Performative design synthesis
İpek Gürsel Dino
Published Jul 24, 2020
Design process is significant in design activity since it allows designers and architects to be more efficient and effective in their decisions and suggestions. Different teams are able to communicate and collaborate to find solutions for design problems in a design process, which covers the entire sequence of design activities from the beginning of a project till its completion, including individual loops of briefing, analysis, and synthesis. When performance is considered as a guiding parameter for architectural design, design processes become more essential. Although the technological developments in the computer-based analysis a significant influence on performative architecture, the developments of tools that support design synthesis - rather than only analysis - manage to bridge the gap between different disciplines, whose involvements are crucial in performative building designs (Kolarevic, 2010, p249). In such interdisciplinary design processes, computational processes in both analytical and generative modes have a potential for a high level of design synthesis. In this respect, research on computational design synthesis, which aims at combining different fields and focuses on analyzing, evaluating, and generating, has gained acceleration for almost half a century. Computational synthesis differs from traditional design processes since it aims at capturing, optimizing, and utilizing design solutions more broadly. The complexity of performative design processes has increased through recent advances in computation, and the increasing computational power provides significant efficiencies in design generation processes (Cagan et al., 2005). Yet, the level of complexity of the solutions provided by a computationally synthesized model still depends on the designer, which is in contrast with the traditional performance models that use computational power only for finding significant performance parameters. In this sense, the computational design synthesis has the potential to allow interpretations and evaluations made by human designers even though they provide a huge amount of information and alternatives (Wyatt et al., 2011).
Models of computational design synthesis can be divided into four major activities, which include representation, generation, evaluation, and guidance (Cagan et al., 2005). In this respect, computational synthesis activities are not quite different from the design activities that humans follow in their design process. In computational design synthesis; representation corresponds to the mental model of the object that can be represented through sketches, axonometric drawings, et cetera; generation stands for the creation of the parts and the whole as in the physical or digital 3D Models according to certain parameters; evaluation is the analysis of the ways in which the design meets the goals and constraints such as daylight analysis, wind analysis, et cetera; and guidance covers the feedback on improvements to the design for the next iteration as most the computer tools provide. In spite of these similarities, computational design synthesis differs from design processes depending on human decisions in terms of its speed, accuracy and the high number of design alternatives that are essential in performative design processes.
In each design synthesis process, the design space is in need of an ordered structure, in which each instance is a solution to a common design problem (Campbell et al., 2003). The solutions may be pursued through fully realized designs or more abstract visualizations. Although modeling for performative design is a key component in performative design processes, visualization remains a challenge for the effective communication of performance data and analysis results. Visual representations aim at bridging the concrete and the abstract with their communicative and interactive properties (Ewenstein and Whyte, 2009). They help to articulate, exchange, and understand design ideas throughout the design synthesis process. In this respect, another main shift in computational design synthesis occurred in the representations formulated by the developer of the computational design method (Cagan et al., 2005). Within visualized spaces, solutions that have similar configurations are organized proximately which enables the designer to reach, understand, and modify the design model easily. Therefore, in order to examine the design space and solutions, little transformations are made to designs to arrive surrounding solutions. By conducting numerous modifications in design space, a wide range of solutions can be visited. In this sense, the computational representations are used for capturing the forms or attributes of the design space. In general, a number of representational structures are embedded into a single representation. The representation structures that form the design space vary from underlying ordering media such as grids or zones to functional orders and relations that are often present implicitly in design representations (Oxman, 1997). In performative architecture, these structures are expanded to various performance parameters in design models. In computational design synthesis for performative designs, the design space is infinite and includes past, present, and future design states for creating, designing, or inventing in design problems (Cagan et al., 2005, p172). In these systems, the challenge is to effectively find the set of solutions that best meets the demands of the design problem with performative concerns.
Consequently, computational design synthesis provides models for performative designs, which connect analysis, evaluation, and generation processes and represent them within design spaces. These representations enable a dialog between designers and other individuals and integrate different data sources from all disciplines while structuring performance data in order to make modifications. The performative design evaluation process is pursued within the same medium, and all the changes are conducted through computational representation tools. This synthesis process provides direct communication, analysis, modification, and data integration models that structure the backbone of performative design processes.
CAGAN, J., M. I. CAMPBELL, S. FINGER, and TOMIYAMA T. (2005) A Framework for Computational Design Synthesis: Model and Applications. Journal of Computing and Information Science in Engineering 5 (3), pp. 171-181.
CAMPBELL, M. I., J. CAGAN, AND K. KOTOVSKY (2003). The A-Design Approach to Managing Automated Design Synthesis. Research in Engineering Design 14 (1), 12-24.
EWENSTEIN, B. AND WHYTE, J. (2009) Knowledge Practices in Design: The Role of Visual Representations as `Epistemic Objects’’, Organization Studies, 30(1), pp. 07–30.
KOLAREVIC, B. (2010). Performative Architecture: Beyond Instrumentality. Spon Press.
OXMAN, R. (1997). Design by Re-Representation: A Model of Visual Reasoning in Design. Design Studies 18 (4), 329-347.
WYATT, D. F., D. C. WYNN, J. P. JARRETT, AND P. J. CLARKSON (2011). Supporting Product Architecture Design Using Computational Design Synthesis with Network Structure Constraints. Research in Engineering Design 23 (1), 17-52.