Home | english  | Sitemap | Impressum | KIT

Philosophy of Models in Engineering Design

Philosophy of Models in Engineering Design
Veranstaltungsart:

Workshop

Tagungsort:

ITAS, Karlstr. 11, 76133 Karlsruhe, room 418

Datum:

27.06.17 - 28.06.17

 

Description

Motivation: Engineers interact with their products and processes largely through models. Consequently model-based reasoning takes center stage in shaping our technological future. However engineers only rarely reflect about the nature of these models and how technical possibilities and actions are affected by the models’ properties and characteristics. Thereby models in engineering describe not only the product, i.e. the designed artefact, but also the generating process (via so-called process models). At the same time the models also shape and create both the artefact as well as the process. This clearly distinguishes them from scientific models that primarily aim to describe a certain target system.

The importance of modelling has been steadily increasing over the past decades with improving computer technology. However a further step change is expected in the coming decade with the increasing prominence of industry 4.0, which brings together engineering from different engineering disciplines. Big data will also play an increasing role through the introduction of sensors monitoring and directing the use of the product and connecting products together. One of the engineering approaches to this is model-based system engineering where the aim is to model and simulate product properties and behavior from the onset. However underlying questions about the nature and influence of models have rarely been asked.

Over the last decades or so, there has been a growing body of literature on models in the sciences. Much less research has been done on models on engineering design. The workshop is supposed to fill in this gap.

Focus: With this workshop we aim to bring together design scholars, engineers and philosophers who have worked on model-based reasoning. The guiding question is as to what knowledge can be derived from models in engineering?And building on that: What actions do models afford? These questions are of relevance beyond the model debates as they may shed some light on the classical question as to what distinguishes scientific from engineering practice.

Consequently relevant subquestions include, but are not limited to the following issues

  • How, if so, do results derived from models differ from more descriptive knowledge by means of normativity, functionality, and other?
  • What are the relations between these aspects?
  • What is the relation of process models to product models and thus to the designed artefact?
  • How do process models structure the knowledge-generation in engineering design?
  • What is the role of tacit knowledge in using a model and utilizing models-based result?
  • What role can or should ethical or social values play in engineering modelling?
  • What problems arise from a wrong/incomplete understanding of the role of model?
  • Given that the epistemic status of model is interpreted differently by individuals and communities, what are the substantive barrier that arise for model based system engineering?

Interdisciplinary: The research question we addressed with this workshop is formulated within philosophy of engineering, but it can only be answered in an inter- and transdisciplinary fashion because it requires expertise in both, philosophy and the engineering sciences. The workshop thus invites design researchers, engineers (particularly system engineers) both from industry and academia to discuss epistemic questions together with philosophers of engineering.

Model-based reasoning takes center stage in shaping our technological future. Particularly in the fast developing field of system engineering and in what has been coined industry 4.0 models play a central role. We thus think that an enhanced understanding of what type of engineering knowledge can be created by models and how (and whether) models afford certain technological actions is thus one central question when reasoning about technological futures as done at ITZ. Thereby the focus of the workshop is on epistemic issues, but also ethical ones (via non-epistemic values in modelling) will be touched and thus the question as to what kind of technological future we want to create.

The presented research will be published an edited volume with Springer.

Program

Tuesday, 27.6.2017

12:00-12:45 Lunch (in front of lecture room 418)
12:45-13:30 Intro to Workshop & Presentation of participants (PDF)
13:30-14:15 Marcus Popplow: Engineering Models in Historical Research
14:15-15:00 Spencer Breiner, Ira Monarch, Anne-Francoise Schmidt and Eswaran Subrahmnian: Epistemological Perspectives on Modeling in Engineering and Science
15:00-15:30 Coffee break
15:30-16:15 Chris Paredis: The Value of Modeling in Systems Engineering and Design (PDF)
16:15-17:00 Sjoerd Zwart: Prescriptive Engineering Knowledge & Models; (PDF)
17:30-18:30 Evening lecture by Pieter Vermaas: Two models for philosophy of engineering (PDF)
19:00 Workshop dinner

Wednesday, 28.6.2017

9:00-9:45 Andreas Hein: Breakthrough Technologies: Analogies, scalability, technology-dependent physics (PDF)
9:45-10:30 Pieter Vermaas: Modelling knowledge claims of design methods and their steps, and benchmarking these claims (PDF)
11:00-11:45 Timothé Sissoko, Marija Jankovic, Chris Paredis & Éric Landel: Using Models in Decision Making Process Under Uncertainty (PDF)
11:45-12:30 Sabine Ammon & Rainer Stark: Model development, model pluralism and knowledge generation in product design: A case study
12:30-14:00 Lunch (in front of lecture room 418)
14:00-14:45 Thea Morgan & Chris McMahon: Constructivism and Complexity: A philosophical basis for experimental learning models in engineering design education? (PDF)
14:45-15:30 Wyob Houkes: Models in collaborative design projects: Boundary objects or make-believe? (PDF)
15:30-16:30 Final discussion (moderation by Kilian Gericke)

Speakers and moderators

Speakers, in alphabetic order

Moderators, in alphabetic order

Abstracts (in alphabetic order of speaker)

Sabine Ammon & Rainer Stark: Model development, model pluralism and knowledge generation in product design: A case study

Due to the complexity of future products and the assessment of life cycles, model-based systems engineering integrates numerous models which need to interact with each other in product design. By drawing on a case study, we want to investigate this model pluralism and its impact on the generation of knowledge. We will start with a brief overview of model types and model usage in processes of product development, followed by the analysis of the case study. We will end our contribution with an outlook on the future role of models in engineering design which need to face, on the one hand, challenges of intelligent models and learning algorithms and, on the other hand, the quest for a unified general model.

Spencer Breiner, Ira Monarch, Anne-Francoise Schmidt & Eswaran Subrahmnian: Epistemological Perspectives on Modeling in Engineering and Science

We identify three epistemological perspectives that characterize the use of models. The first can be seen as informed by classical philosophy of science where the emphasis is firmly placed on the relations of theory and experience. In this theory-centric epistemology, models have been cast as mediators between theory and experience where models are seen as disappearing when theory is established or only come back as a conditions of theory applications. This perspective may acknowledge that models have some autonomy from both theory and experience but only in the sense that they are capable of guiding theory to fit experience better.

The second epistemological perspective on models is informed by historical and sociological case studies that challenge the first perspective’s pure conception of science in arguing that engineering is an essential part of scientific experiments and that engineering like science builds theories, in this case of artifacts, and evaluating them according to data from the entire life cycle (from design to use and back). In this perspective science is no longer pure but viewed as a hybrid that might be called techno-science. Moreover, the use of models in engineering practice both helps in formulating and evaluating engineering theories and guides building and use of instruments and other devices to expose new engineering data. Engineering and science mutually interact in a techno-scientific matrix.

The third epistemological perspective challenges the first perspective’s theory centered focus and role of models as mediators of theory and evidence. It also challenges some of its methodological precepts involving repeatability, crucial experiment, control group etc. The focus is on the creation of models, whether natural or artificial, that are no longer mono-disciplinary or theory inspired. Materials for creating these models are the result of decomposing propositions from multi-disciplinary and non-disciplinary participants into heterogeneous, but somehow compatible conceptual fragments. Models are abductively suggested, not necessarily by a surprising empirical result but through the use of conceptual fragments obtained across disciplines that can be used to represent underdetermined objects that is, objects without a key property (e.g., light waves without a medium that sound waves, air and water waves do have). We call this new epistemological perspective generic since it countenances transforming conceptual fragments of minimal compatibility into models of underdetermined objects in a non-disciplinary generic space. New forms of collective intimacy and shared memory (archives) are needed to support multi-disciplinary and non-disciplinary participants in generic space.

We explore these perspectives by discussing their applicability to several short case studies (including the role of models in the wave theory of light, in strength of materials and in the interrelations among climatology, dynamical systems and chaos theory). We analyze and evaluate which perspective or, more likely, which parts of each perspective best characterize the role of models in the cases reviewed. It may be, for example, that the generic epistemology perspective on modeling rather than the techno-scientific perspective characterizes cases that intuitively would be considered purely engineering practice. It may also turn out that certain cases of modeling are best characterized from a techno-scientific perspective and not a pure scientific perspective. We may also find that not only does the generic epistemological perspective enhance the techno-scientific one, but that also the techno-scientific one (in the form of computerized archives) enhances generic epistemology.

Andreas Hein: Breakthrough Technologies: Underdetermination of Models and Gap-Filling

Breakthrough technologies are technologies that introduce radically new capabilities or a performance increase of at least an order of magnitude. Examples are the turbojet, inertial navigation, and autonomous vehicles. However, a remarkable pattern for these technologies is that their feasibility seems to have been almost always initially contested. One characteristic of breakthrough technologies is that they are initially associated with large uncertainties and numerous unknowns. Hence, initial feasibility debates make heavy use of analogies (x is feasible in context y, as it is feasible in context z; self-replicating in nanomachines is feasible, as it is feasible for biological cells (Kurzweil, 2003)), metaphors (x is like y, y has attribute z, hence x can have attribute y; DNA is like the book of life. As a book can be read, DNA can be read), and the reuse of knowledge from different contexts in order to fill knowledge gaps (x is infeasible, as it has been shown infeasible in context y; Continuous aim firing on ships is infeasible, as it has been shown infeasible for a coastal gun (Morison, 1966)). In this paper, we are specifically interested in the context and domain-dependence of mental models in feasibility arguments of breakthrough technologies. For example, assumptions of linear vs. exponential performance improvement of a component technology can lead to different feasibility results for a breakthrough technology. Prominent examples are the recent exponential decrease of solar cell prices and the exponential increase in the durability of light bulbs in the 19th century, making gas light obsolete. By using various historical and current examples, we illustrate, how different mental models for breakthrough technologies can lead to different feasibility conclusions. We focus on the context-dependence of these mental models (domain and model types) and how they might no longer be applicable to changes in the context and argue that gap-filling is a phenomena that is common in decision-making under uncertainty and in particular regarding breakthrough technologies. Work in this area would not only have merit for the philosophy of engineering and technology but also practical implications, as it may lead to methods and tools for supporting companies in evaluating the potential of breakthrough technologies.

References

Kurzweil, R., 2003. The Drexler-Smalley debate on molecular assembly [WWW Document]. Kurzweil - Accelerating Intelligence. URL http://www.kurzweilai.net/the-drexler-smalley-debate-on-molecular-assembly (accessed 4.23.17).

Morison, E., 1966. Gunfire at sea: a case study of innovation. Men, Machines, and Modern Times.

Wybo Houkes (Eindhoven University of Technology): Models in cross-disciplinary design projects: boundary objects or make-believe?

Modelling is a focal point of recent research in the philosophy of science, yet the role of models in engineering design has gone largely unnoticed. By contrast, sociological analyses of design practices have long recognized the importance of models and other epistemic objects. Here, the notion of ‘boundary object’ (Star and Griesemer 1989) has been used to great effect, significantly improving our understanding of how differences in interests, background knowledge and organizational roles are effectively negotiated in cross-disciplinary collaborations (e.g., Henderson 1991; Carlile 2002; Ewenstein and Whyte 2009; Nicolini et al. 2012).

In this paper, I argue that philosophical work on scientific modelling can illuminate aspects of cross-disciplinary modelling practices in collaborative design projects that have not and cannot be addressed in the influential boundary-object perspective.

In particular, I draw on work that regards modelling as authorized games of make-believe (following Walton 1990), in which participants issue prescriptions for imagining properties of possible concrete objects (Frigg 2010; Toon 2012; Levy 2015). Analyzing modelling in cross-disciplinary design projects as such authorized games reveals how different parties communicate commitments and constraints, while leaving open properties for other competent model users to decide on – aspects of design practices that boundary-object analyses necessarily overlook (cf. Stacey and Eckert 2003). I conclude by identifying some ways in which a make-believe analysis of modelling in some design projects needs to be developed: specifically, to account for the non-epistemic aim of design projects and their possible adversarial character.

Throughout, I use a case of Building Information Modelling (BIM) technology in architectural design (Verstegen, Reymen and Houkes 2016) to illustrate the shortcomings of the boundary-object perspective and the utility of (and room for improvement in) the make-believe perspective.

References

Carlile, P.R. (2002) “A Pragmatic View of Knowledge and Boundaries: Boundary Objects in New Product Development”, Organization Studies 13: 442-455.

Ewenstein, B. and J. Whyte (2009) “Knowledge practices in design: The role of visual representations as 'epistemic objects'.” Organization Studies 30: 7-30

Henderson, K. (1991) “Flexible Sketches and Inflexible Data Bases: Visual Communication, Conscription Devices, and Boundary Objects in Design Engineering”, Science, Technology & Human Values 16: 448-473

Frigg, R. (2010) “Models and Fiction”, Synthese 172: 251-268

Levy, A. (2015) “Modeling without Models”, Philosophical Studies 172: 781-798

Nicolini, D., J. Mengis and J. Swan (2012) “Understanding the Role of Objects in Cross-Disciplinary Collaboration”, Organization Science 23: 612-629

Stacey, M. and C. Eckert (2003) “Against Ambiguity”, Computer Supported Cooperative Work 12: 153-183

Star, S.L. and J.R. Griesemer (1989) “Institutional Ecology, ‘Translations’, and Boundary Objects”, Social Studies of Science 19: 387-420.

Toon, A. (2012) Models as Make-Believe. London: Palgrave-Macmillan

Verstegen, L., W. Houkes and I. Reymen (2016) “Configuring the Technology-Organization Relationship: Collective Affordances of Digital Technology in Creative Work Practice”, Proceedings of the 32st EGOS Colloquium, July 7-9, 2016. Naples, Italy (pp. 1-29). Naples: EGOS

Walton, K. (1990) Mimesis as Make-Believe. Cambridge, MA: Harvard University Press

Thea Morgan & Chris McMahon: Complex constructivism as a philosophical basis for engineering design education and practice

In recent years there have been calls in engineering design education for an increased recognition of the socio-technical context of design, asking what theoretical models could be used for design education and questioning the epistemology of design practice and its influence on methodology. These reflect similar calls in the philosophy of design, recognising the ethical and political context of design and the constraints placed on design by social institutions. These considerations suggest that the predominantly positivistic stance in engineering design education should be challenged. This talk will present a view, based on detailed ethnographic studies of undergraduate engineering design group projects, that design education should be grounded in notions of complex constructivism. Modelling natural design activity as a form of constructivist inquiry, akin to case study research methodology, could help promote the reflective acquisition and utilisation of tacit design knowledge, and reveal the influence of the self in design and associated ethical and political implications.

Chris Paredis: The Value of Modeling in Systems Engineering and Design

In this paper, the role of modeling in systems engineering and design (SE&D) is considered from a normative perspective. Rather than describing how models are currently used in SE&D practice, we aim to identify how models should be used.

We start the discussion by first recognizing that it is human nature to aim to improve one’s situation. Herbert Simon captured this in his view of the human-as-designer who “devises courses of action aimed at changing existing situations into preferred ones.” Noting that human preference can be measured as value, this aim can also be expressed as value maximization. The second starting point for the discussion is the observation that the activities necessary to transform a current situation into a preferred one require resources, resource that are themselves valuable and therefore need to be considered when pursuing value maximization. SE&D is therefore not just about arriving at a more preferred situation, but also about getting to this situation in an efficient manner.

This is why designers use models — because models add value. Models allow us to transform situations efficiently and effectively through planning. As Benjamin Franklin once recognized: “If you fail to plan, you are planning to fail!” Models serve as plans for action. They allow us to specify in detail how to go about transforming the world, without actually expending the time and resources required for the transformation process. In addition, models allow us to think through the consequences of a proposed transformation process so that we can gain confidence that the expected outcomes are indeed preferred over the current situation. While it is impossible to predict with certainty what the outcomes will be, models do allow us to reduce the uncertainty and therefore increase the expected value significantly.

Planning is best performed incrementally through elaboration. From an initial intuition about a potential value opportunity, engineers gradually develop and refine a plan for how best to take advantage of this value opportunity. Over the ages, engineers have come up with strategies for how to develop such plans efficiently. They have identified modeling formalisms that allow plans to be expressed at different level of abstraction and from different viewpoints. These formalisms support a gradual refinement of the specification of the plan, and correspondingly, a gradual improvement in the accuracy of the predictions of the consequences. In this paper, we will explore different models, modeling formalisms and the mechanisms through which these models add value in systems engineering and design. From a normative perspective, the paper supports the claim that systems engineers and designers should use those models that are most valuable.

Marcus Popplow: Engineering Models in Historical Research

In the last two decades, the employment of material models in engineering and architecture has raised considerable interest among historians of technology, historians of science, and architectural historians. In this context, a number of classifications has been proposed to systematize the epistemic functions of such models as well as their various modes of employment. At the same time, the multitude of functions one and the same model could fulfill has often been highlighted.

The contribution attempts to give an overview of such classifications to stimulate discussion on which aspects might be useful for the research questions raised by the Workshop. To a certain extent, emphasis will be given on the functions of engineering models in the early modern period, as my own research has primarily focused on this historical period.

Timothé Sissoko, Marija Jankovic Chris Paredis, Éric Landel: Using Models in Decision Making Process Under Uncertainty

In the automotive industry, as in complex systems design industries more generally, the design process can be seen as a series of decisions largely supported by simulation. During the development phase, models enable the investigation and prediction of architectural consistency and overall vehicle performances. The corresponding design decisions are often based on these modeling and simulation results. Through an iterative process, a vehicle synthesis model is refined by adding detail to the specifications until a physical prototype can be manufactured. Although the physical test results are supposed to be consistent with the M&S predictions, this is not always the case in practice. The difference between the predicted and the actual outcomes is due to uncertainty in the form of lack of knowledge. Consequently, new issues are often discovered late in the cycle, when the cost of solving these issues is high.

From the decision maker’s perspective, uncertainty is partly related to the lack of information or understanding of certain M&S characteristics, such as the assumptions, the precision, the fidelity, the robustness of results, etc. In addition, the data consistency, the framing of the problem, or the feasibility and likelihood of success of the alternatives.

Performed in a multinational car manufacturing company, our research aims to support decision-making processes for vehicle development based on modeling and simulation. We first conducted an observational study for the purpose of identifying the difficulties in existing decision making. Initial observations underscore the need and the difficulty of formalizing and integrating different types of uncertainties in the M&S processes in support of vehicle design.

Our future work includes a comparative study of two different uncertainty characterization methods in order to investigate their overall efficiency and acceptability. Questions related to this comparative study necessarily include uncertainty types that need to be modeled, the relationship between the vehicle decision making and M&S process, or the confidence a decision maker has in a given model. Hence, research towards a credibility assessment of simulation results will be considered. Additional questions pertain to the best representation, the level of granularity, and the level of precision for a given decision situation. This leads to the final question of how models are or should be interpreted according to their representation of the target system.

Pieter Vermaas: Modelling knowledge claims of design methods and their steps, and benchmarking these claims

In my contribution I aim to model the knowledge claims represented by models of design processes as given by design methods. I do so for both design methods as a whole as well as for the design steps these methods are composed of. Finally I make a connection to the idea of benchmarking design methods.

Design methods regularly model design processes that consist of sequences of design steps. Through descriptions of these steps, and a warning that in actual design the steps may have to be carried out in more iterative ways, the design method is then defined. The accompanying knowledge claim is that the method works, a claim that is typically supported by some cases of applications of the method that have led to successful outcomes.

Using the framework of design science research by Joan van Aken, I first analyze this knowledge claim about a method M in three separate propositions, a first about effectiveness and two about efficiency:

M enables designers with specific experience E to find a specific design D and also other designs {D'}
Other methods {M'} enable designers with specific experience E to also find a design D, and M does so more efficiently than {M'}
Other expertise {E'} also enables designers to find a design D with M, and the expertise E that M requires is minimal as compared to {E'}.

Let a method consist of n steps S1 to Sn. The knowledge claims represented by a step Si can similarly be captured as:

Si enables designers with means X to find a specific outcome Y and also other outcomes {Y'}
Other steps {Si'} enable designers with means X to also find outcome Y, and Si does so more efficiently than {Si'}
Other means {X'} also enable designers to find outcome Y with Si, and the means X that Si requires is minimal as compared to {X'}.

Benchmarking design methods means comparing these methods and their steps, yet can have two opposite aims: comparison for singling out the better candidate, or comparison for finding improvements for each of the methods or steps. The above articulation of knowledge claims enables realizing both aims of benchmarking. The propositions 2, 3, 5 and 6 can be a basis for falsifying design methods and their steps. And the propositions 5 and 6 can be a basis for improving design methods by making their steps more efficient.

Sjoerd Zwart: Models to Justify Prescriptive Knowledge

Deepening our comprehension of engineering epistemology and methodology requires a distinction between descriptive (DK) and prescriptive engineering knowledge (PK) (Vincenti 1990 and Mokyr 2002). A lot has already been written on the DK-modeling relationship, but that between modeling and PK remains hitherto mostly unexplored; the DK-vs-PK distinction is mostly ignored in the modeling literature. In this paper I address the question how design engineers use models and modeling techniques to justify PK where engineering is taken to be problem solving in the development of technological innovation.

To do so I first characterize PK in a sufficiently precise way and contrast it with other categories featuring in recent engineering knowledge taxonomies (Meijers and Kroes 2013, Hansson 2013); these are: means-end and functional knowledge (not normative), know how knowledge (subjective); tacit knowledge (not expressible); and ad-hoc knowledge (not generalizable). PK will be illustrated with maxims of Sadi Carnot (efficiency of steam engines), Reginald Fessenden (radio transmissions) and Froude (extrapolation method). The case is made that engineering practice coerces engineering epistemologists to recognize PK as the specific knowledge category.

Next I will take the instrumental view on models according to which models are (should be) designed just like other artifacts (goal identification; working principle; prototype; model behavior; and assessment). In engineering design, models have many types of goals and the one studied here is PK justification. Three PK properties hamper its justification: its means-end character, normativity and context dependence. These aspects render PK justification different from DK confirmation. Finally, by examining various real life case studies, we will learn what roles models can play in gaining trust of design engineers. Sometimes investigators use computational models as a playing ground to determine the consequences of alternative solutions to a design problem (simulations as computational experiments). Alternatively, agent-based models are applied to validate rules of social technical system interventions. Another strategy is to model causal structures to identify the appropriate point of intervention within a process. An interesting outcome is that computation models provide a combination of an empirical-like support from below (by implementing the context) and theoretical support from above (by using standard scientific theories) (Niiniluoto 1993). Thus, PK is “grounded” (Bunge 1966) by computational models into scientific theory.

Connections with the philosophical modeling and simulation literature fall outside the scope of the presentation but will be amply referred to in the underlying paper. For instance the justification of PK is much more similar to Winsberg’s (2010) reliability account, illustrated by computational workarounds in Computational Fluid Dynamics, than the make-believe account that features in various forms of fictionalism described by Weisberg (2013). The latter account is to “make sense of what scientists are doing when they model the world without positing any object that satisfies their modelling assumptions.” (Toon 2012 p. 34) Additionally, the contrast between controlled experiments and computational simulations will be applied to the justification of PK by modeling in engineering.

Bunge, M. (1966). Technology as Applied Science. Technology and Culture, 7(3), 329–347.

Hansson, S. O. (2013). What is Technological Knowledge? In I.-B. Skogh & M. J. D. Vries (Eds.), Technology Teachers as Researchers (pp. 17–31). SensePublishers.

Mokyr, J. (2002). The gifts of Athena: historical origins of the knowledge economy. Princeton N.J.: Princeton University Press.

Meijers, A. W. M., & Kroes, P. A. (2013). Extending the Scope of the Theory of Knowledge. In M. J. de Vries, S. O. Hansson, & A. W. M. Meijers (Eds.), Norms in Technology (pp. 15–34). Springer Netherlands.

Niiniluoto, I. (1993). The aim and structure of applied research. Erkenntnis, 38(1), 1–21.

Toon, A. (2014). Models as make-believe: imagination, fiction and scientific representation. Place of publication not identified: Palgrave Macmillan.

Vincenti, W. (1990). What engineers know and how they know it : analytical studies from aeronautical history. Baltimore: Johns Hopkins University Press.

Weisberg, M. (2013). Simulation and similarity: using models to understand the world. New York: Oxford University Press.

Winsberg, E. B. (2010). Science in the age of computer simulation. Chicago; London: University of Chicago Press.

Pieter Vermaas, Delft University of Technology: Evening lecture: Two Models for Philosophy of Engineering

In this talk I present two models for doing philosophical research on engineering. Model 1 is doing philosophical research on engineering for addressing topics in philosophy. Model 2 is doing philosophical research on engineering for addressing topics in engineering. I argue that both models may lead to results engineering researchers are hardly interested in. The ways in which philosophers address their own topics need not be useful to engineering researchers, making Model 1 primarily one for contributing to philosophy. The topics that philosophers deem relevant to engineering may alienate engineering researchers; Model 2 becomes only productive when philosophical research focuses on topics that engineering researchers find relevant.

For giving the argument I consider efforts in philosophy of technology to analyze the engineering concept of technical function. This analysis was part of the Dual Nature program at Delft University of Technology, aimed at capturing the ontology of technical artefacts. The motivation was simple: technical functions are related to both the physical properties of technical artefacts and the aims of their designers and users. Hence, an analysis of this concept would give a starting point in understanding the relation between the structural and intentional nature of artefacts. Capturing the engineering meaning of technical functions proved however a harder job than anticipated, since engineers give different meanings to functions, which they may even use sequentially and simultaneously. A Model 1 response was to fix a philosophically precise meaning for technical functions, which turned out to be of limited interest to engineering. A Model 2 response was to give a precise analysis of the different meanings that technical functions have in engineering, and turned out to irritate engineering researchers. Results became appreciated by engineering researchers only when this Model 2 analysis was extended to one of understanding why engineers do attach different meanings to the concept of technical functions.

Philosophy of engineering has more topics to address than technical functions, such as the use of models in engineering. I end with extending the argument to research on these further topics.

Organizers

The workshop is financed by the ITZ, the Institute of technological Futures at the KIT.