

Complementing trends in industrial practice, DTM—originally coming from mechanical engineering—widened its focus: The products considered became more ‘mechatronic’, existing expertise and experiences were transferred to the development of multi-domain products and systems (see e.g. [62]). Next, the development of combinations of material products and services was (and still is) considered (so-called Product-Service Systems, PSS).

The issue of Design for X (DfX) has been broadened considerably in the last couple of decades. In addition to ‘traditional’ topics like Design for Manufacturing and Assembly (DfM, DfA, sometimes combined to DfMA) new aspects were introduced, e.g. Design for Quality (DfQ) and Design for Environment (DfE). However, also here we find rather weak links between general DTM research and the development of DfX guidelines.

Since 2000, the area of DTM has gained new impetus, maybe due to a new generation of scientists having taken over. A remarkable number of new approaches have been introduced and are being discussed in the community.

In 2000, the author and H. Werner presented the approach of modelling products and development processes based on product characteristics and properties for the first time [76] which was only later called CPM/PDD (Characteristics-Properties Modelling, Property-Driven Development). The focus of this first article was looking at support tools for product development/design (CAx-systems) from a new perspective—still an important, but not anymore the major topic of CPM/PDD.

Andreasen and his group developed earlier views [4] into the so-called Domain Theory [5, 26]. ‘Domains’ were defined as a set of dedicated views onto a product (in particular: the domains/views of activities, organs and parts) that are used as the skeleton of a procedure for product development/design.

In 2001 Maier and Fadel formulated the concept of affordance-based design [40], more extensively explained and put into context in Maier and Fadel [41]: ‘Briefly stated, an affordance is what one system (say, an artefact) provides to another system (say, a user). The concept of affordance is relational because of the complementarity entailed between two interacting systems.’ Thus, the user becomes integral part of considerations in a product development/design process. Thus, product and user is seen in context. At the same time, the concept of affordances opens extended views on requirements and properties of products and systems. Finally, the relation concept is extended in order to map affordances against product/system components, using a matrix approach derived from Design Structure Matrix practices [42].

Albers und Matthiesen introduced the Contact and Channel Model (C&CM), see [2, 43]. C&CM takes up earlier work (dating back to Hubka [32]) on working surface pairs as carriers of functions, but concentrates on extensions in two dimensions: First, based on working surface pairs (‘contacts’) and the structures to connect them (‘channels’) new design methods are presented. Second, the new approach is not confined to mechanical contacts and channels (like in the past), but broadens the view to fluidic, electrical, even information flows.

In addition, [1, 3] described the so-called SPALTEN methodology as a comprehensive approach to handle problems of different boundary conditions and levels of complexity. From that, Albers recently developed the approach of ‘advanced systems engineering’.

Hatchuel and Weil [30] introduced their Concept-Knowledge (C-K) Theory; it explains product development as mutual interplay between extending the ‘concept space’ (i.e. simplified: generating solutions) and the ‘knowledge space’ (generating knowledge about the concepts’ behaviour via analysis).

Lindemann [39] presented the Munich Procedural Model (MPM) for product development processes. Among other new ideas, it propagates procedural flexibility and the use of a multitude of methods (including computer-supported methods and tools). Another new focus is on the management of product development processes.